Nontourism Services in Barbados: 1970-97.

DARRIN DOWNES (*)

This paper initially refers to some of the classifications of services used by various authors, followed by the components that comprise nontourism services in Barbados and their linkages to the other sectors of the Barbados economy. Using regression analysis, variables that

may influence nontourism services in the long and short run are then identified. In the long run, the variables that impact nontourism services are relative real incomes, tourist arrivals, merchandise imports, commercial bank credit, relative prices, and government policy. The short run is influenced by relative real incomes, tourist arrivals, and merchandise imports. Since tourist arrivals influence nontourism services in the long and short run, Barbados must still focus on tourism activity to maximize its foreign exchange potential from nontourism services. (JEL O50)

Introduction

Prior to 1970, earnings from exports of goods were the primary source of foreign receipts in Barbados. Since then, services have earned more foreign exchange than sales of goods that have been domestically produced. Undoubtedly, travel has been the major category of these source items, accounting for approximately 65 percent of service inflows. Over the years, the major export markets have been the United Kingdom, the U.S., Canada, and Caricom. However, as the Barbadian tourism product matures to a stage where rapid growth is limited, it will become necessary to maximize the foreign exchange potential from nontourism services. In this regard, increasing contributions are already being made by the data processing, communications, and financial services sectors, to name a few. However, the remaining categories of services also considerably impact Barbados' foreign exchange earning capacity and therefore cannot be overlooked.

Since 1970, nontourism services have contributed an average of 25.5 percent of total service inflows and 82 percent of total service payments. These ratios originated with the switch from the fourth to the fifth edition of the Balance of Payments Manual [International Monetary Fund (IMP), 1977, 1993]. When the fourth edition was in operation, the ratio of nontourism inflows to service receipts actually fell between 35 and 40 percent. This was because port disbursements (especially sales of oil and bunker fuel to ships and aircrafts) and investment income fell under this category. However, port disbursements are now treated under the goods category while income from investment and personal sources are recorded under the income table, a recent addition to the balance of payments [IMP, 1993; Campbell and Bynoe-Mayers, 1997, pp. 113-38]. Nontourism payments were somewhat lower under the previous edition of the IMF manual, accounting for approximately 55 percent of service payments to nonresidents. However, the ra tio of these payments to total outflows has risen significantly due to the inclusion of a new subcomponent, freight payments, originally captured under merchandise imports. Therefore it is easily concluded from these statistics that nontourism services have generally placed pressure on the balance of payments (BOP) and, by extension, the country's foreign reserves.

This paper analyzes the role of nontourism services in the Barbadian economy from 1970 to 1997. Given the significant contribution of these services to foreign exchange movements and employment, this study provides important insights not only of the origin and nature of nontourism service flows, but also as an indication of their long-term growth trend. There appear to be no previous studies as such on the Barbadian economy.

The remainder of this paper is organized as follows. The second section reviews some of the literature on services as a whole, and the third section examines the components and definitions in the nontourism services sector. The fourth section addresses the relationships between some of the items of nontourism services and the Barbadian economy, and the fifth section describes the data sources and presents some regression results for the long and short run. Conclusions follow in the sixth section.

A Brief Review of the Literature on Services

In the last century, David Ricardo formulated the theory of trade, which essentially dealt with trade in commodities or goods. With the passage of time, considerable changes have occurred in the global economy to the extent that increasing trade in goods has led to higher demand for services, which account for a large proportion of international transactions. In particular, studying the services sector in the Barbados economy is therefore critical to an understanding of the economic health and development of the country [Clarkson and Craigwell, 1997, pp. 29-40].

One of the main problems associated with services is that there is no consensus on its definition. This is because services constitute a heterogeneous bunch of activities. Perhaps the most common characteristic of services is their intangibility. Charles [1994, pp. 2-8] contends that one of the critical issues related to the further development and management of the services sector is that of measurement. According to her, measurement problems abound in services given their peculiar characteristics. These are compounded in the Caribbean by the highly aggregated nature of the data. Furthermore, she states that the presence of the residual category "other services" limits meaningful analysis since data on a number of activities of interest (for example, the nontraditional services) are not readily available. Specific reference is made to offshore banking activities.

