Small Island Tourist Economies

Across the Lifecycle*

Jerome L. McElroy

Professor of Economics

Department of Administration and Economics

Saint Mary's College

Notre Dame, Indiana 46556 USA

TEL: 574-284-4488

FAX: 574-284-4566

EMAIL: jmcelroy@saintmarys.edu

*An earlier version of this paper was presented to the International Conference, ‘Beyond MIRAB: The Political Economy of Small Islands in the 21^st Century,’ School of Economics and Finance, Victoria University, Wellington, New Zealand, 23-25 February 2004.

Abstract

Using an abbreviated three-stage version of the destination lifecycle as a theoretical backdrop, this study applies the Tourism Penetration Index (TPI) to 36 small islands less than one million in population and 5,000 km2 in area. Most developed are primarily Caribbean, least developed Pacific and Indian Ocean, and intermediates from all regions. Tourism characteristics of each group (stage) are detailed and changes recorded between 1991 and 2001. Descriptive profiles reveal sharp differences in socio-economic and demographic modernization up the cycle. Rising tourism development is associated with rising income, in-migration, literacy and life expectancy and falling unemployment, fertility and infant mortality. These same differences surface when islands are dichotomously categorized by political status with dependencies outperforming sovereign islands. This evidence indirectly suggests successful tourism-driven small islands represent a special insular development case and an alternative to MIRAB. Key words: small, island, tourism, development, lifecycle.

Introduction

Establishing a distinctive island development paradigm has been hindered by disparate disciplinary and methodological approaches, varying definitions of size, and ‘unusually severe data constraints’ (Armstrong and others, 1998c: 639). As a result, island literature remains controversial and largely peripheral to the mainstream (Tsai and Clark, 2003). However , a highlight review of the historical evolution of island social science suggests progress has been achieved. Following on the pioneering work of Kuznets (1960) and Seers (1964), Demas (1965: 22) writing about the Caribbean declared: ‘The pattern of growth in a very small country must of necessity be different from that in a large continental country.’ Since small size precluded capital goods capacity and internally self-sustaining growth, he emphasized export growth engines and domestic policy innovation ‘whether cost-reducing or product-introducing’ (1965: 61).

A number of subsequent authors focused on other size constraints. For example, Knox (1967) identified the diseconomies of scale imposed by small domestic markets, the so-called high-cost sub-optimality problem further extended by Armstrong and Read (1998a). Benedict (1967) stressed the difficulty of achieving policy consensus and objective decision-making in closed insular societies characterized by intense kinship ties and ethnic cleavages. Soon after Selwyn emphasized island vulnerability due to thin domestic linkages (1975) and later (1978) disadvantageous geographic location and remoteness, a theme that dominated the work of Brookfield (1990: 27): ‘It is this transport stranglehold that is the sharpest single distinguishing constraint of island development today.’ Finally, Briguglio (1995) and Crowards and Coulter (1998) have both stressed the macroeconomic instability inherent in characteristic insular export specialization and concentration, marked import dependence and disaster proneness.

In a more positive vein, Dommen (1980) identified a distinct island demography based on the propensity to migrate and progress along the demographic transition. Bertram and Watters (1985) sketched out the special MIRAB case of South Pacific islands sustained by aid and remittances. In follow-up work, Bertram (2004) argued that long-run performance of Pacific economies was more due to metropolitan links than to regional influences. Very recently a number of authors maintain that small countries and islands out-perform larger states (Armstrong and Read, 2002), and dependent islands outdo their sovereign neighbors (Bertram, 2004; Armstrong and Read, 2000; McElroy and Mahoney, 2000). These studies suggest much of the insular advantage is due to domestic policy, i.e. restructuring the colonial economy toward services (tourism, off-shore banking) and, to a lesser extent, export manufacturing. Similarly, in an examination of North Atlantic islands, Baldacchino and Milne (2000) contribute the seminal notion that the distinctive feature of dependent island political economy is the deliberate use of jurisdiction or autonomy to legally manipulate metropolitan links for local economic benefit.

Scope

Despite this progress, questions persist whether islands are a useful category of analysis (Selwyn, 1980) and whether small size matters (Jalan, 1982; Srinivasan, 1986). Further study is warranted. This paper focuses on a cluster of small, tourist-driven island economies or so-called ‘Small Island Tourist Economies (SITES).’ It builds on earlier work involving a subset of Caribbean islands (McElroy and de Albuquerque, 1998) that applied the Tourism Penetration Index (TPI) to measure the degree of tourism development and (indirectly) assumed socio-environmental intrusion. Using an abbreviated three-stage version—least, intermediate, and most penetrated—of Butler’s (1980) lifecycle as a theoretical backdrop, the TPI was introduced as an early warning signal particularly for those high-density destinations risking potentially damaging levels of visitation. The model was later extended to a global sample of small islands (McElroy and de Albuquerque, 1999; McElroy, 2003).

