Tourism Penetration Index

in Small Caribbean Islands

by

Jerome L. McElroy

Professor of Economics

Department of Business Administration and Economics

Saint Mary's College

Notre Dame, Indiana 46556

e-mail:jmcelroy@saintmarys.edu

and

Klaus de Albuquerque

Professor of Sociology

Department of Sociology

The College of Charleston

Charleston, South Carolina 29424

e-mail:albuquerque@cofc.edu

 

Biographical Note

Professors McElroy and de Albuquerque have had extensive teaching, research, and policy experience in the Caribbean. Their research has focussed on small-scale agriculture, inter-island migration, the implications of political status change and, more recently, sustainable tourism strategies.

Introduction

The post-1960 restructuring of tropical islands worldwide, from colonial export staples to tourism, has been facilitated by a confluence of forces: metropolitan affluence, multinational investment by invitation, aid-financed transport infrastructure, and the advent of low-cost jet travel. Recent research, however, has emphasized the deleterious effects of over rapid and extensive tourist development particularly in the smaller more accessible island destinations (Briguglio, et. al. 1996a). Such negative impacts have included deforestation, erosion and wild life extinction (Lean 1994), near-shore pollution and reef destruction (Robbins 1994; Acharya 1995), land alienation, subsistence disruption, and socio-cultural intrusions that threaten to irreversibly alter insular traditions and identity (Wheat 1995; Mansperger 1995; Lanfant, et. al 1995). The response has been increasing calls for improved monitoring, more effective comprehensive planning, and more sustainable tourism styles (de Albuquerque and McElroy, 1995c; Conlin and Baum 1995; Briguglio, et. al. 1996b).

The Setting and Scope of this Study

The small-island Caribbean was chosen because it represents a subset of relatively homogeneous microstates similar in size and socio-cultural history and emphasizing mainly mass tourism development to achieve economic modernization. Twenty islands met the small size criterion of less than 500,000 in population and 2,000 square kilometers in area (Figure 1). They included the northern British dependencies of Bermuda, Cayman Islands and Turks and Caicos Islands, and the southern Dutch dependencies of Aruba, Bonaire, and Curacao, and most of the Eastern Caribbean (Figure 2) with the exception of Trinidad (large size) and Saba and St. Eustatius (lack of data).

Tourism in the small-island Caribbean has been especially advanced and dominates the landscape. It accounts for a third of all trade, a fourth of foreign exchange earnings, and a fifth of all jobs (de Albuquerque and McElroy, 1995b). Its development, however, has been environmentally invasive for a number of reasons: the fragility of the insular ecosystem (Beller and others 1990) due to a history of deforestation (to plant sugar), erosion and species loss "of exceptional rapidity and severity" (Watts 1993:134), a proclivity for earthquake and volcanic activity and periodic hurricanes and droughts (Barker and McGregor 1995), and a propensity for large-scale transformational infrastructure and resort complexes concentrated along delicate coastlines (Cohen 1978; Pearce and Kirk 1986; Miller and Auyong 1991).

As a result, the region is characterized "by degrees of environmental degradation ranging from significant to extreme" (Barker and McGregor 1995:5). Tourist-induced intrusions have been documented particularly in the Leeward and Windward Islands of the Eastern Caribbean (Figure 1). They include: the scarring of mountain faces with condominium clusters and roadworks (McElroy and others 1990), filling-in of wetlands and mangrove destruction from resort construction (de Albuquerque, 1991; Bacon 1995), beach loss and lagoon pollution from sand mining, nearshore dredging, and hotel sewage dumping (UNEP 1985; de Albuquerque and McElroy, 1995a), and reef damage from diving, yacht and cruiseship anchoring and marina development (Wilkinson 1989). Critics are reassessing the industry's sustainability because of these indelible footprints across the landscape (Holder 1988; Pattullo 1996).

Complex external and internal factors are responsible for this state of affairs. These include: the high-volume growth imperatives of multinational air, cruise and hotel interests that drive international travel (McElroy and de Albuquerque 1996), tourism's import discontinuities whereby short-run economic benefits disguise cumulative longer run costs (Mattheison and Wall 1982), and limited insular infrastructure and planning capacity (Jackson 1986; de Albuquerque and McElroy, 1995b). Complicating matters, because of tourism's pervasive and differential impact on the economy, society, and environment of small islands, there is no single standard integrated measure of tourism penetration (U.S. Congress 1992). (Penetration is defined in terms of scale/dominance). Such a measure could help policy makers gauge tourism's progress and scope and provide an early warning signal for anticipating dangerous thresholds and for mounting mitigating strategies.

The purpose of this paper is five-fold: (1) to construct a simple index of tourism penetration, (2) to apply the index to a sample of 20 small Caribbean islands, (3) to group these islands into subsets based upon levels of tourism penetration, (4) to present four case studies that illustrate the different characteristics of islands at varying levels of tourism penetration, and (5) to highlight the policy implications of the index. To these ends, the paper contains six sections: a literature review, the construction of the index, its application to 20 small Caribbean islands, the grouping of these islands into subsets, four island profiles, and a conclusion noting the wider implications of the index and its importance for strategic planning in small island microstates.

Literature

Most attempts to measure tourism's scale and impact are linked in the literature to the concept of carrying capacity. Some authors stress the conceptual and methodological problems of measuring tourism (Getz 1983) while others search for more precise descriptions of overrun. In the former case, one difficulty involves understanding (de Kadt 1979) and measuring (Mattheison and Wall 1982) tourism's dynamics because of its pervasiveness. Another, involves disentangling tourism's effects from other modernizing influences (Wall 1996). A third concerns capturing the different types of impacts (not necessarily linear nor simultaneous) on the host community: physical, economic, environmental and social (Hall 1974; Cooper 1993). Others argue that the numbers and size of the visitor industry are less important than the different types of visitors and their respective behavior (Wall 1996), what change is acceptable to the host community (Stankey and McCool 1984), and/or the pace or intensity of tourist development (Lundberg 1974a). Despite the long history of interest in quantitatively measuring tourism penetration and capacity, there has been little real advance in this area (Johnson and Thomas 1996).