Another problem with services is that there is still no universally accepted rule that determines that services are actually internationally traded. This is a consequence of the continuing debate over which characteristics distinguish services from goods [Codrington, 1993, pp. 272-97]. It was even suggested that services sold within the domestic economy to foreigners were not traded [Sapir and Lutz, 1980]. These difficulties have made international transactions in trade more problematic to record than international trade in commodities. They have also affected efforts to standardize the recording of international transactions in services. Accordingly, several classification systems exist, but reference will be made herein to three approaches, namely the functional method, the embodiment method, and the locations approach.

The functional method groups traded services according to the tasks they perform. This method has been promoted and refined by the IMF and is largely used by BOP compilers in its member countries, including Barbados. This allows for easy comparison of data from different countries. This approach also allows for making important distinctions between factor and nonfactor services. A substantial part of service transactions is income earned directly from employing production factors, namely labor and property. Under the previous edition, investments would also have been included. After these deductions are made, the residual items, nonfactor services, are considered a more accurate reflection of trade [Kindleberger, 1963]. Functional classification also allows traded services to be separated into primary services (that is, unskilled labor services such as domestic help, guards, and the like), intermediate services (for example, transport, nonlife insurance, advertising, communications, databases, and business se rvices), and final services such as travel, life insurance, and real estate rentals.

The embodiment approach groups internationally traded services according to physical characteristics with which they are associated when the transaction occurs [Bhagwati, 1984, pp. 133-44; Grubel, 1987, pp. 319-30], thought of as factor-embodied, goods-embodied, and people-embodied services. An example of factor-embodied services is tourism (where the factor is either consumption related or construction when the factor is production related (as regarding a piece of equipment)). In this situation, these services bring together consumer and producer when either of them is internationally mobile for a short period. Goods-embodied services facilitate the use of the service by a consumer who is "separated in time and space" from the producer [Gray, 1989]. An example of this type of service is a computer program that must be embodied in or used with goods at the point of sale. People-embodied services are exchanged with nonresidents through telecommunications, for example, data processing.

The location approach suggests that trade in services may be classified according to locational characteristics [Sampson and Snape, 1985, pp. 171-81], identified as:

1) services that require the supplier to move to the location of the user;

2) services that require the user to move to the location of the supplier;

3) services that are joined locations such as communications; and

4) professional services that are made available to consumers abroad by transmission over communications equipment or by being temporarily embodied in a good, for example, information stored on a computer disk.

A Description of the Components of Nontourism Services

In Barbados, nontourism services presently comprise categories such as transportation, government services (not included elsewhere), and other services. This composition is consistent with the fifth edition of the IME's Balance of Payments Manual, which was published in 1993 but implemented by Barbados two years later.

Transportation

Inmost cases, this category covers transportation services (sea, air, and other) performed by residents of one economy for those of another. Since these transactions result in a negative balance (shown in Figure 1), it puts pressure on the country's foreign reserves. Transportation is further subdivided into passenger services, freight, and other transportation.

Passenger services include, among other things, international transportation of nonresidents by resident carriers outside resident economies and international transportation of residents by nonresident carriers outside their economic boundaries. This is one of the traditional items of nontourism services. In Barbados, such information is provided by airlines, travel agents, and shipping companies. Activity in passenger services results in a net BOP debit since Barbadians incur significantly more expenditure in passenger fares going abroad than when foreigners come to Barbados. Prior to 1995, passenger fares paid by Barbadians represented the major transportation item on the debit side.

Freight transportation is a new service item appearing in this category, the result of the implementation of the fifth edition manual. Previously, on the credit side, this item was one of the components that formed part of miscellaneous services, whereas on the debit side, it was included under imports in the merchandise trade account. On the credit side, information on freight is supplied from direct inquiry, whereas on the debit side, freight is currently estimated at 10 percent of total imports. This 10 percent rate replaced the original rate of 7 percent used prior to 1986. Freight has now replaced passenger services as the largest debit item under transportation and, just as passenger services, leads to severe nontourism service outflows. As a result of the direct link between merchandise imports and freight payments, any movements in imports will directly affect net transportation flows.

Other transportation includes items such as renting vessels, freight cars, and other commercial vehicles with crews for limited periods (such as a single voyage), towing, cargo handling, landing fees, harbor dues, and so on. The landing rate fee and rate of harbor dues are predetermined by the government of Barbados.

