Chapter 4. Cuba Awakening: Potential Risks and Opportunities

Krishna Srinivasan, Inci Otker, Uma Ramakrishnan, and Trevor Alleyne
Published Date:
November 2017
  • ShareShare
Show Summary Details
Sebastian Acevedo and Joyce Wong 


The growth of Caribbean tourism other than to Cuba is partly a post–Cuban revolution phenomenon. After the Cuban revolution in 1959, U.S. travel restrictions in 1963 closed U.S. tourism to one of the preferred Caribbean destinations of U.S. travelers. In 1953, the last year of tourism statistics in Cuba before the revolution,1 Cuba received almost half of all tourist arrivals to the Caribbean; by 1980 Cuba had less than 3 percent of the market compared with the same set of countries.2 For example, in The Bahamas—despite a long history of tourism promotion that started with the Tourism Encouragement Act of 1851—it was the U.S. embargo on Cuba that provided “the main stimulus to the tourism industry,” with U.S. tourists switching to The Bahamas (The Bahamas Ministry of Tourism 2016). Tourist arrivals to The Bahamas grew from about 150,000 in 1954 to more than a million in 1968. Mexico also followed suit with the directed development of Cancun as a tourism destination.

Just as the closing of U.S.-Cuba relations was a boon to other Caribbean tourism destinations, could the normalization of U.S.-Cuba relations reverse these gains? Since the announcement in December 2014 that the United States and Cuba were normalizing relations, the United States has relaxed travel restrictions to Cuba by allowing travel without prior authorization for 12 categories, while still banning outright tourism flows. Nevertheless, 2015 was a record year for Cuba’s tourism sector, with growth in arrivals of 17.4 percent (including growth of 21.8 percent in the “other” category, in which the United States is grouped).3 Despite this sharp increase in arrivals to Cuba, the rest of the region still fared quite well, with average growth of 6 percent from all tourism sources, and 6.6 percent growth in U.S. tourist arrivals.

These short-term effects are encouraging for the rest of the Caribbean; however, given Cuba’s sheer size, complete removal of travel restrictions between the United States and Cuba could potentially entail a much deeper structural change to tourism flows from the United States to the Caribbean, with varying effects across Caribbean destinations and over time. Thus, the potential impact across each Caribbean island of U.S. tourists’ switch to Cuba will depend on (1) each island’s dependence on U.S. tourists, (2) its gain from the Canadian and European tourists displaced from the Cuban market, and (3) the extent to which more U.S. tourists shift toward the Caribbean, generating an overall growth in Caribbean tourism.

Against this background, this chapter estimates the potential impact from a full removal of U.S. travel restrictions using two approaches:

  • An updated gravity model following the approach of Romeu (2008, 2014)
  • A calibrated structural setup in which the Caribbean market is the outcome of preferences with constant elasticity of substitution across destinations

Stylized Facts

Cuba’s tourism market has been dominated by Canada, whose tourists account for 40 percent of the 3 million visitors to Cuba annually. Canada has also been the fastest growing source market for Caribbean tourism in the past 20 years (Figure 4.1), which suggests that it is possible for tourist arrivals from a particular source market to expand rapidly in Cuba while also growing and benefiting all other countries in the region. Going forward, more tourism from the United States to Cuba will likely have a positive impact on the region’s share in global tourism; however, it will also likely result in a rearrangement of market shares within the region. Caribbean destinations most affected by Cuba’s possible rapprochement with the United States are likely to be those most dependent on the U.S. market (for example, The Bahamas, Bermuda, Jamaica, Belize, St. Kitts and Nevis).

Figure 4.1.Evolution of the Canadian and Cuban Tourism Market, 1995–2014

Sources: Acevedo, Alleyne, and Romeu 2017; Caribbean Tourism Organization; and World Tourism Organization.