Butler argues that successful destinations pass through a sequence of six successive stages—emergence, involvement, growth, consolidation, and maturity or stagnation followed by decline or rejuvenation—analogous to the product lifecycle (Heywood, 1986). Increasing development is associated with expanded facility scale, visitor saturation and cumulative ecological impact (Butler, 1991). Widespread early application of the model to individual islands/resorts and even to oceanic regions (Holder, 1988; Minerbi, 1988; Choy, 1992) uncovered several problems about empirically defining the stages, the inevitability of the progression, multiple products and cycles in a given destination and so on (Argawal, 1994). More recent research has emphasized that the lifecycle is but one of ‘a multitude of evolutionary scenarios’ (Papatheodorou, 2004: 233). Despite these limitations, Butler’s theory remains the most-quoted framework for describing the general pattern of tourism development (Hovinen, 2002). It is employed here not as a predictive model but as a heuristic device or classification scheme for describing the behavior and characteristics of small island destinations at various stages of tourist evolution.

The main objective of the present study is to explore the empirical validity of SITE experience as a special island development alternative to the MIRAB model. This is accomplished in two broad but related tasks. The first details, as completely as the aggregate data allow, the average tourism behavior of three subgroups of SITES at different stages of tourism evolution. The second and more important details the socio-economic and demographic contours of each subgroup to determine whether tourism specialization indeed is a viable engine of economic development, a relatively neglected issue in the mainstream literature (Sinclair, 1998; Tisdell, 2000). If the latter holds true, the study provides initial groundwork for treating these SITES—especially the most successful—as a special island development case.

Methodology

This section has four parts: (1) the first discusses the criteria chosen to define small size; (2) the second identifies the global island sample and data sources; (3) the third briefly reviews the construction of the index, and (4) the fourth discusses the analyses performed. In the first case, contrary to most previous island research, both population and area limits were used to define small size. This was done for three reasons: (1) to design as far as possible a relatively homogeneous sample; (2) to conform to procedures followed by two pioneers in the field—Demas (5 million and 30,000 km², 1967:22) and Brookfield (100,000 and 1,000 km², 1990:30)—and (3) to reduce the potential for distortion observed in previous TPI research. For example, because the index employs one land-based indicator (hotel rooms per km²), it may down-rank islands like Bahamas and Fiji with relatively developed tourism sectors but relatively large land areas. For purposes of this analysis, the small size limits were designated as population less than one million inhabitants and land area less than 5,000 square kilometers.

These two size criteria eliminated a number of islands used in previous TPI research (McElroy, 2003): Hawaii, Mauritius and Trinidad (over 1 million population), and Bahamas, Cyprus, Fiji, Iceland, New Caledonia, Solomon Islands, and Vanuatu (over 5,000 km²). Five other small islands were dropped because of excessive data deficiencies: Niue, Palau, SaoTome/Principe, Saba and St. Eustatius. The resulting sample included 36 islands: 19 in the Caribbean, nine in the Pacific, four in the Indian Ocean (Comoros, Maldives, Reunion, Seychelles), Bermuda and Cape Verde in the Atlantic, Malta in the Mediterranean, and Bahrain in the Persian Gulf.

The TPI is a comprehensive measure of tourism’s overall impact on small island societies. It is constructed in threes steps. As in past work, the three variables used to formulate the index were: (1) in-country visitor expenditure per resident population to broadly measure economic impact; (2) average daily visitor census or density per 1,000 population to indirectly measure crowding and socio-cultural pressure; and (3) hotel rooms per km² as a proxy for tourism’s environmental footprint. Next the three variables were normalized using a standard MAX-MIN formula. Then the TPI scores and destination rankings were developed by taking the unweighted average of the three impact indices. To operationalize the index, and to minimize bias, data were taken primarily from two standard sources: tourism data from the Compendium of Tourism Statistics (WTO, 2003), and demographic and other data from The World Factbook (CIA, 2003).

Finally, four analyses were performed. The first develops the TPI from most recently published data available (2001) and positions the 36-island sample across three levels of increasing tourism penetration. The second analysis develops the TPI using 1991 data to identify: (1) increasing levels of tourism development for the sample as a whole over the decade; (2) destination position changes over time; (3) and to further define dynamic aspects of the three stages along the cycle in so far as the limited data allow. The third draws demographic and socio-economic profiles of the three stages using over a dozen aggregate indicators to uncover patterns of societal and welfare progression associated with rising tourism intensity. Since recent island research links political status to insular economic performance in general and service (tourism) orientation in particular (Armstrong and Read, 2000; Bertram 2004), the fourth briefly compares profiles of politically dependent versus sovereign small islands to indirectly explore the role of status and/or autonomy in the SITE experience.

Results: TPI 2001

Table 1 presents the seven basic data series required for the TPI calculations for the 36-island sample. These include: area and population, number of overnight or stayover visitors, their average trip stay, number of one-day visitors, hotel rooms, and aggregate annual tourism expenditure. The left side Table 2 develops the three basic tourism indicators for the TPI: average visitor spending per island resident (Spend/pop), average daily visitor density per 1,000 population (Density/1,000), and rooms per square kilometer (Rooms/km2). On the right side of Table 2, the three related impact indices are derived and then averaged into combined TPI scores. The overall destination rankings are presented with the TPI loosely grouping the 36-island sample into three levels of development—most, intermediate and least—that correspond roughly to the three major stages of the resort cycle, namely emergence, growth, and maturity. In similar previous research (McElroy, 2003) on small islands, the index ranked Caribbean destinations at the top and Pacific islands at the bottom. This application is no different. The more accessible, traditional and popular Caribbean, Mediterranean and North Pacific resort areas dominate the top half of the rankings since they represent the established ‘Pleasure Periphery’ of North America, Europe and Japan respectively (Turner and Ash, 1976). The more remote Pacific and Indian Ocean newcomers populate the bottom.