Government Services (not included elsewhere)

Government services comprise transactions between the government of Barbados and nonresidents on one hand, and residents of Barbados and the agencies of foreign bodies on the other. Such transactions became important just after independence when a deepening of international links resulted in a rapid increase in the number of diplomats stationed in Barbados. Government services mainly include wage payments to locals (foreigners) by foreign (local) embassies, services under aid programs, real estate rentals, and diplomatic personal expenditure, provided that such expenditure is incurred in the particular country of operation. Although there have been fluctuations in net foreign receipts from government services, this category has remained a net foreign exchange earner for Barbados over the 26-year period under discussion (also shown in Figure 1).

Other Services

This category includes all those services not accounted for in the previous categories. Other services have earned net foreign revenue for Barbados with the exception of six years. Some of the services recorded under this heading include insurance, financial, computer and information, communication, construction, cultural and recreation, miscellaneous business, and management fees.

In the Balance of Payments of Barbados [Central Bank of Barbados, various], insurance services are further subdivided into life insurance and pension funds, freight insurance (originally part of merchandise imports under the fourth edition), and other types of insurance (for example, housing and marine). Presently, Barbados calculates insurance services based on the difference between premiums and claims. However, this is somewhat at variance with the fifth edition of the IMF's Balance of Payments Manual where life and nonlife insurance are calculated differently.

In the case of nonlife insurance, the service charge is recorded in the same manner (in Barbados) as for the exporter, but for the importer, the insurance service charge is computed as gross premiums paid to nonresident insurers multiplied by the ratio of insurance services to gross premiums for exports of services. If claims exceed premiums for the export of insurance services, implying a negative service charge, then the service charge is calculated using a long-term relationship between premiums and claims. The difference between total premiums and the service charge is then treated as a current transfer.

As far as the insurance service on life insurance is concerned, the fifth edition states that service charges are a relatively small percentage of premiums. Hence, if this is the case, the service can be ignored altogether, otherwise ratios from BOP compilers in the other countries can be used to determine this item. The difference between total premiums and the service charge in this case is treated as a financial transaction.

Financial services represent the most important item recorded under other services as far as Barbados is concerned. Such services comprise, for example, commissions and fees for letters of credit, financial leasing services, brokerage services, and the transactions of offshore companies, which are very important to Barbados.

In the 1960s, the first fiscal incentives were introduced to Barbados to attract light industries, such as assembly operations for the computer industry. The government of Barbados was encouraged by the benefits of this initiative to develop an international business and financial sector as part of a strategy to widen the economic base and generate additional revenues, foreign exchange, and employment opportunities. However, according to the Ernst and Young [1994] study, it was not until the mid 1980s that the industry began to develop rapidly, the result of, among other factors, the passing of legislation to govern the operation of offshore banks, exempt insurance companies, foreign sales corporations, and shipping companies. In addition, the study reports that by the end of 1988, the number of incorporations in the international business sector totaled 966. International business companies accounted for 439; foreign sales corporations, 330; exempt insurance companies (EICs), 153; exempt insurance management companies, 39; and offshore banks, 5. Five years later, the number of licensed companies, as recorded by the Financial Services Unit of the Ministry of Finance, rose to 2,338. International business companies increased to 1,171; foreign sales corporations, 926; EICs, 190; and EICs and exempt insurance management companies, 190 and 19, respectively. By the end of 1996, the number stood at 4, 137.

Since the 1990-92 recession, it was estimated that the offshore sector had grown by an annual average of 3.3 percent and generated a significant number of jobs for Barbadians. In the mid 1980s, the sector employed approximately 1,300 persons or 2 percent of total persons employed. By 1997, the number of employed persons in that sector had increased to almost 4,000 persons or approximately 5 percent of the total employed labor force.

Companies incorporated under the International Business Companies Act are subject to rates of tax varying from 1 to 2.5 percent on income earned from their activities. Some of these activities include general trading of goods, financing, publishing, information or data processing, and manufacturing electronics. EICs are exempt from Barbadian taxes and were first established in 1986. These companies are subject to strict reporting requirements and regulations. Audited financial statements must be filed with the supervisor of insurance along with a report from the auditors stating whether the company has met the solvency requirements of the Act.