Note: In the case of destinations, “Other” includes Anguilla, Antigua and Barbuda, Aruba, Barbados, Belize, Bermuda, British Virgin Islands, Cayman Islands, Curaçao, Dominica, Grenada, Guyana, Haiti, Martinique, Montserrat, Puerto Rico, St. Kitts and Nevis, St. Lucia, Sint Maarten, St. Vincent and the Grenadines, Suriname, Turks and Caicos, and United States Virgin Islands. In the case of sources, “Other” includes the rest of the world.

The changes in U.S. policy toward Cuba since December 2014 are quite significant: travel has been facilitated under much more general licenses for 12 categories,4 the permitted level of remittances to Cuba has been increased, and commercial relations and authorized transactions have been broadened. However, tourism is still banned, although cruise ships and scheduled commercial airline traffic have resumed within the 12 categories of travel allowed. Although many of these changes are quite recent, 2015 was a record year for Cuba tourism—arrivals increased by 17.4 percent, with across-the-board increases in Canadian and European tourists, as well as U.S. visitors.

Some Caribbean destinations are more at risk of experiencing disruptions in their tourism sectors if a change in U.S. policy allows unrestricted tourism travel to Cuba, thereby increasing competition for U.S. tourists. Figure 4.2 shows the countries whose tourism source is predominantly the United States (that is, more than 50 percent). All the countries to the left of the white line depend heavily on U.S. tourists, with the U.S. Virgin Islands, Puerto Rico, Turks and Caicos, The Bahamas, the Cayman Islands, and Bermuda receiving more than 70 percent of their tourists from the United States. The tourism destinations on the right side of the white line, on the other hand, are more diversified away from the United States; hence, if the number of U.S. tourists visiting them were to decline, the shock would be smaller, and it would likely be easier for them to attract visitors from other countries to compensate for the decline in U.S. visitors.

Figure 4.2.Caribbean Tourist Arrivals by Source, 2014

(Percent of total)

Sources: Acevedo, Alleyne, and Romeu 2017; Caribbean Tourism Organization; World Tourism Organization; and authors’ calculations.

Note: “Other” includes the rest of the world. Data labels in figure use International Organization for Standardization (ISO) country codes.

The extent to which a country’s tourism strategy has recently concentrated on targeting the U.S. market matters as much as its current dependency on U.S. tourists. To study this aspect, the change in the share of tourists from all source markets was calculated, and is presented in Figure 4.3. All the countries to the left of the black bar have become more dependent on the U.S. tourism market since 2000. The countries to the right have diversified away from the United States, and thus are better prepared if a U.S.-Cuba opening were to result in a decline of U.S. tourists to the region. The countries to the left of the bar seem to have focused their efforts on attracting more U.S. tourists, so they are less prepared to diversify their visitor sources.

Figure 4.3.Caribbean Change in Share of Tourism Sources, 2000–14


Sources: Acevedo, Alleyne, and Romeu 2017; Caribbean Tourism Organization; World Tourism Organization; and authors’ calculations.

Note: “Other” includes the rest of the world. Data labels in figure use International Organization for Standardization (ISO) country codes.

1 For Guyana the base year is 2001, and for Suriname the base year is 2006.

Not surprisingly, there is some overlap between the “most at risk” countries identified in Figures 4.2 and 4.3. Among those countries, the U.S. Virgin Islands, Puerto Rico, and Turks and Caicos stand out as the most vulnerable to a potential disruption in U.S. tourist flows. It is important to note that while some countries for which more than 50 percent of their visitors come from the United States have been diversifying away from the U.S. market (for example, The Bahamas and Bermuda), other countries with relatively low U.S. dependency (for example, Dominican Republic and Antigua and Barbuda) have actually been targeting the U.S. tourism market more in the past 15 years.