(Table 1 about here)

The most developed comprise a subgroup of nine internationally popular destinations. Clearly they are the smallest of the sample, averaging roughly 250 km2 in area with a population just over 100,000. On the other hand, these islands are relatively crowded. Their average population density approaches 400 persons per km2, i.e. over 50 percent higher than the intermediate average (253) and over twice the average (166) for the least developed subgroup. The most developed are also the most tourism-driven. For example, average per resident in-country visitor expenditure approaches $12,000, and average daily visitor density exceeds 180 per 1,000 residents. This is nearly equivalent to a 20 percent increase in the daily year-round population. In addition, tourism leaves a visible imprint on the insular ecology. They average over 35 rooms per km2 of area. The subgroup includes six Caribbean destinations plus Bermuda in the Atlantic as well as Malta and Guam. The British Virgins’ top ranking is somewhat overstated by the TPI because the high visitor density figure (413 per 1,000 population) fails to capture the dominant yatching nature of their tourism activity. On the other hand, the heavily built environments of St. Maarten, Bermuda and Malta (respectively 75, 65 and 64 rooms/km2) demonstrate their designation as most developed destinations.

(Table 2 about here)

These affluent islands share a distinct tourism profile consonant with their advanced position in the lifecycle. First, they dominate the small-island tourism industry. As a group these nine islands account for over 50 percent of total one-day visitors and visitor expenditure for the entire 36-island sample, and over 40 percent of tourists and hotel rooms. To accommodate these high densities, they possess large-scale transport and facility infrastructure and attract visitors with a strong preference for large, full-service hotels and man-made attractions. According to the additional indicators assembled in Table 3 ( left-hand columns) to flesh out the picture, they also maintain the highest ratio of day-trippers (mainly cruise ship passengers) to stopover visitors. As an indication of their international visibility and commercial consolidation, as a group they boast the highest average share of holiday visitors (85%) as well as the highest reported average hotel occupancy rate (61%). Their tourists, however, also exhibit the shortest average stay (6.7 nights). This is a function of many factors: crowding, high incidence of package tours, business travel in some cases (off-shore finance in Bermuda and Turks/Caicos), and the transient character of some man-made attractions (shopping, conventions, golf weekends, gambling).

As a further index of their maturity, research on high-density Caribbean islands (McElroy and de Albuquerque, 1992) indicates they exhibit the highest per visitor advertising spending as well as the least seasonal visitation by promoting summer carnivals, regattas, sporting events, honeymoon getaways and the like. Not surprisingly, however, they ‘are also among the most frequently cited in the literature for tourism-induced ecosystem damage, marine pollution, overcrowding, host tensions and declining vacation quality’ (McElroy, 2003: 237). In this sense, they are partially victims of their own developmental progress, a success rooted primarily in five factors: (1) an abundance of natural and man-made assets; (2) geographic proximity to affluent origin markets in North America (Caribbean and Bermuda), Europe (Malta) and Asia (Guam); (3) historical commercial ties with metropolitan factor and capital markets; (4) pro-growth domestic policy and liberal investment/tax incentive legislation; and (5), since eight of nine top destinations are politically dependent, favorable metropolitan political affiliation that has fostered aid-financed transport infrastructure for airports, deep-draft harbors and overland roadways. These expensive physical investments form the indispensable capital base of a thriving visitor industry.

In contrast to these small affluent tourism leaders, the least developed islands recently emerging in the marketplace are considerably larger in area and population. Although there are exceptions (Marshalls, Tuvalu), they average 2.5 times larger in population and over six times larger in area than the most developed subgroup. Tourism is marginally visible on the landscape. For example, in-country visitor expenditure averages roughly $120 per resident, only one percent of the affluent island average. Likewise they average less than one hotel room per km2, and visitors add less than one percent (7/1,000) to the daily island population. They comprise relatively remote destinations including five in the Pacific (Marshalls, Kiribati, Samoa, Tonga, Tuvalu), two in the Indian Ocean (Comoros and Reunion), and Cape Verde in the Atlantic. These last three plus Samoa represent some of the largest small islands in the sample and typically exhibit relatively diversified agricultural economies: Reunion (sugar), Samoa (coconut oil), with Cape Verde and Comoros poor MIRAB economies dependent on subsistence farming/fishing and spices respectively (McElroy and Morris, 2002).

Despite their heterogeneity, the least penetrated destinations broadly share characteristics common to the early stages of the lifecycle. These include small-scale facilities and infrastructure and the absence of man-made attractions. One of their most distinctive markers, with the exception of Reunion, is the absence of day-trippers or cruise traffic, in great part of function of their remoteness. On the other hand, they boast the longest average length of visitor stay (9 nights) with over two weeks for Reunion and Tonga. This is partly due to heavy sunk travel costs on the part of visitors as well as the islands’ less disturbed biocultural diversity special appeal to longer-staying niche market segments. Since the majority of their visitors are business travelers and visiting relatives/friends, barely over a third (35%) on average are true holiday-makers (Table 3). Without strongly established international commercial ties, hotel occupancy averages less than 50 percent. In addition, they likely spend the least on promotion and suffer most from seasonality. Their slow progression up the lifecycle results from a constellation of self-reinforcing factors: remoteness (restricted air access) and limited infrastructure, modernization and promotion.