On the other hand, there are a variety of qualitative definitions of saturation. A sampling suggests most writers mention the appearance of serious environmental, economic and socio-cultural problems. These include infrastructural (water, electricity, etc.) disruptions (Jackson 1986), the displacement of traditional economic activities (Johnson and Thomas 1996), the stage when growth exceeds local labor supply (Kakazu 1994), real estate inflation, congestion and noise (Wall 1982), the increase in man-made attractions to replace lost natural amenities (Butler 1980), escalating crime, prostitution, the steady erosion of cultural traditions, and the appearance of inauthentic cultural attractions (de Albuquerque and McElroy, 1995c; Pattullo, 1996). On the perceptual side, saturation is associated with sustained declines in visitor satisfaction (Lundberg 1974b; UNEP 1985) and rising host hostility (Gray 1974; Doxey 1976; Knox 1982) expressed actively in visitor harassment and passively as a sense of alienation in the face of foreign control (Pattullo 1996). However, this literature has provided limited guidance for objectively measuring how the scale or style of tourism influences these changes.

Actual measures of tourism penetration have been relatively crude and uni-dimensional. Those emphasizing economic impact include tourism's contribution to GDP, the balance of payments, employment, and tax revenues (Bryden 1973; Matthieson and Wall 1982). Other standard indicators include the ratio of visitor receipts to GDP, merchandise exports, outstanding debt, and spending per capita (CTO 1993) plus other common industry indices: average length of stay, hotel occupancy rate, extent of large hotels and foreign ownerships, promotional expenditures and the like (de Albuquerque and McElroy, 1992). The use of such measures is limited because they are stand-alone and single-dimension indicators, that is they are not combined to produce either an overall index of economic penetration or, with non-economic variables, socio-environmental pressure.

The most common indicators of social and environmental pressure are the number of "stayover" visitors (the commonly used expression for over night visitors in the Caribbean) per resident population and land area respectively. For Lundberg (1974a), the "Travel Intensity Index" is simply the ratio of visitors to local population. The more common measure, Harrison's Tourism Intensity Rate (1992), is the number of visitors per 1,000 population and per square kilometer of total or arable land area. The most vigorous indicators are the Tourism Penetration Ratio (visitors x the average length of stay divided by the population x 365) and the Tourism Density Ratio (visitors x average length of stay divided by land area x 365) (de Albuquerque and McElroy, 1992; CT0, 1993). These measures can be interpreted as average daily visitor densities per 1,000 population and per square kilometer. Although all such indicators do tend to correlate closely with visitor expenditure per capita (UNEP 1985; Liu and Jenkins 1996), they usually omit day-trippers and cruise passengers and make no allowance for seasonal and spatial concentration of visitors, features characteristic of island tourism. In addition, they suffer the same single-dimension problems of the economic indicators, and since their impacts are more likely to be considered negative, it is very difficult to establish critical values at which carrying capacity is exceeded (in the past we have subjectively operated with 100 visitors per 1,000 population and 50 rooms per square kilometer). Defining the latter is particularly problematic because at their macro level of abstraction they fail to capture key dimensions of pressure. These include visitor seasonality, the geographic concentration of facilities, as well as socio-cultural settings. Indeed, different natural and socio-cultural settings can sustain vastly different levels of visitation, with some settings reaching saturation at very low levels of visitation. For example, very high visitor densities in Westernized affluent Singapore may be much more tolerable than very low densities in Bhutan (see Harrison 1992).

The few efforts to overcome these shortcomings have attempted to construct more complex and integrated indicators of tourism penetration. Bryden and Faber (1971) developed a visitor density index by ordinally ranking 8 Caribbean islands according to the ratio of tourist to local population. They found the index related positively to an unfriendliness index developed by Zinder (1969) from resident surveys. Likewise Doxey (1976) constructed his Index of Tourism Irritation by comparing visitor levels with results of host surveys in Barbados and Niagara Falls, Ontario.

More recently, the authors (de Albuquerque and McElroy 1992) constructed a three-stage tourism penetration model loosely based on Butler's (1980) life cycle formulation and applied it to 23 Caribbean islands using a dozen standard economic and density measures plus annual rates of change in stayover and cruise visitors. The 23 islands were roughly grouped into three levels of increasing tourism penetration which were called low-density, intermediate and high-density. A follow-up study using only two separate density measures revealed a similar loose three-stage cluster for 38 Caribbean and Pacific Islands (McElroy et. al. 1993). The methodological limitations of earlier work is the basis for developing a more comprehensive index of tourism penetration.

The Tourism Penetration Index

Four criteria guided the construction of the Tourism Penetration Index (TPI) so that it conforms broadly with earlier attempts to measure levels of socio-economic development (McGranahan and others 1972; Morris 1979; and de Albuquerque and D'Sa 1986) and the more recent construction of the human development index (UNDP 1994). The criteria were that the index be (1) simple to formulate using unidirectional indicators (for a simpler normalizing scheme) from standard accessible data, (2) easy to interpret, (3) sufficiently comprehensive to capture the major dimensions of tourism penetration, and, (4) following Briguglio (1995), suitable for wide applicability. To these ends, it was decided to develop a simple index constructed from three separate but unavoidably connected and overlapping sub-indices which measure economic, socio-cultural and environmental penetration.

To measure economic penetration, visitor spending per capita was selected because of its uniformity of estimation, standard use to measure tourism's overall impact, and its close correlation with levels of overall economic development measured by per capita GDP (Liu and Jenkins, 1996:96).

For similar reasons of uniformity, accessibility, and common usage, average daily visitor (stayover and excursionists) density per 1,000 population was chosen to measure social penetration. Although this is only an indirect proxy measure of host irritation (and by assumption reduced visitor satisfaction), and although its level of aggregation masks important seasonal and spatial visitor concentrations, it is useful in capturing host-guest pressures in the small island context. For example, high visitor densities can cause major host irritation on heavy cruise ship days in places like Philipsburg in St. Maarten, Charlotte Amalie in St Thomas (U.S. Virgin Islands), Cruz Bay in St. John (U.S. Virgin Islands), and Roadtown in Tortola (British Virgin Islands), when traffic congestion and parking problems make the conduct of daily business for locals especially difficult.

As a measure of environmental penetration the number of hotel rooms per square kilometer was selected. The choice of a more direct measure was precluded by the unavailability of data. Although this measure does provide some indication of tourism's impact on the physical environment, landscape and infrastructure, this impact is not evenly distributed spatially. In addition, unlike the other two sub-indices, hotel rooms per square kilometer is less sensitive to annual fluctuations in visitor arrivals.