The enforcement—or relaxation—of U.S. travel restrictions to Cuba has had an important effect on the number of flights (panel 1 of Figure 4.4).5 Thus, one of the concerns surrounding the opening of U.S.-Cuba travel is that air travel services might be diverted from the rest of the Caribbean to Cuba. Acevedo and others (2016) use a panel structural vector autoregression and find that more flights between the United States and Cuba has no statistically significant impact on the availability of flights to the rest of the Caribbean in the first few months (panel 2 of Figure 4.4).6 Similar results were found for an increase in the number of U.S. flights to the rest of the region, suggesting that there is no substitution effect such that one destination’s gain is another’s loss. In other words, airlift supply between the United States and the Caribbean is not a zero-sum game.

Figure 4.4.U.S. Flights to Cuba and the Impact on the Rest of the Caribbean

Some possible explanations for the absence of a substitution effect could be that the airline industry expands its fleet to increase the number of flights to a Caribbean destination, airlines shift flights from other regions (domestic, Latin America), or they accommodate greater demand by scheduling more flights without requiring an increase in the fleet.

Potential Effects of Liberalizing Tourism Between the United States and Cuba

This section examines the possible effects that a change in U.S. travel policy toward Cuba could have, not only on U.S.-Cuba tourism flow, but also on the rest of the Caribbean. Two methods—a gravity model and a structural model— are used to analyze these effects.

It is important to be mindful of some characteristics of the Cuban economy that could affect the transition:

  • In the short term, Cuba will likely be constrained by capacity and hindered by its dual exchange rate system wherein tourists transact at a 1–1 exchange rate with the U.S. dollar, while locals transact at 24–1 with the U.S. dollar. Under such a setting, a large inflow of U.S. tourists (some of whom may substitute Cuba for other Caribbean islands) could potentially mean significant tourism-related price hikes in Cuba, potentially displacing current tourists from Canada and Europe. The winners and losers from this short-term switching effect across Caribbean islands will hinge on each country’s dependence on U.S. tourists and the extent to which Canadian and European tourists are willing to substitute another island for Cuba (as discussed below).
  • In the long term, as Cuba builds capacity, exchange rate adjustments take place, and the “curiosity factor” wanes for U.S. tourists, the region will likely settle into a new equilibrium determined by structural factors. Gradual changes are already taking place, and so far the region has coped well with the relaxation of travel restrictions from the United States to Cuba with little or no negative effects registered.

Gravity Model

This modeling approach explicitly accounts for the U.S. travel restrictions to Cuba and then calculates a counterfactual scenario in which those restrictions are removed.7 Bilateral tourist arrivals are a function of the distance between the destination and source country capitals, and factors that affect trade links such as cultural and historical relations (for example, whether the countries share a common language, have a colonial history or a common colonizer, or are part of the same country, such as the United States and Puerto Rico). To capture the effect of the travel restrictions imposed by the United States, the inquiry includes dummy variables that reflect periods when U.S. travel restrictions toward Cuba were more tightly enforced (1996–97 when the Helms–Burton Act increased sanctions, and 2004–08 when travel restrictions to Cuba were enforced more strongly).8

The results indicate that the full removal of travel restrictions would increase tourism flows from the United States to Cuba in the range of 3 million to 5.6 million people a year.9 Most of the increase would come from new tourists to the region—that is, the overall number of tourists traveling to the Caribbean would increase. Nonetheless, some Caribbean destinations might see some temporary decline in U.S. arrivals (as is shown next). Table 4.1 suggests that part of Cuba’s gain in U.S. tourism would be at the expense of other destinations (the ones with red cells in the U.S. column), but some Canadian and European tourists who currently visit Cuba could decide to visit other countries in the region once prices in Cuba start rising (countries with green cells in the Canada and U.K. columns). Interestingly, Miami could benefit from Canadian and U.K. visitors switching away from Cuba because it is the closest Caribbean destination to both source countries and has close cultural and historical ties.