(Table 3 about here)

The intermediate islands fall between the most and least penetrated subgroups both in size and level of tourism development. Compared to the affluent islands, they average 150 percent larger in area, two-thirds larger in population, and record an average population density 33 percent lower. Their tourism impacts are also less visible. Average daily visitor density (55/1,000) is only 30 percent of the high-density average while per resident in-country visitor expenditure ($2,200) and hotel rooms per km2 (7.5) are only 20 percent of the most developed average. Nonetheless, these figures are seven-plus times larger than the least developed averages. They comprise a heterogeneous mix of 19 islands, many with substantial tourism experience and diversified economies in transition. Several face declining preferential markets for bananas (Dominica, Grenada, St. Lucia, St. Vincent) while others are restructuring away from traditional pursuits: Barbados, Guadeloupe, Martinique, and St. Kitts (sugar); Curacao (oil refining), Seychelles (tuna), Maldives (fishing) and Polynesia (farming, pearl diving). They also include: the Cook Islands, a MIRAB economy heavily dependent on migrant remittances and government subsidies from New Zealand; Montserrat down-ranked since 1995-1997 volcanic eruptions destroyed over half the island; and Northern Marianas down-ranked from previous TPI analyses because of a sharp 50 percent drop in tourism since the peak of 1996, a decline partly associated with Japan’s weakening economy and the Asian currency crisis.

The intermediates are also marked by a diversity of tourism styles. In addition to the mass sun-lust visitation common to most of these tropical resort areas, some of these niche markets include the following: dive tourism in Bonaire and the Maldives, bird-watching in the Seychelles, ecotourism in Dominica, sailing in St. Vincent, volcanic exploration in Montserrat, and honeymoon weekends in the Marianas. Despite this variety, many intermediate destinations share several common tourism traits characteristic of their middle-rung position along the lifecycle. Their distinguishing feature is rising international visibility as suggested by rising visitor densities, expanding infrastructure and hotel scale and the growth of man-made attractions (especially shopping). According to Table 3, over 80 percent of tourists visit for holiday purposes and average hotel occupancy is 60 percent. Day-trippers represent 43 percent of all visitors. In the Caribbean, these destinations also tend to have significantly higher levels of promotional spending than the least developed islands (McElroy and de Albuquerque, 1992). Their declining average visitor stay (8 nights) is also somewhat indicative of their increasing mass market appeal. Since many of these destinations are in the rapid expansion phase of the cycle, many experiencing growth pressures face resource-use conflicts and planning challenges as factors rapidly migrate from traditional activities to the modern and more lucrative tourism sector.

TPI 1991 and 2001

In order to test the stability of the TPI over time and explore some of the dynamic aspects of tourism evolution, another TPI was constructed using 1991 data to monitor decade trends. Because a comparative statics (not longitudinal) framework was employed, and the two years selected were somewhat unusual—worldwide recession (1991) and U.S. terrorist attack impacts (2001)--results should be considered suggestive only and cautiously interpreted. For this purpose, Table 4 records the basic 1991 data for the 36-island sample, and Table 5 presents the three basic TPI indicators, their impact indices and resulting TPI outcomes. In addition, decade rates of change are recorded in the right-hand side of Table 3 for four variables.

Between 1991 and 2001, several patterns are discernible. Increasing tourism impacts are substantial. First, in the most developed and intermediate islands, average spending per resident rose 50 and 30 percent respectively, and visitor density 15 and 20 percent respectively. Increases in the least developed subgroup were marginal. Second, in the two top groups, the growth in cruise traffic (day visitors) was three times faster than stopovers in the most developed, and twice as fast in the intermediates (see Table 3). Roughly two thirds of these destinations experienced an increasing share of one-day to total visitors. This indicates increasing industry consolidation, and certainly the successful promotion, affordability and security provided by the cruise lines in a more uncertain global political economy. Third, in the two developed subgroups, average room growth (37-38%) lagged average tourist growth (49 and 40%) somewhat while the difference for the least developed islands was more substantial. For all groups the average stay declined between 1991 and 2001 further suggesting accommodation to mass market demand.

(Tables 4 and 5 about here)

For the most part, between the two years the islands retained their basic TPI positions in the three subgroups with three exceptions. The Marianas fell from the most developed group in 1991 to the high end of the intermediates in 2001 largely as a result of declining Asian origin markets and the 2001 terrorist attacks. It was replaced by Turks/Caicos, which enjoyed the most robust growth in the sample. Stopover visitors there increased 200 percent while hotel rooms doubled and spending rose five-fold as a number of exclusive, full-service resorts came on stream over the decade. As evidence Duval (2004: 15) reported that Turks/Caicos maintained the highest average per trip visitor expenditure ($1,879) in 2000 among the 29-island insular Caribbean. Finally, Dominica graduated from low-density emerging status to the bottom of the intermediates mostly because of increasing cruise traffic. The number of one-day visitors rose over 200 percent between 1991 and 2001.