The three separate sub-indices of tourism penetration were selected from a large potential list compiled from the economic, social and environmental indicators reviewed above plus the following additional measures: socio-cultural--population density, crime rates per 10,000 population, crime rates against visitors, levels of visitor harassment, and number of motor vehicles per 1,000 population; environmental--number of motor vehicles per square kilometer, per capita energy production (kwh), per capita solid waste production (kg), annual rate of change in arable land, and annual rate of deforestation. Three indicators were selected from a relatively small subset of measures because appropriate data were not available for all destinations in the sample.

In selecting these three indicators, a priori, the literature and 50 plus years of experience researching and working in small Caribbean islands served and guides. In addition, data on visitor arrivals, visitor spending, and number of hotel rooms are available for most tourist destinations, making the index easy to construct for a large number of destinations. A preferred approach would have been to employ a statistical technique, such as factor analysis, that would help us (1) identify the important components of tourism penetration, and (2) from the factor loadings or factor scores, construct an index (see de Albuquerque and D'Sa, 1986). Neverthelesss, because a large number of indicators of tourism penetration could not be assembled for most islands in the sample, either because the data were not collected, or were missing or incomplete, and given the preliminary nature of this study, it was felt that the a priori determination of the three indicators to be used in constructing the index was methodologically adequate and defensible (see also UNDP 1994; Briguglio 1995).

Several further caveats need to be mentioned. First, some of the indicators that could have been used, such as social irritation and environmental fragility, were excluded because they are difficult to measure. Second, given the frequency of hurricanes and other natural disasters in the region and the substantial impact of such events on tourist infrastructure and demand, it was necessary to choose a year relatively free of major disruptions. The year 1993 was selected because it was considered sufficiently long after hurricane Hugo (1989) and predated the 1994, 1995, and 1996 seasons when the Eastern Caribbean experienced numerous destructive storms and hurricanes (Debbie, Luis, Marilyn, etc.). It also preceded the eruption (August 1995) and continuing ash fallout and pyroclastic flows from the Soufriere Hills volcano in Montserrat. Third, it should be recognized that aggregate data masks individual island differences for archipelagic states, of which there are a number in this small-island sample. They include: Antigua and Barbuda, the British Virgin Islands (BVI), Cayman Islands, Grenada, St. Kitts and Nevis, St. Vincent and the Grenadines, Turks and Caicos and the U.S. Virgin Islands (USVI). Such aggregation also masks seasonal variations in visitor flows (Harrison 1992) and the significant concentration of tourist facilities in island destinations (Pearce 1995). Fourth, caution especially should be used in interpreting the data on the BVI since a significant portion of its tourism is water based (yacht visitors comprise the largest segment of stayovers). To illustrate, an indicator like average daily visitors per 1,000 population tends to overstate the actual social impact on the BVI host population. Finally, in this cross-sectional analysis, none of three measures selected capture the time dimension of tourism such as the length of a destination's experience with or exposure to tourism development.

Constructing the Index

The index was constructed according to the methodology used by de Albuquerque and D'Sa (1986) in developing indices of socio-economic development and by Briguglio (1995) in developing a small-island vulnerability index. The three selected indicators, visitor spending per capita, average daily visitor density per 1,000 population, and number of hotel rooms per square kilometer (see Table 1) were standardized using the following formula:

TPIij = (Xij - Min Xi)

(Max Xi - Min Xi)

Where:

TPI stands for the degree of tourism penetration for the j th island with respect to the th variable, taking on the value of 1 to 3 since 3 variables are used in constructing the index, and j taking on the value of 1 to 20 for the 20 islands in the sample. Xij stands for the value of the th variable for island j. Max Xi and Min Xi stand for maximum and minimum values of the th variable for all islands in the sample.

If a given island has a value of Xij equal to the maximum, its value for TPIij would be one, and conversely if it has a value of Xij equal to the minimum, its value for TPIij would be zero. The resulting standardized values of these sub-indicies are presented in Table 2.

An unweighted average Tourism Penetration Index was defined according the formula below, because weights could not be established on apriori grounds or from guidelines in the literature:

3

TPIij = I ij

=1

This unweighted formulation assumes that each indicator was as important as the other two in contributing to overall tourism impact. A differentially weighted TPI was also constructed by assigning visitor spending per capita twice the weight of the two measures. This was justified on the grounds that economic impact was the most direct measure of tourism penetration among the three indicators.

Results

The TPI scores in Table 2 suggest the weighted (2:1:1) and unweighted (equivalent weights) formulations yield similar rankings since the three sub-indices are highly correlated. The differences in the rankings are negligible and involve only one -position deviations. According to the rankings, the 20 islands are loosely arranged from the most penetrated (St. Maarten) to the least penetrated (Dominica). The three relatively large Windward Islands of Dominica, St. Vincent and Grenada predictably rank at the bottom of the scale and constitute previously identified (de Albuquerque and McElroy 1992) low-density destinations with limited infrastructure and international visibility and small-scale tourist facilities.

At the top of the ranking are the most penetrated high-density islands. They comprise the traditionally popular resort areas of Bermuda and the USVI along with more recent entrants (1980s) into the mass tourism market like St. Maarten and the Cayman Islands. In all cases, these are highly affluent destinations distinguished by their much smaller size and high population densities. As the case study of St. Maarten to follow indicates, these islands tend to be the most tourism-dependent in the sample and exhibit large-scale facilities and an increasing propensity for man-made attractions such as duty-free shopping, gambling, and water sports (de Albuquerque and McElroy 1992).

Like all indices/ranking schemes, the TPI, when applied to small Caribbean islands, produces some anomalies. For example, the BVI's high ranking is the result of the high ratio of daily visitors to local population (226 per 1,000 population, Table 1). As argued earlier, this impact is significantly diminished by the water-based nature of much of BVI tourism. Additionally, rankings are unlikely to be static over the short term or long term, being responsive to a variety of changes and unpredictable events in both origin and destination markets. A case in point is St. Maarten's top ranking in 1993, which may have changed, following the tourism downturn because of the destruction wrought by hurricane Luis in 1995. The same is true for the USVI which experienced the brunt of hurricane Marilyn, also in 1995. Perhaps the best example is Montserrat, whose 1993 ranking has undoubtedly slipped considerably. With continuing eruptions (ash fall and pyroclastic flows) from the Soufriere Hills volcano, visitor arrivals have declined drastically.

Identifying the intermediate group and the precise boundaries between the most and least penetrated islands at one specific point in time is difficult because these destinations are transitional by definition. They are also generally distinguished by rapid growth in arrivals and hotel construction (de Albuquerque and McElroy 1992). In order to more objectively determine island groupings according to levels of tourist penetration, a cluster analysis was performed on the three standardized sub-indices in both weighted (2:1:1) and unweighted (equivalent weights) formulations.