Table 4.1.Actual versus Predicted Arrivals
DestinationUnited StatesCanadaUnited KingdomOtherTotal
Antigua and Barbuda
The Bahamas
British Virgin Islands
Cayman Islands
Dominican Republic
Puerto Rico
Sint Maarten
St. Kitts and Nevis
St. Lucia
St. Vincent and the Grenadines
Trinidad and Tobago
Turks and Caicos
United States Virgin Islands
Note: The table compares actual tourist arrivals in 2013 with the predicted arrivals from model (1) in Annex Table 4.1.1. A red cell indicates that the model predicts fewer arrivals than the ones that actually took place in 2013, while a green cell indicates that the model predicts more arrivals than the actual.

U.S. arrivals to Cuba based on the unrestricted model.

Note: The table compares actual tourist arrivals in 2013 with the predicted arrivals from model (1) in Annex Table 4.1.1. A red cell indicates that the model predicts fewer arrivals than the ones that actually took place in 2013, while a green cell indicates that the model predicts more arrivals than the actual.

U.S. arrivals to Cuba based on the unrestricted model.

Countries most vulnerable to possible spillovers from the U.S.-Cuba opening are (1) those with larger dependence on the United States as a source market in 2014, (2) those whose share of U.S. tourists has increased in recent years, and (3) those that could lose some U.S. tourists after a change in U.S. travel policy toward Cuba as identified by the gravity model. The results are summarized in the Venn diagram in Figure 4.5, identifying Anguilla, Belize, Sint Maarten, and the U.S. Virgin Islands as the more vulnerable group. This result does not imply that the other destinations are completely safe, or that this group will see declining U.S. tourism flows; it only highlights the countries that would need to be more alert to possible spillovers from changes in the U.S.-Cuba relationship.

Figure 4.5.Destinations Most Vulnerable to a Change in U.S.-Cuba Travel Policy

Sources: Acevedo, Alleyne, and Romeu 2017.

Note: Labels in figure use International Organization for Standardization (ISO) country codes.

Structural Model: A Change in “Preferences”

A structural approach to examining the effects from the normalization of U.S.-Cuba relations confirms the findings above. In this setup, the United States, Canada, and Europe were assumed to have constant elasticity of substitution preferences10 across Caribbean destinations:

where qi is the quality and preference indicator for destination i, pi is the price of a visit to destination i, and visit si is the number of visits to destination i.

Using data on visitors’ arrivals from each market (United States, Canada, Europe) to each destination in the Caribbean and the Week-at-the-Beach Index (see Chapter 3 in this book) for prices at two times (2014 and 2016), three sets of implied “preferences” (United States, Canadian, European) for each Caribbean destination are calibrated ({q1*,,q1*}j,j=US,CAN,EUR).11

Each country’s implied preference rankings are shown in Figure 4.6. In general, the preferences of the three markets are quite similar: The Bahamas ranks first for all three markets (United States, Canada, Europe) with Dominica and St. Vincent and the Grenadines ranking last. The places where U.S. preferences differ significantly are Puerto Rico (second for the United States and seventh and eighth for Canada and Europe, respectively) and, of course, Cuba (11th for the United States and third and fourth for Canada and Europe, respectively, where the U.S. ranking reflects the U.S. institutional ban).

Figure 4.6.Implied Preferences for Destinations

Sources: Caribbean Tourism Organization; and authors’ calculations.

The possible impact of U.S. tourists in the absence of any institutional restrictions is analyzed by assuming that U.S. tourists, in this case, would have the same preferences as Canadian tourists (including for Cuba), and Cuba is ranked third. In this case, arrivals to Cuba from the United States would increase by more than 3.3 million a year (from the current 96,000). Not all of this increase, however, would mean losses for other destinations: jointly the other 12 destinations would lose about 2 million visitors per year, with the largest losses in nominal terms coming from the main U.S. destinations.

Thus, out of Cuba’s total gain in visitors, about 60 percent would be due to trade diversion while 40 percent would be due to trade creation. Why is there trade creation? Because of Cuba’s relatively low prices: since the model assumes that the overall resource envelope for Caribbean destinations does not change, one trip to The Bahamas could be replaced by trips to Cuba and Belize, for example.