There are a few noteworthy destination changes within the subgroups. The British Virgins (BVI) overtook St. Maarten as the most tourist-penetrated small island. This resulted from two factors: (1) the doubling of tourists in the BVI largely for sailing charters and overstated by the land-based TPI; and (2) the 27 percent decline in tourists and ten percent drop in hotel rooms for St. Maarten over the decade (Table 3). This down-ranking is likely due to a variety of forces: (1) increased crowding and decreased repeat visitation, (2) intense competition from newer resorts in Turks/Caicos, Cuba and the Dominican Republic—the current ‘hot spot’ and tourism leader in the region (Padilla and McElroy, 2005)—(3) a nearly 30 percent decline in European visitors partly because of the rising strength of the dollar over these years, and (4) slow recovery from Hurricane Luis (1995). Bermuda experienced a steady, decade-long decline in tourists, rooms and visitor spending, in part the result of a deliberate policy to restrain growth, preserve natural/historical assets and retain its upscale image (de Albuquerque and McElroy, 1995).

Most position changes occurred among the intermediates, the most heterogeneous and dynamic destinations along the lifecycle. Four are notable. Bahrain moved up the scale doubling its hotel plant and one-day visitors because of increased trade following communism’s dissolution and its emergence as an international banking center and home to multinationals serving Gulf region countries. Maldives advanced five positions—more than doubling its tourists, rooms and spending because of an aggressive growth strategy and the attraction of its marine assets and private, low-density, segregated resorts (Domroes, 1999). Montserrat fell several steps because of the volcanic disaster. Over the decade tourists declined over 40 percent and hotel rooms fell over 80 percent (Table 3). Guadeloupe dropped three places because of a general slowdown, notably in stopovers which declined over 20 percent since the peak in 1997 (WTO, 2003). In addition to the impact of 9/11, the rising dollar was also partly responsible since roughly 70 percent of Guadeloupe’s tourists are Europeans. In summary, despite data limitations and small inter-group sample sizes, this descriptive two-year decennial analysis supports the basic stability of the TPI and provides evidence of increasing tourism penetration particularly among SITES in the most penetrated and intermediate stages of the lifecycle. Within subgroups, there were examples of growth and decline among the most developed, noticeable movement among the intermediates in transition, and stability for the least developed where no position changes occurred.

Socio-Economic Profiles

In order to further identify from aggregate data the contours of the tourist-driven island economy and demonstrate the impact of rising tourism intensity on small-island development and modernization, Table 6 presents recent (2003) socio-economic data across 13 variables for the 36-island sample and three island groupings established by the TPI. Despite fragmented data and possible mean distortion from small group sample sizes, predictable patterns emerge. Based on a comparison of average inter-group variable values, the most developed are the most economically advanced and demographically progressive while the intermediates far outpace the least developed destinations. For example, the affluent islands boast an average per capita GDP above $22,000 (PPP), nearly three times higher than the intermediate level. Their average number of telephones (main lines only) per 1,000 population is 60 percent higher while their average unemployment rate (6.1%) is only half the intermediate level (13.5%). Most importantly, their share of services to GDP—that is, orientation toward tourism (and off-shore banking to a lesser extent)—is significantly higher, 85 to 72 percent.

As a result of their advanced economic development, the affluent islands demonstrate a somewhat higher level of demographic maturity and social modernization compared to the intermediates. For example, they exhibit a slightly lower average rate of natural increase (crude birth minus crude death rates) and infant mortality as well as a slightly higher average life expectancy (77.6 years) and literacy rate (97.2%), levels that approach the most advanced continental societies. In addition, there are two other distinctive markers that contrast the most developed SITES from their intermediate counterparts. The first is a higher labor force participation rate (LFPR = LF/pop) of 47 to 41 percent. Part of this difference may be due to multiple family earner and multiple job holding patterns characteristic of heavily tourist-penetrated societies (McElroy and de Albuquerque, 1992). The second is contrasting net migration behavior, i.e. the difference between the population growth rate and the rate of natural increase. The most penetrated SITES average six immigrants per year per 1,000 population compared with only one migrant per 1,000 population for the intermediates. Both of these demographic differences in part result from more robust and pervasive labor-intensive tourist development in the former over the latter.

If these average values can be considered suggestive indicators for discrete levels of economic development, the distance between the intermediate and least developed is just as pronounced as between the affluent and the intermediates. For example, the intermediate average per capita income ($7,970) is more than three times the level of the least developed ($2,375), and the same gap holds for telephone availability, i.e. 307 to 115 per 1,000 population. Predictably, for these more tourist-penetrated islands the average service bias (72%) is higher (66%), and the average unemployment rate (13.5%) significantly lower (22.9%). In addition, there are similar demographic and social differences. For the intermediates these include lower average birth (18.8 vs. 26.6), total fertility (2.3 vs. 3.7), and infant mortality rates (16.6 vs. 35.7), and measurably higher literacy (92.8 vs. 85.7%) and life expectancy (73.2 vs. 67.6 years).