A standard software program (NTSYS-pc 1993) was used to calculate the unweighted average distance between each pair of islands across the three sub-indices. This produced a matrix of pairwise average distances or so-called taxonomic distances (Sneath and Sokol 1973). On the basis of this matrix, the program grouped islands into similar clusters as shown in Figure 3. The initial cluster of all 20 islands is successively split into smaller clusters and when the required number of clusters is reached, islands are reallocated iteratively to the nearest cluster. The clusters formed may be distinguished by levels of tourism penetration (high, intermediate, low) represented by the indicators.

Only clusters using the weighted sub-indices are reported since the unweighted sub-indices yielded almost identical results. The results suggests two primary groupings. These include: (1) the six most penetrated islands identified by the TPI; (2) the rest of the sample. The latter 14 islands are further clustered into two groups--four medium-high penetrated destinations consisting of Anguilla, Antigua/Barbuda, the Turks and Caicos, and Bonaire, and a medium-low to least penetrated group composed of Barbados at the top and Dominica at the bottom. This last group is further broken down into two clusters--the medium-low group of Barbados, St. Kitts/Nevis, St. Lucia, and Curacao, and the least penetrated group comprising Guadeloupe, Martinique, Grenada, St. Vincent and Dominica. Dominica and St. Vincent are at the bottom, while Guadeloupe and Martinique form their own final cluster.

These results broadly confirm direct observations from the TPI (Table 2) and earlier work (McElroy and de Albuquerque 1991; de Albuquerque and McElroy 1992), although some questions remain about the exact boundary between least penetrated, medium-low and medium-high destinations. Some of these issues are clarified in the contrasting experiences of Dominica and St. Lucia in the case studies. However, some anomalies between the rankings and clusters exist, but these are not unexpected given some of the methodological caveats discussed earlier.

The BVI's higher than expected ranking and inclusion in the most penetrated cluster is the result of its small size and the use of data that does not discriminate between BVI's land-based and water-based tourism. To most tourism professionals familiar with the Caribbean, Barbados is clearly more penetrated than the BVI; it has a longer history of tourism and extensive facility investment. Yet, because of its larger population and geographic size, it ranks much below the BVI. Likewise, Guadeloupe and Martinique because of their greater size (population and geographic) and absorptive capacity rank below islands like Anguilla, Turks and Caicos, Bonaire and St. Kitts/Nevis with a shorter history of tourism and with lower levels of infrastructural development (airports, cruise terminals, roads) and visitor facilities (shopping complexes, large hotels, restaurants, etc.). The latter islands have small populations and because two of the sub-indices (economic and social penetration) are based on population the impact of tourism is greater. Anguilla's high ranking is partly due to its high score on visitor density per 1,000 population which derives principally

from the large number of day trippers (twice the number of stayovers, see Table 1), mostly from St. Maarten.

Since the TPI, like the Human Development Index (HDI) and other such composite indices, is based on cross-sectional data, it is useful to present rankings for several years so one can gauge how individual states move up or down based on their scores on the sub-indices. Table 3 compares the rankings of the same 20 islands in our sample for 1988 (a year that preceded hurricane Hugo) and 1993. While there are a few differences, for example,

the Cayman Islands and Aruba move up a few positions, the two rankings are remarkably similar. A Spearman's rank-order correlation coefficient (a measure of association for differences between ranks) computed between the two ranks yielded a value for rho equal to +.96, statistically significant at the .001 level.

Clearly there is short-term stability in the TPI rankings.

Four Cases

Four cases are presented to empirically and descriptively flesh out the different levels of tourism penetration and to highlight the different planning challenges along the penetration continuum. They include Dominica and St. Maarten at the bottom and top of the rankings, and St. Lucia and Antigua/Barbuda representing middle islands in the medium-low and medium-high clusters respectively . Their profiles, constructed with over 20 indicators (Table 4), demonstrate a tourism process of increasing density, scale, and industry consolidation across the four islands.

Dominica is a Windward island situated in the middle of the Lesser Antilles (Figure 1). Its rugged topography, rural ambience, and limited plantation history and urbanization make it an ideal nature destination: "[Its] mostly unspoiled landscape is considered...to be among the most dramatically beautiful and pristine in the world. The country's undisturbed vegetation is more extensive than any other island in the Lesser Antilles...Its forests...undoubtedly the finest in the Caribbean" (CCA 1991:1).

Dominica represents an emerging destination with a small tourism sector. Per capita visitor spending in 1993 was less than $500, and hotel/restaurant activity less than 3 percent of GDP (Table 4). Export agriculture (bananas), small-scale produce, and copra and soap processing accounted for roughly one-half of GDP. Hotel tax receipts are negligible, and there is only one hotel room per square kilometer. During the 1980s population growth was zero indicating emigration was equal to the rate of natural increase. This demographic stability places Dominica far from the high-growth labor-intensive migration transitions characteristic of many rapidly developing intermediate destinations (McElroy and de Albuquerque, 1988).

Growth in stayover visitors since 1970 has been a steady 6 percent per year with most occurring after the mid-1980s when the island's popularity as an ecotourism destination took hold. Dominica's tourism promotional effort has been modest. Annual room expansion since 1984 has been robust and rising, despite the low average hotel occupancy rate of less than 40 percent. Typical of emerging destinations, Dominica has a high average visitor stay (10 days), small-scale facilities, and few man-made attractions to service the small but rapidly accelerating cruise (shopping) trade. Over 90 percent of accommodations are locally owned (Wiley, 1996). Most overnighters stay with friends/relatives or in their own homes while less than 30 percent frequent hotels. There are no white sand beaches, no large hotels (100+ rooms), and no international airport to provide direct access to major origin markets. Only 11 percent of the promotional staff work abroad. As a result, Dominica's visitor profile is dominated by nearby West Indians (especially from Guadeloupe and Martinique) who tend to stay and spend less than North Americans and Europeans who dominate the more penetrated destinations. These two lucrative overseas markets comprise less than one-half of Dominica's profile, an indication of modest development and brand name recognition.

As a newly emerging resort area, Dominica illustrates the major challenge of achieving international visibility. Specifically this means: (1) determining product identity, (2) developing assets appropriate to that image, (3) targeting a desirable visitor clientele, and (4) carefully providing the necessary infrastructure to access its natural attractions. Dominica has progressed far on the first two tasks and enjoys an increasing reputation as a rugged exotic ecotourism paradise. It is expected to be listed soon as a World Heritage Island.