However, based on this simple model, the change in U.S. preferences would imply significant loss in market shares for Caribbean destinations highly dependent on U.S. tourists (Figure 4.7). Destinations most affected include the Dominican Republic, Puerto Rico, Jamaica, and The Bahamas (whose current U.S. market shares range from 17 to 27 percent), which could see their shares of the U.S. market drop by an average of 9 percentage points. Under the new preferences, Cuba would become the main U.S. destination, receiving more than 40 percent of U.S. tourists to the Caribbean. The results of this static analysis should be viewed with caution because (1) it fails to account for the duration of this transition, which, in turn, depends on Cuba’s supply-side responses; and (2) it makes a simplifying assumption that the Caribbean islands currently receiving large proportions of U.S. tourists also make no supply-side responses to become more attractive and competitive destinations.

Figure 4.7.Change in U.S. Market Share

Source: Authors’ calculations.

Note: Data labels in figure use International Organization for Standardization (ISO) country codes.

Conclusions and Policy Implications

With the United States being the single largest tourism market for the Caribbean and, for most countries, the most important source of tourists, a full removal of U.S.-Cuba travel restrictions would undoubtedly bring about significant shifts in total tourists to the Caribbean and relative market shares. However, these shifts will not necessarily cause only negative outcomes for the rest of the Caribbean. There will be a period of adjustment and more intense competition, which, as in the past (when Cancun and the Dominican Republic became dominant tourist destinations), the Caribbean destinations must confront with sensible policies.

  • In countries where the dependence on the U.S. market is large, a diversification strategy that targets other advanced economies and large emerging markets in Latin America would be beneficial. Tapping into new markets, historical links, and diaspora resources could help provide a much-needed boost.
  • Improving competitiveness and reducing the costs of the tourism sector will be crucial. Supply-side reforms to decrease reliance on imports, increase domestic links (for example, with domestic agriculture and manufacturing), and strengthen physical infrastructure could support more market diversification.
  • Upgrading quality and improving product marketing and differentiation (for example, by fostering cultural tourism instead of just “sun and sand”) will help countries compete with a nascent low-cost provider like Cuba and attract tourists outside of the all-inclusive model.
  • Finally, putting in place regional strategies to facilitate intraregional travel (for example, through a hub-and-spoke airline model) would help nurture the possibility of multidestination vacations. Such a model would help the rest of the Caribbean benefit from the new tourists who will start visiting the region when the United States opens free travel to Cuba.

It is encouraging that the region is actively addressing some of these challenges. The tourism authorities and local hoteliers are proactively embarking on efforts to enhance their tourism product by tapping into new markets, developing new products, promoting investment, forging new partnerships, and developing human capital. In addition, the Caribbean Hotel and Tourism Association has been actively engaging with Cuban authorities to explore partnerships in promoting multidestination initiatives. These various initiatives should mitigate the risk of decline in tourist arrivals from the United States to the Caribbean.

In addition, the whole process is likely to be gradual because Cuba will also need to adjust its economic policies to scale up investment and improve the quality of its tourism services. In the short term, higher U.S. tourism demand in Cuba may push prices up and potentially displace some Canadian and European tourists who would have otherwise visited Cuba but may instead travel to other Caribbean destinations. This displacement would partly offset any potential loss of U.S. tourists that some destinations might suffer in the adjustment phase to the new equilibrium. These shifts will provide support for other Caribbean destinations as they put in place reforms to adapt to the new equilibrium. In the long term, the change in U.S. policy is expected to benefit the whole region as aggregate tourism flows grow.

Nevertheless, the region should not be complacent about the changes that could come later. Many of the necessary reforms are significant and would take time to yield results; the earlier countries embark on these reforms, the readier they will be for Cuba’s reopening.