(Table 6 about here)

Just as immigration was a common feature of the most advanced SITES, so emigration is a common propensity for both intermediate and emerging destinations. In fact, the average positive net migration rate of 1.3 per 1,000 for the intermediates discussed above is actually overstated because of the unusual circumstances of Montserrat, which has been experiencing large annual waves of returning residents since the Soufriere Volcano has become quiescent. In 2003, its net migration was approximately 35 persons per 1,000 increasing the overall island population by nearly four percent. If Montserrat’s unusual experience is omitted from the calculation, average net migration becomes negative (-2.1/1,000) for the intermediate subgroup. Clearly many different factors condition the pattern of livelihood mobility in island nations (King and Connell, 1999). They include: geography, historical ties, ethnic and familial traditions, regional dynamics, political changes, legal constraints and the like. Yet weak economies are the dominant influence. Despite this similarity, overall the evidence indirectly suggests that tourism and related activities have played a major role in the intermediates’ advance ahead of the least penetrated—not only along the lifecycle but also along the continuum of economic development and socio-demographic modernization.

Contrasting profiles of course are even sharper when the most successful SITES are compared with the least developed island economies where agriculture is often the dominant activity and the tourism sector newly established or still under construction. In this case, average per capita income is over nine times higher, services orientation 20 points higher and unemployment 70 percent lower (see Table 6). The affluent islands’ higher level of economic development is also associated with greater progress along the demographic transition: markedly lower natality, fertility and infant mortality largely as the result of a burgeoning working–age population (15-64 years): 67 versus 59 percent. In contrast to the bulging population pyramid in the most developed destinations—swollen by immigration to service the labor-intensive demands of expanding tourism—the least penetrated islands are experiencing a wasting of the workforce cohorts through persistent emigration. As one consequence of these demographic processes, the affluent islands exhibit significantly higher literacy (97 vs. 86%) and life expectancy (77.6 vs. 67.6 years) than their poorer counterparts. Finally, absent from the most developed destinations are the MIRAB contours familiar to many of the least developed islands. To illustrate, remittances account for 20 percent of GDP in Cape Verde and exceed tourism earnings in Tonga (CIA, 2003). Between 1986-2002, the Marshall Islands received over $1 billion in U.S. assistance. Underemployment, symptom of a weak private economy, is estimated to be 70 percent in Kiribati. These stark differences draw contrasting contours between the SITE and the MIRAB models and further emphasize the special development case of the most tourism-penetrated islands.

Status Impact

Recent island research has emphasized the positive economic benefits deriving from the exercise of jurisdiction in the context of political affiliation in general (Armstrong and Read, 1998b, 2000; Baldacchino, 2004; Bertram, 2004), as well as for facilitating tourism development in particular (McElroy and Mahoney 2000). Specific advantages:

“…could include geographic proximity to and ease of travel (no passports, same currency) from major mother country origin markets, ready access to investment capital and aid-financed transport and communications infrastructure, special tax and duty-free concessions…that deserve further examination” (McElroy, 2003: 241).

Casual observation of the 36-island sample across the destination lifecycle suggests this further look. Eight of the nine most developed resort islands are dependencies while six of the eight least developed are sovereign nations.

To indirectly explore the connection between political status and tourism and socio-economic/demographic development, Table 7 divides the sample into two subgroups of islands—dependent and sovereign—and presents detailed tourism and socio-economic profiles for each across 17 variables. For purposes of this analysis, status was defined dichotomously with no varying degrees of affiliation for the dependent islands. As such, associated states like the Cook Islands (New Zealand) and the Marshall Islands (United States) were included with the colonies, overseas territories and departments in the dependent category. As constituted, there were 19 dependencies and 17 sovereign islands.

The results are suggestive even employing simple means of the aggregate data. Not surprisingly, on average the independent countries are measurably larger, i.e. 50 percent in population and 33 percent in area, indicating a more diversified resource and economic base and less dependence on tourism. On the other hand, the dependents are considerably more affluent and advanced along the lifecycle. Their average per capita income ($14,138) is over twice the average level of the independents ($6,429), a result similar to Poirine’s (1998) study that argued differences were due primarily to contrasting aid levels. On the other hand, Armstrong and Read (2000) emphasized the influence of policy and service (tourism, finance) orientation, as in this case where the average percent of GDP in services is significantly higher (80 vs. 68%) for dependents than for independents. Although average unemployment levels are similar, the number of phones per 1,000 population is much higher, i.e. 408 to 222 for dependencies. Although it appears that the sovereign islands possess somewhat larger tourism sectors and hotel plants, when Malta—with its 20,000 plus rooms and one million plus tourists—is excluded from the independent subgroup, outcomes are reversed. What is clear is that the dependents are more tourist-penetrated. In comparison with their sovereign counterparts, they average over six times more in-country visitor spending per resident ($6,899 to $1,026), three times average daily visitor density (133 to 36), and nearly two times more rooms per square kilometer (16.9 to 9.6).