St. Lucia is situated south of Dominica and Martinique (Figure 2) and blessed with a unique blend of natural (climate, beaches, scenery) and man-made assets (shopping, yachting) that increasingly appeal to the mass market. It also boasts a rich history of Franco-British colonial rivalry, plantation sugar, and transatlantic shipping. In the postwar era, the economy has successfully diversified into bananas, textiles, and electronic assembly, tourism and related construction (CCA 1991b). Through aggressive promotion, aid-financed transport/ utility/ communications systems, and investment tax incentives, St. Lucia appears to have achieved international visibility and the high-growth trajectory common to many intermediate destinations. During the last two decades, stayover visitors and new room construction expanded at eight and six percent per year respectively, with a tendency to accelerate in both cases in recent years. In the past decade, cruise passengers increased nearly 16 percent per year. By 1992 the number of cruise passengers equalled stayover visitors. In 1993 there were 43 daily visitors per 1,000 population and 10 per square kilometer, roughly one-half of the tourist density levels of Antigua, in the next higher cluster. To service the labor-intensive needs of this expansion, annual population growth (1.4%) was one of the highest in the region, probably as a result of significantly reduced emigration, return migration, and some immigration.

St. Lucia's tourism sector also exhibits increasing scale and consolidation. Over one-half of all rooms available are in large hotels (100+ rooms). A rising number of stayovers (59%) choose hotels, and there are 8 large hotels contain nearly two-thirds of all room capacity. Fifty-one percent of all rooms are in all-inclusive hotels. The 68 percent average occupancy rate was at the high end for the 20-island sample. This relatively high occupancy rate (by Caribbean standards) is partly due to expanded promotion of the all-inclusives and an increased marketing presence overseas. St. Lucia has double the promotional staff of Dominica and five times the number of employees (10 to 2) in overseas visitor bureaus. Partly as a result of this effort, affluent North Americans and Europeans (40%) make up three-fourths of all island visitors. The island is presently positioned to address the key policy challenge usually faced by intermediate-penetrated destinations, that is, managing sustainable tourism growth. Specifically this will require strong government initiatives to (1) implement comprehensive planning that will ensure development appropriate to the island's image and natural landscape and cultural patrimony, and (2) control the spatial and temporal pace of growth. This growth caution is especially important to reduce valuable asset losses and infrastructural bottlenecks (e.g., traffic). St. Lucia's laissez-faire strategy to date, and its increasing orientation toward all-inclusive resorts, are not promising trends, although the island's modest tourist penetration suggests there is sufficient time to adjust.

Antigua, and its small sister island, Barbuda, lie 250 miles east-southeast of Puerto Rico roughly midway between the Leewards and the Windwards (Figure 2). Antigua's coastline is deeply indented with abundant wetlands, beaches and fringing reefs. The coastal region also contains over 100 pre-Columbian settlement sites as well as rich (and some restored) remnants of sugar mills and fortifications that testify to its plantation past and role as colonial headquarters to the British Navy. As sugar's profitability waned in the first half of the 20th century, some coastal plantations were gradually converted to resort properties (Weaver, 1988). With the advent of jet travel, the demise of export agriculture in the late 1960s, and aggressive promotion and tax incentives, tourism came to dominate the economy and the landscape. Growth was especially brisk in the 1980s. Although it took over three decades to produce the first 100,000 annual stayover visitors (1950-1983), the second 100,000 was achieved between 1983 and 1988 (Butler, 1993). This untrammelled colonization of coastal areas with hotels, marinas and infrastructure has destroyed and/or damaged salt ponds, beaches and archeological sites and polluted harbors and reefs (de Albuquerque and McElroy, 1995a).

Antigua is clearly more tourist-dependent than St. Lucia in terms of visitor expenditure per capita and the density indicators. The visitor industry, related construction and services, and residential development account for over one-half of all economic activity (de Albuquerque and McElroy, 1995a). Daily visitor density per 1,000 population approaches the low end of most penetrated destinations. Since 1988 the annual growth rates of stayover visitors and hotel rooms have stabilized or fallen whereas they have accelerated somewhat for St. Lucia (Table 4). Antigua also exhibits increasing facility size and environmental alteration, the scale of which, implies external market control and other planning problems confronting mature destinations. The island has 50 percent more large hotels (100+ rooms) than St. Lucia, and 90 percent of all rooms available are in hotels. Four out of five stayovers are relatively affluent North American and European visitors.

With a propensity for large-scale resort/marina complexes, high-volume visitation, and abundant promotion and incentives, Antigua will likely cross into the most penetrated mass tourism cluster in the coming decade. This passage will be precarious since Antigua has not yet solved its primary challenge of managing sustainable growth. Despite vigorous efforts of nongovernmental organizations and ecologically aware citizens' groups, illegal sand mining continues, environmental legislation is either ignored or unenforced, and resource monitoring is underfunded (de Albuquerque and McElroy, 1995a). In the absence of a major policy effort to reverse past trends, land and resource use conflicts will likely intensify and coastal assets will be degraded further. In the island's present context of heavy external indebtedness and constrained diversification alternatives, failure to heed the early warning signals may jeopardize the stability of the island's economy still reeling from the after effects of hurricane Luis in September 1995.

Many hotel properties remain closed and occupancy rates have plummeted to 33 percent, yet the Government is proceeding with a massive new tourism development project, involving several offshore islands, and the construction of two golf courses, a marina, and 1,000 villas. This project, to be financed by Malaysian interests, is slated for the still pristine North sound area and has drawn a lot of local opposition.

The challenges facing Antigua are myriad and call for a number of planning initiatives such as (1) initiating and implementing a comprehensive land-use plan to include a master plan for sustainable tourism development and coastal zone management, (2) establishing a new coordinating agency, with Cabinet-level authority, to monitor all development and enforce various regulations, and (3) institutionalizing public participation so that community concerns are reflected in most development decisions (de Albuquerque and McElroy 1995a). Bermuda's model of participatory planning, tourism growth controls, and coastal conservation, demonstrate that policy reversal is possible in small, highly tourism-dependent economies (de Albuquerque and McElroy 1995c). What is lacking in Antigua and Barbuda is a serious commitment to sustainable development by the political directorate.