Annex 4.1
Annex Table 4.1.1.Gravity Estimations
Variables(1) In TA(2) In TA(3) In TA
U.S.-Cuba Restrictions−3.41***−3.61***−3.59***
Tightening of Restrictions−0.75***−0.73***−0.75***
Common Language1.13***1.08***1.09***
Common Colonizer0.51**0.54**0.53**
Colonial Ties1.36***1.40***1.41***
Same Country1.19**1.15*1.11*
Puerto Rico0.781.04
H1N1 Epidemic−0.12−0.180.10
Natural Disasters−1.81*−2.20***−1.92***
Low-Income Country−1.18***−1.27***−1.37***
9/11 Attacks−7.43***−7.53***−7.62***
Antigua and Barbuda0.04−0.03−0.04
The Bahamas−0.05−0.12***−0.12***
British Virgin Islands−0.13***−0.18***−0.19***
Cayman Islands−0.19***−0.19***−0.19***
Dominican Republic0.03−0.02−0.02
Puerto Rico−0.11−0.09−0.13***
Sint Maarten−0.07−0.13***−0.13***
St. Eustatius−0.27***−0.35***−0.35***
St. Kitts and Nevis−0.17**−0.24***−0.26***
St. Lucia−0.05−0.14***−0.15***
St. Vincent and the Grenadines−0.07−0.16***−0.17***
Trinidad and Tobago−0.16**−0.25***−0.26***
Turks and Caicos−0.30***−0.31***−0.32***
United States Virgin Islands−0.24***−0.25***−0.26***
Adjusted R20.900.900.90
Note: βUS captures the effect of the log of U.S. tourist arrivals to each destination’s other sources, that is, the elasticity of non-U.S. arrivals to a change in U.S. arrivals for each destination. CAFTA = Central America Free Trade Agreement; CARICOM = Caribbean Community; CBI = Caribbean Basin Initiative; TA = tourist arrivals.*** p < 0.01, ** p < 0.05, * p < 0.1.
Note: βUS captures the effect of the log of U.S. tourist arrivals to each destination’s other sources, that is, the elasticity of non-U.S. arrivals to a change in U.S. arrivals for each destination. CAFTA = Central America Free Trade Agreement; CARICOM = Caribbean Community; CBI = Caribbean Basin Initiative; TA = tourist arrivals.*** p < 0.01, ** p < 0.05, * p < 0.1.
Annex Table 4.1.2.Price and Visitor Data for Calibrated Model
Visitors (per month)
Price IndexUnited StatesCanadaEuropeOtherALL
Antigua and Barbuda1.748,6381,6198,0303,45621,743
The Bahamas2.5292,36810,4177,5549,415119,754
Dominican Republic1.15112,50828,12659,900104,711305,244
St. Lucia1.2113,2162,5156,6536,19628,580
St. Vincent and the Grenadines1.231,9206321,8052,1276,483
St. Kitts and Nevis2.215,4423828492,4319,103
Puerto Rico1.46134,1841,6933,72010,843150,440
Source: Caribbean Tourism Organization and author’s calculations.
Source: Caribbean Tourism Organization and author’s calculations.

    Acevedo, Sebastian, TrevorAlleyne, and RafaelRomeu. 2017. “Revisiting the Potential Impact to the Rest of the Caribbean from Opening US-Cuba Tourism.” IMF Working Paper 17/100, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

    Acevedo, Sebastian, LuHan, MarieKim, and NicoleLaFramboise. 2016. “Flying to Paradise: The Role of Airlift in the Caribbean Tourism Industry.” IMF Working Paper 16/33, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

    Anderson, J., and E.Van Wincoop. 2003. “Gravity with Gravitas: A Solution to the Border Puzzle.” American Economic Review93 (1): 170–92.

    • Search Google Scholar
    • Export Citation

    The Bahamas, Ministry of Tourism2016 “The History of the Ministry of Tourism”

    Baldwin, R., and D.Taglioni. 2006. “Gravity for Dummies and Dummies for Gravity Equations.” NBER Working Paper 12516, National Bureau of Economic Research, Cambridge, MA.