Offshore banks were established in Barbados in 1979 with the passing of the Offshore Banking Act. This genre of bank differs from a regular commercial bank in that it may not accept deposits from or manage assets of the residents of Barbados. As long as it is granted a license to operate, the offshore bank is required to submit to the Central Bank of Barbados quarterly statements of its assets and liabilities and any returns that may be requested by the Central Bank. The profits of offshore banks are subject to tax rates ranging from 1 to 2.5 percent.

Foreign sales corporations are of U.S. origin and are principally engaged in foreign trade transactions. These companies are not required to pay taxes in Barbados. Instead, they contribute to government revenue through the payment of initial fees ranging from $1,025 to $1,525 in addition to license fees of $500 or $1,000, depending on their size.

Communication services include international settlements for the use of postal, telegraph, telephone, and satellite services. Construction services cover construction and installation work on a temporary basis. On the other hand, management fees take into consideration the share of branches in head offices, expenses of parent companies, as well as payment for services under management contracts.

Computer and information services relate to hardware consultancy, software implementation (that is, data processing, database services, and the like), and maintenance and repair of computers and other related equipment. Other business services cover merchanting and other trade-related services, operational leasing, and miscellaneous business, professional, and technical services. Merchanting is the purchase of a good by a resident from a nonresident and the subsequent sale of that good to another nonresident. During this process, the good in question does not enter or leave Barbados. The difference between the value of goods when acquired and when sold is the value of the merchanting services. In Barbados, the main subcomponent of other trade-related services is agent commissions. Operational leasing (which excludes financial leasing) is the rental of items such as aircrafts, cars, and vessels (without crews) by residents to nonresidents for limited purposes such as a single voyage, particularly in the case o f a vessel. Miscellaneous business, professional, and technical services address transactions between residents and nonresidents in services such as accounting, legal, advertising and marketing research, architectural, management consulting, and other services not mentioned previously.

It is of some interest to examine some of the traditional and new components of other services to see how their foreign activities impact foreign exchange movements in Barbados. Some of the traditional components falling under this category are insurance services, communication services, management fees and merchanting, and other trade-related services, while the new ones comprise financial services, computer and information services, construction services, and personal, cultural, and recreational services. Between 1970 and 1997, the balance on insurance services has been negative except on three occasions, emphasizing the extent to which nonresident insurers have been penetrating the local market. Net outflows moved from a mere $16 million in 1990 to $63 million three years later, mainly because of rising insurance premiums introduced after a spate of disasters within that period.

Communication services have always been a net earner of foreign exchange for Barbados. Indeed, this component picked up significantly in the mid 1980s when it was facilitated by the modernization of telecommunications facilities, which made Barbados the regional leader in this field. Barbados has received net foreign earnings from management fees during 1982 to 1992 but the pattern has been reversed since then, while with the exception of 1986 and 1989, net foreign exchange receipts have been received from merchanting and other traderelated services.

The new sectors have been performing very well for Barbados to date. Financial services have provided Barbados with net foreign revenue of almost $100 million within the last three years, while an average of approximately $35 million can be attributed to computer and information services. Although the foreign exchange inflows from construction services and personal, cultural, and recreational services can in no way be compared with the previously mentioned two services, their activities still result in BOP credits to this country.

The Linkages Between Some Components of Nontourism Services and Other Sectors of the Barbados Economy

While transactions in nontourism services will directly affect the BOP through the current account, their impact on the other sectors of the Barbadian economy cannot be overlooked. As such, some attempt will now be made to see how some of the main components of nontourism services can be related to these other sectors. Since the removal of port disbursements from the transportation account, this table is now dominated by outflows, thereby becoming a strain on Barbados' foreign reserves. Indeed, in 1981, transportation payments to nonresidents were four times higher than inflows, and this ratio has risen sharply ever since then to approximately nine times higher. The major items responsible for this level of expenditure undertaken by Barbadians are freight transport in the first instance and passenger fares in the second. Freight transport is linked directly to the trade account since it is estimated at 10 percent of total imports. However, when an increase in imports occurs, it is reflected immediately in the nontraded sector since this is the sector that uses foreign exchange. This means that an indirect link exists between freight transport and the nontraded sector. In addition, rising imports would also affect the financial sector by way of a decline in liquidity either through an increase in private sector credit or reduced deposits, implying an indirect relationship between liquidity and freight. With a fall in liquidity in the banking system, there would be more reliance on the Central Bank to finance the government's fiscal deficit. Freight and the fiscal account could also be connected if the movement in imports and, by extension, freight was due to higher wages and salaries, which would affect current expenditure, or to the government's capital works program, which would impact on capital expenditure.