The Dutch and French island of St. Maarten/St. Martin (Figure 2) represents the classic example of a small tropical island overwhelmed by tourist development in the space of a generation. With the decline of plantation cotton in the late nineteenth century, the island became an emigrant backwater exporting labor continuously, through the pre-war and immediate post-war era, to oil refineries in Aruba and Curacao. Like in the nearby USVI, the U.S. embargo of Cuba, prompted hotel construction in the 1960s. The construction of the 600-room Mullet Bay Resort in 1969/1970, however, transformed the island's small-scale tourism plant and ushered in a take-off with few parallels in the region.

For example, the growth in arrivals more than doubled in the 1960s and 1970s and nearly tripled in the 1980s. The number of hotel rooms more than doubled during the 1970 and 1980 decades. The latter witnessed "...a virtual free-for-all in tourism construction...Any semblance of rational planning soon disappeared..." (de Albuquerque and McElroy 1995c). While most of the early growth was fueled by the draw of the island's beaches and casinos, the 1980s building boom was markedly intensified by explosive growth in cruise arrivals which increased five-fold over the decade lured by duty-free bargains. The burgeoning demand for construction and service labor created a massive migration transition whereby workers streamed back from Aruba and Curacao and others migrated from nearby islands, as well as from Haiti and the Dominican Republic (Samson, 1989; de Albuquerque and McElroy, 1995c). The population doubled in each of the two decades between 1970 and 1990.

According to the TPI, St. Maarten ranks as the most tourist-penetrated in the small-island Caribbean. Population density, including the 262 visitors per 1,000 population on a daily basis, is the highest in the sample. With visitor spending above $11,000 per capita, the island has taken its place as a premier large-scale resort area alongside traditional leaders like Bermuda and the USVI and larger destinations like the Bahamas and Puerto Rico. Over two-thirds of overnight visitors stay in hotels and over two-thirds of hotel accommodations are in large (100+ rooms) facilities (Table 3). Recent experience, however, suggests a slower-growth transition is underway, similar to Butler's (1980) stagnation stage. Since the 1980s there has been a sharp deceleration in annual cruise and stayover growth and a precipitous fall-off in room construction. Visitor satisfaction is falling as average hotel occupancy has declined from 77 percent to 55 percent, and the average length of stay is less than five days.

Past excesses have also exacted a heavy toll on social and environmental stability. Crime has become endemic and there is sporadic industrial strife. Power outages and traffic congestion are common, and haphazard suburban sprawl, derelict cars, litter, uncleared brush and low-income immigrant ghettos have replaced the "natural genius of the place" touted in the promotional literature (de Albuquerque and McElroy 1995c:83). The landscape has been marred by extensive devegetation, hillside and beach erosion, sewage discharge and significant marine pollution in the highly urbanized areas of Philipsburg and Marigot and coastal resort enclaves. Meanwhile, excess capacity, hotel closures, high property turnover and conversions to non-tourist uses depress the investment climate. Decapitalization of the infrastructure continues as government budget and trade surpluses from the previous boom years have been reversed. The passage of hurricane Luis in September 1995 resulted in more hotel closures and significant damage to the infrastructure. By the end of the 1996-97 winter season, some hotels were still closed, but the island had made a significant comeback.

As a lagged response to its lost luster, St. Maarten has aimed an aggressive marketing program at tour operators and stepped up discounting airline seats, hotel rooms and pre-paid vacations. New emphasis has been placed on creating man-made attractions: marinas, cabarets, a tiny zoo and botanical garden; to the point where "imported artificial facilities" present an image increasingly "divorced from its geographical- environment" (Butler 1980:8). As a result, St. Maarten typifies the common syndrome facing overrun high-density destinations: rising densities causing declining amenities and visitor satisfaction spawn increased promotion efforts, new amusements, and even higher densities to stave off falling profits, occupancy rates and market share. This case also illustrates the basic long-run challenge mature highly penetrated tourist destinations must address: sustaining quality. This will require a difficult and sharp reversal of past laissez faire policy and effective government and citizen cooperation to accomplish four things: (1) a new community consensus and planning effort limiting further tourism development to thoroughly vetted and sustainable projects; (2) strong government intervention to control external hotel, airline etc. propensities to violate insular scale; (3) a commitment to resource restoration to correct past mistakes, improve amenity management and mitigate future construction impacts; and (4) the enforcement of building codes and planning regulations (de Albuquerque and McElroy, 1995c). St. Maarten's tourism plant experienced significantly more damage from hurricane Luis precisely because many hotels and tourism-related enterprises had been erected during the building boom when hurricane and other building codes were not being enforced.

Conclusions

The post-1960 tourist restructuring of Caribbean island microstates has produced two disturbing trends: (1) the older more affluent and successful destinations have sacrificed social and environmental stability in favor of bottom line economics, and (2) the newer destinations seem bent on the same non-sustainable mass-tourism path. One aspect of the problem is limited understanding of tourism's complex dynamics, and this stems partly from the absence of comprehensive integrated quantitative measures of tourism's pervasive economic and socio-environmental impacts. Without an early warning system in place, island decision-makers rush to embrace the varied economic benefits of tourism but fail to anticipate the destructive intrusions of mass tourism practice: in particular, how the interplay of inadequate facility, infrastructure and amenity planning and management, with the high-volume propensity of capital-intensive travel interests, tends to cumulatively overrun the delicate insular carrying capacity.

The Tourism Penetration Index developed here is a preliminary response to the problem. Despite its simple construction and inevitable methodolgical flaws, it represents a first effort to measure tourism penetration and is certainly not limited to the small-island Caribbean. Like the first construction of the HDI, which elicited considerable controversy over its analytic and statistical basis, criticism is anticipated over the choice of dimensions of penetration (economic, social, and environmental), selection of indicators, and the lack of firm guidelines to establish threshold levels for two of our indicators--average daily visitors per 1,000 and hotel rooms per square kilometer. The debate this article will engender, should lead to constructive suggestions for the formulation of a more sophisticated index.

The TPI roughly clustered the 20 small islands into four groupings from low to high tourism penetration (technically two initial clusters, most penetrated and less penetrated, were first identified). The four case studies drawn from these four clusters detailed the process of tourism evolution in the region somewhat analogous to the resort cycle framework: (1) from small-scale, low-density, long-stay emerging Dominica, (2) through rapidly growing, larger scale, medium penetrated St. Lucia and Antigua, to (3) mature, high-density, high-impact, short-stay St. Maarten/St. Martin. The four cases also highlighted the strategic policy challenges at each level of tourism penetration: achieving visibility, managing growth and sustaining tourism quality.