    • Search Google Scholar
    • Export Citation

    Romeu, R. 2008. “Vacation Over: Implications for the Caribbean of Opening US-Cuba Tourism.” IMF Working Paper 08/162, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

    Romeu, R. 2014. “The Vacation Is Over: Implications for the Caribbean of Opening US-Cuba Tourism.” Economίa14 (2): 1–24.

    Wong, JoyceCheng. 2017. “A Giant Awakens: Structural Shifts to the Caribbean from Opening US-Cuba Tourism.” Mimeo.

This chapter is partly based on Acevedo, Alleyne, and Romeu (2017), and on Wong (forthcoming).


The Cuban revolution started in July 1953, and the rebel forces seized control in January 1959.


In 1953 only seven Caribbean countries reported tourist arrivals to the World Tourism Organization: The Bahamas, Barbados, Cuba, the Dominican Republic, Haiti, Puerto Rico, and Trinidad and Tobago. In 1953 tourist arrivals to the Caribbean were 649,911, but by 1980 arrivals to these seven countries had reached 4 million.


Cuba’s office of statistics does not report a separate line for U.S. visitors, but the “other” category is a large residual after reporting the 17 largest source markets, and it is believed to be mostly U.S. visitors.


These 12 categories comprise family visits; official business of the U.S. government, foreign governments, and certain intergovernmental organizations; journalistic activity; professional research and professional meetings; educational activities; religious activities; public performances, clinics, workshops, athletic and other competitions, and exhibitions; support for the Cuban people; humanitarian projects; activities of private foundations or research or educational institutes; exportation, importation, or transmission of information or information materials; and certain export transactions that may be considered for authorization under existing regulations and guidelines.


Figure 4.4 also illustrates that there have been rapid changes in the supply and demand for fights between the United States and Cuba, which allows an estimate to be made of the impact in other Caribbean destinations of the changes in the number of U.S.-Cuba fights.


Chapter 3 has a more detailed description of the work of Acevedo and others (2016).


The idea behind gravity models is that trade flows are determined by trade costs; this approach can be applied to the movement of physical goods or to the movement of people for trade in tourism services. This chapter follows the work of Romeu (2014), which is based on the work of Anderson and Van Wincoop (2003) and Baldwin and Taglioni (2006).


The model also controls for the participation of destination countries in different trade agreements (that is, CARICOM, the Central America Free Trade Agreement, and the Caribbean Basin Initiative) that facilitate travel and investment, including in tourism facilities. Also, to get a sense of the impact that U.S. tourists have on non-U.S. arrivals, this elasticity is estimated for each country. The model also includes destination-year and source-year indicators that capture nonsystemic tourism determinants of destination and source countries (for example, GDP). The model also controls for the effects of natural disasters; for the September 11, 2001, attacks that disrupted travel around the world; for the H1N1 outbreak in 2009; and for low-income country destinations that might have insufficient infrastructure capacity for a well-functioning tourism sector.


The results of the estimations are presented in Annex Table 4.1.1, where column (1) shows the main estimation, and columns (2) and (3) show some alternative specifications that serve as robustness checks. To estimate the potential gain in U.S. tourist arrivals to Cuba we calculate a counter-factual scenario using these estimations. We set both variables measuring the restrictions to zero while at the same time setting the CBI variable for Cuba to one; that is, we assume that the United States includes Cuba as part of the CBI initiative. Model 1 predicts the highest gains (5.6 million), while Model 2 predicts gains of 3.2 million, and Model 3 predicts an increase of 3 million.


Constant elasticity of substitution preferences are widely used in the literature to convey a preference for a basket of different goods rather than homogeneous consumption (“taste for variety”).


”Preferences” should be used carefully in this setup: while the implied values for Canada and Europe likely reflect preferences for destinations, for the United States they also reflect the current institutional environment in which U.S. tourists are restricted from visiting Cuba; thus, the outcome of prices and numbers of visits reflects more than just market factors.

    Other Resources Citing This Publication