Passenger fares paid by Barbadians normally depend on an individual's level of income, deposits, or credit availability. If this income originates from the government's wage bill, then the relationship between the fiscal position and passenger fares goes without saying; whereas the other two items relate passenger fares to the financial sector.

Regarding the components of government services (not included elsewhere), services provided by and to Barbados under aid programs will also be included under current transfers. Earnings from real estate rentals by the government of Barbados will be captured in the fiscal account under nontax revenue while expenditure incurred by foreign diplomats in Barbados will result in a rise in domestic deposits, thereby leading to higher liquidity in the banking system. On the other hand, wage payments made by the government of Barbados to foreign embassy employees in addition to the rental of its overseas embassy office form part of the government's current expenditure and reduce the level of domestic liquidity in the commercial banking system. The final category, namely other services, comprises components that mainly serve to either increase domestic liquidity in the case of inflows or reduce liquidity where there are payments.

Possible Determinants of Nontourism Services in Barbados

Prior to making forecasts on nontourism services, a brief discussion is needed regarding some of the possible explanatory variables of this category. In the absence of any economic theory on this subject, the variables immediately identified are relative real incomes, that is, the ratio of the average real income of Barbados' major trading partners (the U.S., the United Kingdom, and Canada) to real income of Barbados, tourist arrivals, total credit extended to various sectors in Barbados by commercial banks, relative prices (that is, the ratio of the consumer price index of Barbados to the consumer price index of the U.S.), merchandise imports, and a dummy variable representing the periods in the early 1980s and 1990s when the government of Barbados implemented measures to dampen aggregate demand after approaching the IMF for BOP assistance.

A positive relationship is expected between real income and net nontourism services because as real income of the major trading partners expands, which implies an increase in relative incomes, foreigners have more income available to them. This can result in higher demand for nontourism services of Barbados. Tourist arrivals are expected to positively impact nontourism services since an increase in this category by air or sea would result in improved airport and harbor dues from a rise in aircrafts and ships. In addition, such arrivals could lead to greater revenue from package tours, which are captured in transportation receipts. As relative prices rise, say, from an increase in the consumer price index of Barbados, then Barbados' nontourism services become more expensive to the rest of the world. This should lead to a reduction in nontourism service earnings. Rising merchandise imports would be accompanied by higher freight costs, which would, in turn, cause transportation expenditure incurred by residents of Barbados to foreigners to rise and worsen nontourism services. An increase in commercial bank credit would allow Barbadians to purchase additional nontourism services from abroad and reduce net nontourism services. The coefficient of the dummy variable should be positive since the government of Barbados introduced measures in the early 1980s and 1990s to slow down aggregate demand, which should, in turn, cause nontourism outflows to contract and thereby improve the nontourism balance if nontourism credits remain unchanged.

Model Specification

Formally, a model of the nontourism service sector can be expressed as:

NTSN = NTSN(Y+, ARR+, RP-, IMP-, TCR-, DUM+), (1)

where NTSN is net earnings from nontourism services, Y is real income, which is proxied by the ratio of average real foreign income of the aforementioned countries to real income of Barbados, ARR is tourist arrivals to Barbados, RP is relative prices, IMP is merchandise imports, and DUM is a dummy variable representing the 1991-93 period when the government requested IMF assistance for BOP support and implemented measures to dampen aggregate demand in an effort to protect the BOP and Barbados' foreign reserves.

Data and Variables

The data analyzed in this study are annual observations and span almost three decades from 1970 to 1997. Various issues of the Balance of Payments of Barbados and the Annual Statistical Digest, both published by the Central Bank of Barbados, and the IMF's International Financial Statistics database were used to compile the data set. The series for the net nontourism services balance was calculated as the difference between net revenue from total services and net earnings from tourism services. The average real foreign income was proxied by using the average real gross domestic product of the U.S., the United Kingdom, and Canada, whereas real income of Barbados was proxied by using the country's real gross domestic product. The real income and relative prices are estimated at 1990 prices and base year 1990, respectively. The dummy variable will be assigned a value of 1 during the periods when the government requested IMP assistance and zero otherwise. All estimations were performed using the econometric softwa re package Eviews 3.1.