The Index and case analyses underscore the need for the most penetrated destinations to emphasize quality over quantity as a strategy for controlling impacts. Targeting fewer more affluent tourists who stay longer and spend more locally is an approach that can be phased in gradually. Bermuda, one of the oldest and most penetrated destinations, has been implementing such a strategy over the past two decades. This restructuring has involved a moratorium on new hotel development and innovative attempts to expand visitor stay through restoring historical facilities, promoting "green" excursions, and arranging multiple-port cruise ship calls. The adjustment has been costly. Since 1980, the number of visitors has remained essentially flat and tourist expenditure has barely kept pace with inflation. This sacrifice of growth has been possible because residents prize the environment as part of their own individual identity and collective historical memory (de Albuquerque and McElroy 1995c). With better early warning signals in place like the TPI, it is hoped that Bermuda's no growth orientation will become more the rule than the exception in small islands. The outcome is important, as the "growth versus environment" and "quantity versus quality" debates are played out. The decisions made will largely determine whether tourism is a sustainable enterprise or just another short-lived chapter in the boom-bust cycles that have characterized the economic history of most small tropical islands.

References

Acharya, A.

1995 Small Islands: Awash in a Sea of Troubles. World Watch 8(4):24-34

Bacon, P. 1995 Wetland Resource Rehabilitation for Sustainable Development in the Eastern Caribbean. In Environment and Development in the Caribbean: Geographical Perspectives, D. Barker and D.F.M. McGregor, eds. pp. 46-56. Kingston, Jamaica: University of West Indies Press.

Barker, D. and D.F.M. McGregor, eds. 1995 Environment and Development in the Caribbean: Geographical Perspectives. Kingston, Jamaica: University of West Indies Press.

Beller, W., and d'Ayala, P. and P. Hein, eds. 1990 Sustainable Development and Environmental Management of Small Islands. Paris: Parthenon-UNESCO.

Briguglio, L., Butler, R., Harrison, D., and W.L. Filko, eds. 1996a Sustainable Tourism in Small Island States: Case Studies. London: Pinter.

Briguglio, L., Archer, B., Jafari, J., and G. Wall, eds. 1996b Sustainable Tourism in Islands and Small States: Issues and Policies. London: Pinter.

Briguglio, L. 1995 Small Island Developing States and their Economic Vulnerabilities. In International Symposium on Small Islands and Sustainable Development, G. Paoletto and R. Kuhr, eds., pp. 7-51. Tokyo: United Nations University.

Bryden, J. 1973 Tourism and Development: A Case Study of the Commonwealth Caribbean. Cambridge: University Press.

Bryden, J. and M. Faber 1971 Multiplying the Tourist Multiplier. Social and Economic Studies 20(1):61-82.

Burton, R. 1995 Travel Geography. London: Pitman, 2nd. ed..

Butler, R.W. 1993 Tourism Development in Small Islands. In The Development Process in Small Island States, D. G. Lockhart, D. Drakakis-Smith and J. Schembri, eds., pp. 71-91. London:Routledge.

Butler, R.W. 1980 The Concept of a Tourist Area Cycle of Evolution: Implications for Management of Resources. Canadian Geographer 24(1):5-12.

Caribbean Conservation Association 1991a Dominica: Country Environmental Profile. St. Michael, Barbados: CCA.

Caribbean Conservation Association 1991b St. Lucia: Country Environmental Profile. St.Michael, Barbados: CCA.

Caribbean Tourism Organization 1993 Caribbean Tourism Statistical Report. Christ Church, Barbados: CTO.

Cohen, E. 1978 The Impact of Tourism on the Physical Environment. Annals of Tourism Research 5(2):215-237.

Conlin, M. and T. Baum, eds. 1995 Island Tourism: Management Principles and Policies. Chichester, UK: John Wiley.

Cooper, C.P. 1993 The Life Cycle Concept and Tourism. In Choice and Demand in Tourism, P. Johnson and B. Thomas, eds., pp. 145-160. London: Mansell.

de Albuquerque, K.1991 Conflicting Claims on the Antigua Coastal Resources: The Case of the McKinnons and Jolly Hill Salt Ponds. In Caribbean Ecology and Economics, Girvan and Simmons, eds., pp. 195-205. St. Michael, Barbados: Caribbean Conservation Association.

de Albuquerque, K., and E. D'Sa 1986 Spatial Inequalities in Papua New Guinea: A District Level Analysis. Port Moresby, Papua New Guinea: Institute of Applied Social and Economic Research Discussion Paper No. 49, 78pp.

de Albuquerque, K., and J. McElroy 1992 Caribbean Small-Island Tourism Styles and Sustainable Strategies. Environmental Management 16(5):619-632.

de Albuquerque, K. and J. McElroy 1995a Antigua and Barbuda: A Legacy of Environmental Degradation, Policy Failure and Coastal Decline. Washington, D.C.: USAID, EPAT/MUCIA, Supplementary Paper No. 5.

de Albuquerque, K. and J. McElroy 1995b Planning for the Effective Management and Sustainable Development of Coastal Resources in Caribbean Small Island States. Caribbean Dialogue 2(1):11-16.

de Albuquerque, K. and J. McElroy 1995c Tourism Development in Small Islands: Lessons from St. Maarten/St. Martin and Bermuda. In Environmental Development in the Caribbean: Geographical Perspectives, D. Barker and D.F.M. McGregor, eds., pp. 70-89. Kingston, Jamaica: University of West Indies Press.

de Kadt, E. 1979 Tourism Passport to Development. New York: Oxford University Press.

Doxey, G. V. 1976 When Enough's Enough: The Natives are Restless in Old Niagara. Heritage Canada 2(2):26-27.

Getz, D. 1983 Capacity to Absorb Tourism: Concept and Limitations for Strategic Planning. Annals of Tourism Research 10(2):39-63.

Gray, H. P. 1974 Towards an Economic Analysis of Tourism Policy. Social and Economic Studies 23:386-398.

Hall, J. 1974 The Capacity to Absorb Tourists. Built Environment 3:392-397.

Harrison, D. ed. 1992 Tourism and the Less Developed Countries. West Sussex, UK: John Wiley and Sons.

Holder, J. 1988 Pattern and Impact of Tourism in the Caribbean. Tourism Management 9(2):119-127.

Johnson, P., and B. Thomas 1996 Tourism Capacity: A Critique. In Sustainable Tourism in Islands and Small States: Issues and Policies, L. Briguglio, B. Archer and others, eds., pp. 118-136. London: Pinter.