(Table 7 about here)

These same mean differences are apparent across contrasting demographic behavior and levels of socio-economic modernization. The average crude birth (18.0) and total fertility (2.3) rates are somewhat lower than independent levels (21.8 and 3.0). More telling, the average net immigration rate of 6.3 per 1,000 population (3.9 without Montserrat) is significantly higher than the sovereign average net emigration rate (-4.9). This further suggests the dependencies are islands of growth and opportunity with expanding working-age populations and economic momentum. They have passed through the so-called ‘migration transition’ (McElroy and de Albuquerque, 1988) whereby former labor exporters become labor importers. The opposite is indirectly indicated for the labor-exporting independents, characterized by slower-growing economies and chronic labor surplus. Finally, the dependents demonstrate superior health and educational performance. Their average infant mortality rate (11.8) is less than half the sovereign level (26.8) while both average life expectancy (76.6 years) and literacy (96.0%) are markedly higher than levels recorded by the independents (65.6 years and 88.6%). All the above evidence suggests two things: (1) political affiliation may be a determining characteristic of SITE success; and (2) the propensity for dependence in island microstate is long standing. According to Bertram (2004: 353): ‘. . .there is no clear incentive for presently dependent island territories to seek independence, and good grounds for them to hold on to the status quo.’

Summary and Conclusions

This study explored the role of tourism development in small islands defined as less than one million inhabitants and 5,000 km2 in area. With the use of the Tourism Penetration Index, 36 islands were clustered across the destination lifecycle in three stages of increasing penetration, scale and socio-environmental impact. The most tourism-driven islands were the smallest located primarily in the Caribbean and characterized by large-scale facilities, man-made attractions (shopping), high visitor densities and hotel occupancy and short average stays. The least developed were the largest islands primarily in the Pacific and Indian Ocean and characterized by small-scale facilities, the absence of cruise traffic, longer average stays and lower occupancy. The most numerous and heterogeneous intermediates fell between the most and least developed subgroups and exhibited diverse tourism styles, expanding industry scale, cruise traffic, and growth pressures. An analysis of tourism trends between 1991 and 2001 confirmed the basic stability of the TPI analysis.. There was growth and decline (Bermuda and St. Maarten) among the most developed, movement for several intermediates in transition, and stability for the least penetrated.

Further examination of the socio-economic characteristics of the three island subgroups revealed the distinctive profile of the most successful tourism-driven small-island economies, the so-called SITES. In contrast to the intermediates and least developed, in stepwise order they were the most affluent and demographically progressive with the lowest fertility and infant mortality and highest life expectancy and literacy rates. They were also marked by immigration and rising working-age cohorts to service the labor intensive demands of expanding tourism. This was the reverse of the two other subgroups identified by double-digit unemployment and emigration. The same contrasting profiles surfaced when the sample was divided dichotomously according to political status. The smaller politically dependent destinations outperformed their sovereign neighbors across most tourism and socio-economic indicators.

These preliminary results suggest at least three things. First, for small island economies tourism is a viable engine of growth. Advancement up the destination lifecycle, although fraught with potential long-term consequences for socio-environmental stability, is also a path to socio-economic modernization and progress along the demographic transition. Second, the most successful of the tourism-driven insular microeconomies, the SITES, represent an interesting special case of island development and a clear MIRAB alternative that deserves further study. Although often favored by geography and lucrative metropolitan ties, to their credit such islands have aggressively pursued an endogenous policy of tourism promotion and have been swept forward by the juggernaut of the postwar global economy. Third and closely related, three promising directions for that further exploration would be: (1) to provide more rigorous statistical testing of the socio-economic and status profiles descriptively developed here; (2) to examine more in-depth specific case studies probing the variety of tourist advantages linked to political affiliation; and (3) to test for other aggregate determinants of tourism intensity other than status like size, geographic location and so forth. Such research would contribute toward the emerging political economy of small island tourism development.

References

Agarwal, S. (1994) The resort cycle revisited: implications for resorts, in C.P. Cooper and A. Lockwood (eds.), Progress in tourism, recreation and hospitality management. Chichester, UK: Wiley.

Armstrong, H. and R. Read (2002) The phantom of liberty? Economic growth and the vulnerability of small states, Journal of International Development 14(3): 435-458.

Armstrong, H. and R. Read (2000) Comparing the economic performance of dependent territories and sovereign micro-states, Economic Development and Cultural Change 48: 285-306.

Armstrong, H. and R. Read (1998a) Trade, competition and market structure in small states: the role of contestability, Bank of Valletta Review 18: 1-18.

Armstrong, H. and R. Read (1998b) The international political economy of micro-states: an overview. Paper presented at the Islands V Conference, University of Mauritius, Mauritius (July).

Armstrong, H., R.J. de Kervenoael, X. Li, and R. Read (1998) A comparison of the economic performance of different micro-states, and between micro-states and larger countries, World Development 26 (4): 639-656.

Baldacchino, G. (2004) Small islands versus big cities: lessons in the political economy of regional development from the world’s small islands, Journal of Technology Transfer (forthcoming).

Baldacchino, G. and D. Milne (2000) Lessons from the political economy of small islands: the resourcefulness of jurisdiction. London, UK: Macmillan.

Benedict, B. (1967). Sociological aspects of smallness, in B. Benedict (ed.), Problems of smaller territories. London, UK: Athlone Press.

Bertram, G. (2004) On the convergence of small island economies with their metropolitan patrons, World Development 32(2): 343-364.

Bertram, G. and R.F. Watters (1985) The MIRAB economy in South Pacific microstates, Pacific Viewpoint 26(3): 497-519.

Briguglio, L. (1995) Small island developing states and their economic vulnerabilities, World Development 23 (10): 1615-1632.

Brookfield, H.C. (1990) An approach to islands in W. Beller, P. d’Ayala and P. Hein (eds.), Sustainable development and environmental management of small islands. Paris, FR: UNESCO-Parthenon.