Results

The Engle-Granger [1987] two-step method is being applied in this exercise. First, estimate the coefficients from the cointegrating regression, then take the residuals from the estimate and use them lagged in a vector autoregression of the changes of the explanatory and dependent variables. This is preferred to the Johansen maximum likelihood method since it is more powerful in small samples [Inder, 1993, pp. 53-68].

The augmented Dickey-Fuller (ADF) [1979] test was used to determine the order of integration of each data series. This is a test of the null hypothesis of nonstationarity or a unit root (integrated of order d, I(d), where d [greater than or equal to] 1) against the alternative hypothesis of stationarity (or integration of order zero, I(0)). The ADF test employed different lagged structures as was necessary to eliminate any serial correlation in the residuals of each series.

The ADF tests of each variable indicate that all of the variables are integrated of order 1, I(1). The ordinary least squares long-run regression of model 1 is reported in (2). All of the variables have the correct a priori signs. The long-run results are shown as:

NTSN = -500.3 + 8.15 Y + 0.005 ARR - 3.803 RP

-0.093 IMP - 0.12 TCR + 79.45 DUM, (2)

where [R.sup.2] = 0.76, adjusted [R.sup.2] = 0.69, Durbin-Watson = 1.85, ADF = -4.7532 (-3.7076), and Phillips-Perron = -4.56 (-3.6959).

Equation (2) is the Engle-Granger cointegration regression. Given the small size of the sample, 30 observations, the bias in the Engle-Granger estimator of the long-run relationship could be significant. Therefore, the standard errors and t-values of the estimated regression coefficients are not reported since these statistics are not valid [Banerjee et al., 1986, pp. 253-78].

DNTSN = -31.12 + 0.001 DARR + 10.42 [DY.sub.-1]

The results indicate that all of the explanatory variables are cointegrated since the ADF and Phillips-Perron [1988] tests reject the null hypothesis of nonstationarity of the residuals at the 5 percent level. In addition, the Durbin-Watson statistic does not indicate any first order serial correlation. It is interesting to note that in this model, tourist arrivals play an important role in generating foreign revenue from nontourism services. Indeed, this category will benefit by $5,000.00 from each additional tourist that visits Barbados. A rise of $1 million in relative real income will boost nontourism services by $8.1 million. On the contrary, the data suggest that a nontourism service account will contract by $3.80 for every unit increase in relative prices. Rising imports will result in higher freight payments and worsen nontourism services. However, as Barbadians increase their credit from commercial banks, nontourism services will contract. The positive sign of the dummy variable implies the success t hat the government will achieve in depressing the local demand for nontourism services, therefore allowing net tourism services to improve.

The Granger representation theorem states that if a set of variables are cointegrated, then a valid error correction model of those variables can be formulated. This error correction model nests both the long-run behavior and short-run dynamics. An application of the Hendry general to specific methodology is used here. It begins with an over-parameterized model and, using a stepwise process, eliminates insignificant variables until a parsimonious representation of the model is obtained. The results are shown as:

(-2.46) (3.56) (1.96)

(3)

-0.09 DIMP -0.12 [DIMP.sub.-1] + 0.11 [DIMP.sub.-2] -0.88 [ECM.sub.-1]

(-2.089) (-1.77) (2.17) (-3.18)

where [R.sup.2] = 0.60, adjusted [R.sup.2] = 0.47, Durbin-Watson = 1.92, ADF = --3.612 (-2.990), Phillips-Perron = -4.66 (-3.73), norm = 0.19, Breusch-Godfrey Lagrange multiplier = 0.88, and the autoregressive conditional heteroskedasticity = 0.71 (Engle's kth order autoregressive conditional heteroskedasticity test statistic). The numbers under the respective coefficients are the t-values. D is the first difference operator and norm is the Jarque-Bera [1980] test for normality, based on a test of kurtosis and skewness of the residuals.

The results of the diagnostic tests show that the model is adequately specified and that the residuals do not violate the classical assumptions of normality, homoskedasticity, and serial dependence. Three variables impact nontourism services in the short run, namely tourist arrivals, real relative incomes in the previous year, current merchandise imports, and lagged merchandise imports over a two-year period. The data indicate that nontourism services will benefit by $1,000 from each additional tourist that visits Barbados. The data also indicate a $10.4 million rise in nontourism services from a $1 million increase in relative income in the previous year. In the case of merchandise imports, (3) shows that over a three-year period, nontourism services will worsen by approximately $10,000 for every extra million dollars spent on such items. The error correction term is negative and significant, thereby confirming the long-run cointegrating results, and its coefficient of 0.88 reflects a rapid speed of adjustme nt to its long-run relationship.