Kakazu, H. 1994 Sustainable Development of Small Island Economies. Boulder, CO: Westview Press.

Knox, J. M. 1982 Resident-Visitor Interaction: A Review of the Literature and General Policy Alternatives. In the Impact of Tourism in the Pacific, F. Rajotte, ed., pp. 76-101. Peterborough, CA: Trent Universtity.

Lanfant, M. F., Allcock, J. B. and E. M Bruner, eds. 1995 International Tourism: Identity and Change. London: Sage Publications.

Lean, G. 1994 Early Warnings from Small Islands. Choices 3(3):27-30.

Liu, Z. H. and C. J. Jenkins 1996 Country Size and Tourism Development: A Cross-Nation

Analysis. In Sustainable Tourism in Islands and Small States, L. Briguglio, B. Archer and others, eds., pp. 90-117. London: Pinter.

Lundberg, D. E. 1974a Caribbean Tourism: Social and Racial Tensions. Cornell Hotel and Restaurant Administration Quarterly 15(1): 82-87.

Lundberg, D. E. 1974b The Tourist Business. Boston: Cahners Books.

Mansperger, M. C. 1995 Tourism and Cultural Change in Small-Scale Societies. Human Organization 54(1):87-94.

Matthieson, A. and G. Wall 1982 Tourism: Economic, Physical and Social Impacts. London: Longman.

McElroy, J. and K. de Albuquerque 1988 The Migration Transition in Small Northern and Eastern Caribbean States. International Migration Review 22(3):30-58.

McElroy, J. and K. de Albuquerque 1991 Tourism Styles and Policy Responses in the Open Economy-Closed Environment Context. In Caribbean Ecology and Economics, N. P. Girvan and D. Simmons, eds., pp. 143-164. St. Michael, Barbados: Caribbean Conservation Association.

McElroy, J. and K. de Albuquerque 1996 Sustainable Alternatives to Insular Mass Tourism: Recent Theory and Practice. In Sustainable Tourism in Islands and Small States: Issues and Policies, L. Briguglio, B. Archer and others, eds., pp. 47-60. London: Pinter.

McElroy, J., de Albuquerque, K. and L. Dioguardi 1993 Applying the Tourist Destination Life Cycle Model to Small Caribbean and Pacific Islands. World Travel and Tourism Review 3:236-244.

McElroy, J., Potter, B. and E. Towle 1990 Challenges for Sustainable Development in Small Caribbean Islands. In Sustainable Development and Environmental Management of Small Islands, W. Beller and others, eds. pp. 299-316. Paris: Parthenon-UNESCO.

McGranahan, D. and Others 1972 Contents and Measurement of Socioeconomic Development: A Staff Study. UNRISD, New York: Praeger.

Miller, M. and J. Auyong 1991 Coastal Zone Tourism. Marine Policy 15(2):75-90.

Morris, M. 1979 Measuring the Conditions of the World's Poor. Overseas Development Council, New York: Pergammon.

Pattullo, P. 1996 Last Resorts: The Cost of Tourism in the Caribbean. London: Cassell.

Pearce, D. G. 1995 Tourism Today: A Geographical Analysis. Essex, U.K.: Longman.

Pearce, D.G. and R.M. Kirk 1986 Carrying Capacities for Coastal Tourism. Industry and Environment 9(1):3-7.

Potter, R. B.1989 Urbanization, Planning and Development in the Caribbean. London: Mansell.

Robbins, E. 1994 Trouble in Paradise. Environmental Magazine 5(3):36-40.

Rohlf, J. 1993 Numerical Taxonomy and Multivariate Analysis Systems (NTSYS-pc). Setauket, NY: Exeter Software, Version 1.8.

Samson, M. L.1989 The Netherland Antilles. In Urbanization, Planning and Development in the Caribbean, R. B. Potter ed., pp. 161-180. London: Mansell.

Sneath, P. H. A. and R. R. Sokol 1973 Numerical Taxonomy. San Francisco: Freeman.

Stankey, G. H. and S. F. McCool 1984 Carrying Capacity and Recreational Settings: Evolution, Appraisal, and Applications. Leisure Sciences 6(4):453-473.

United Nations Development Programme 1994 Human Development Report. New York: Oxford University Press.

United Nations Environment Programme 1985 Tourism and Environment in Development with Emphasis on the Eastern Caribbean. Port-of-Spain, Trinidad: ECLAC.

United States Congress 1992 Science and Technology Issues in Coastal Ecotourism. Washington, D.C.: U.S. Government Printing Office, Office of Technology Assessment.

Wall, G. 1982 Cycles and Capacity: Incipient Theory or Conceptual Contradiction. Tourism Management 3(3):188-192.

Wall, G. 1996 Rethinking the Impacts of Tourism. Progress in Tourism and Hospitality Research 2:207-215.

Watts, D. 1993 Long-Term Environmental Influences on Development in Islands of the Lesser Antilles. Scottish Geographical Magazine 109(3):133-141.

Weaver, D. B. 1988 The Evolution of a 'Plantation' Tourist Landscape on the Caribbean Island of Antigua. Tijdschrift voor Economische en Sociale Geografie 79(5):319-331.

Wheat, S. 1995 Creating the Right Environment. Leisure Management 15(11):26-29.

Wiley, J. 1996 Preparing for a Post-Banana Future: Economic Diversification in a Caribbean Mini-State. Paper presented to the American Association of Geographers. Charlotte, NC: April.

Wilkinson, P. 1989 Strategies for Tourism in Island Microstates. Annals of Tourism Research 16(2):153-177.

Zinder, H. and Associates 1969 The Future of Tourism in the Eastern Caribbean. Washington, D.C.: H. Zinder and Associate

______________________________________________________________

 

TABLE 3. 1988 and 1993 Rankings on the TPI (Weighted)

_________________________________________________________________

 

1988 1993

Most penetrated St. Maarten St. Maarten

Bermuda Cayman Islands

BVI Bermuda

Cayman Islands Aruba

USVI USVI

Aruba BVI

Anguilla Anguilla

Antigua/Barbuda Antigua/Barbuda

Barbados Turks and Caicos

Turks and Caicos Bonaire

St. Kitts/Nevis St. Kitts/Nevis

Montserrat Montserrat

St. Lucia St. Lucia

Martinique Curacao

Guadeloupe Guadeloupe

Curacao Martinique

Grenada Grenada

St. Vincent St. Vincent

Least penetrated Dominica Dominica

_________________________________________________________________

Source: Table 2 and McElroy and de Albuquerque, 1991.