Butler, R.W. (1991) Tourism, environment and sustainable development, Environmental Conservation 18(3): 201-209.

Butler, R.W. (1980) The concept of a tourist area cycle of evolution: implications for management of resources, Canadian Geographer 24: 5-12.

Central Intelligence Agency (2003, 2001) The World Factbook. Washington, D.C.: CIA. http://www.odci.gov/cia/publications/factbook/index.html.

Choy, D.J.L. (1992) Lifecycle models for Pacific island destinations, Journal of Travel Research 30(3): 26-31.

Crowards, T. and W. Coulter (1998) Economic vulnerability in the developing world, with special reference to the Caribbean. Bridgetown, Barbados: Caribbean Development Bank, (July).

De Albuquerque, K. and J.L. McElroy (1995) Tourism development in small islands: St. Maarten/St. Martin and Bermuda in D. Barker and D.F.M. McGregor (eds.), Environment and development in the Caribbean: geographical perspectives. Kingston, JA: University Press of the West Indies.

Demas, W.G. (1965) The Economics of development in small countries. Montreal, CA: McGill University Press.

Dommen, E. (1980) Some distinguishing characteristics of island states, World Development 8 (12): 931-943.

Domroes, M. (1999) Tourism in the Maldives: the resort concept and tourist-related services, International Journal of Island Affairs 8 (3): 7-14.

Duval, D.T. (ed.) (2004). Tourism in the Caribbean: trends, development, prospects. London, UK: Routlege.

Heywood, D.M. (1986) Can the tourist area lifecycle be made operational? Tourism Management 7(3): 154-167.

Holder, J. (1988) Pattern and impact of tourism on the environment of the Caribbean, Tourism Management 9(2): 119-127.

Hovinen, G. (2002) Revisiting the destination lifecycle model, Annals of Tourism Research 29(1): 209-230.

Jalan, B.M. (1982) Problems and policies in small economies. New York, NY: St. Martin’s Press.

King, R. and J. Connell (1999) Small worlds: global lives. London, UK: Pinter.

Knox, A.D. (1967) Some economic problems of small countries in B. Benedict (ed.), Problems of smaller territories. London, UK: Athlone Press.

Kuznets, S. (1960) Economic growth of small nations in E.A.G. Robinson (ed.), Economic consequences of the size of nations. London, UK: Macmillan.

McElroy, J.L. (2003) Tourism development in small islands across the world, Geografiska Annaler 85 B (4): 231-242.

McElroy, J.L. and K. de Albuquerque (1999) Measuring tourism penetration in small islands, Pacific Tourism Review 3(2): 161-169.

McElroy, J.L. and K. de Albuquerque (1998) Tourism penetration index in small Caribbean islands, Annals of Tourism Research 25 (1): 145-168.

McElroy, J.L. and K. de Albuquerque (1992) An integrated sustainable ecotourism for small Caribbean islands. Bloomington, IN: Indiana University Press.

McElroy, J.L. and K. de Albuquerque (1988) Migration transition in small Northern and Eastern Caribbean states, International Migration Review 22(3): 30-58.

McElroy, J.L. and M. Mahoney (2000) The propensity for political dependence in island microstates, International Journal of Island Affairs 9 (1): 32-35.

McElroy, J.L. and L. Morris (2002) African island development experiences: a cluster of models, Bank of Valletta Review 26(Autumn): 38-57.

Minerbi, L. (1988) Alternative forms of tourism in the coastal zone: searching for responsible tourism in Hawaii. Honolulu, HI: University of Hawaii, Dept. of Urban and Regional Planning.

Padilla, A. and J.L. McElroy (2005) The tourism penetration index in large islands: the case of the Dominican Republic, Journal of Sustainable Tourism (forthcoming).

Papatheodorou, A. (2004) Exploring the evolution of tourism resorts, Annals of Tourism Research 31(1): 219-237.

Poirine, B. (1998) Should we love or hate MIRAB? The Contemporary Pacific 10(1): 65-105.

Seers, D. (1964) The mechanisms of an open petroleum economy, Social and Economic Studies 13 (2): 233-242.

Selwyn, P. (1980) Smallness and islandness, World Development 8 (12): 945-951.

Selwyn, P. (1978) Small, poor and remote: islands at a geographical disadvantage. IDS Discussion Paper No. 123. Sussex, UK: Institute of Development Studies.

Selwyn, P. (1975) Industrial development in peripheral small countries in P. Selwyn (ed.), Development policy in small countries. London, UK: Croom Helm.

Sinclair, M. (1998) Tourism and economic development: a survey, Journal of Development Studies 34: 1-51.

Srinivasan, T.N. (1986) The costs and benefits of being a small, remote, island, landlocked, or ministate economy, World Bank Research Observer 1(2): 205-218.

Tisdell, C. (2000) The economics of tourism. Cheltenham, UK: Edward Elgar.

Tsai, Huei-Min and E. Clark (2003) Nature-society interactions on islands: introduction, Geografiska Annaler 85 B (4): 187-189.

World almanac and book of facts, 1991 (1990). New York, NY: Pharos Books.

World Tourism Organization (2003, 1997) Compendium of tourism statistics. Madrid, SP: WTO.