Conclusion

This paper has attempted to explain several issues that relate to nontourism services in Barbados between 1970 and 1997. One of the points of this study is that nontourism services for the most part has been generally a net user of foreign exchange for Barbados. Therefore, can nontourism services be relied upon in the future as a replacement for tourism? Based on the results of this paper, the answer to this question is quite interesting. If emphasis is being placed on earning foreign revenue from nontourism services, then tourist arrivals and relative income variables must be targeted both in the long and short run. The mere fact that tourist arrivals to Barbados will impact nontourism service inflows means that Barbados can ill-afford to divorce tourism activity from nontourism activity. Hence, this may suggest pursuing tourism and nontourism service activities with equal vigor.

(*.) Central Bank of Barbados--West Indies.

References

Banerjee, A; Dolado, J.; Hendry, D; Smith, G. "Exploring Equilibrium Relations in Econometrics Through State Models: Some Monte Carlo Evidence," Oxford Bulletin of Economics and Statistics, 48, 1986, pp. 253-78.

Bhagwati, J. "Splintering and Disembodiment of Services and Developing Countries," The World Economy, 7, 2, 1984, pp. 133-44.

Campbell, T.; Bynoe-Mayers, N. "Implementing the Fifth Edition Manual: The Barbadian Experience," Money Affairs, 10, 2, 1997, pp. 113-38.

Central Bank of Barbados. Annual Statistical Digest, various.

___. Balance of Payments of Barbados, various.

Charles, S. "Conceptualizing Services," The Services Sector in the Caribbean Economies: Social Development and Management Studies, Barbados, West Indies: Institute of Social and Economic Research, 1994, pp. 2-8.

Clarkson, B; Craigwell, R. "An Analysis of the Service Sector in Barbados," working paper, Central Bank of Barbados, 1997, pp. 29-40.

Codrington, H. "Exports of Services from Barbados: Insights from Classification, Theory, and International Comparisons," working paper, Central Bank of Barbados, 1993, pp. 272-97.

Dickey, D.; Fuller, W. "Distribution of Estimators for Autoregressive Time Series with a Unit Root," Journal of the American Statistical Association, 74, June 1979, pp. 427-31.

Engle, R.; Granger, C. W. G. "Cointegration and Error Correction, Representation, Estimation, and Testing," Econometrica, 55, March 1987, pp. 251-76.

Ernst and Young. "The Financial, Economic, and Social Impact of the International Business Sector of Barbados," report prepared for the Government of Barbados, 1994, pp. 1-30.

EViews 3.1. Quantitative Micro Software, Irvine, CA.

Gray, P. "Services and Comparative Advantage Theory," in Herbert Gierch; ed., Services and World Economic Growth, Tubingen, Germany: J. C. B. Mohr, 1989.

Grubel, H. "All Internationally Traded Services are Embodied in Materials or People," The World Economy, 10, 1987, pp. 319-30.

Inder, B. "Estimating Long-Run Relationships in Economics: A Comparison of Different Approaches," Journal of Econometrics, 57, 1993, pp. 53-68.

International Monetary Fund. Balance of Payments Manual, 4th ed., Washington, DC: IMF, 1977.

___. Balance of Payments Manual, 5th ed., Washington, DC: IMF, 1993.

Jarque, C.M.; Bera, K. "Efficient Tests for Normality, Homoskedasticity, and Serial Dependence of Regression Residuals," Economic Letters, 6, 1980, pp. 255-9.

Kindleberger, C. International Economics, Homewood, IL: Richard D. Irwin, 1963.

Phillips, P.C.B; Perron, P. "Testing for a Unit Root in Time Series Regression," Biometrika, 75, 1988, pp. 335-46.

Sampson. G.; Snape, R. "Identifying the Issues in Trade in Services," The World Economy, 8, 2, 1985, pp. 171-81.

Sapir, A.; Lutz, E. "Trade in Nonfactor Services: Past Trends and Current Issues," staff papers, International Monetary Fund, 1980.

[Figure 1. Omitted]