Tourism competitiveness measurement. A perspective from Central America and Caribbean destinations

Tourism competitiveness measurement. A perspective from Central America and Caribbean destinations Víctor Ernesto Pérez Le on, Flor M a Guerrero and Rafael Caballero Abstract Purpose –This study aims to present diverse proposals for the measurement of tourism destination competitiveness that serve as alternatives to the travel and tourism competitiveness index (TTCI). Design/methodology/approach –The proposal includes principal component analysis, the DP2- distance method, goal programming, data envelopment analysis and the Borda count. The study evaluates 17 destinations from Central America and the Caribbean. Findings –These include the feasibility that the methodologies provide reliable competitiveness rankings and the possibility of using less information due to the strength of the statistical methodologies. International tourist arrivals, income from international tourism and travel and tourism contribution to the gross domestic product could be used as approximations of tourism destination competitiveness. Research limitations/implications –The main limitation is the absence of major destinations from the region that constitutes fierce competitors. Practical implications –New aggregation methods can build composite indicators for competitiveness measurement and their presentation in a more comprehensible way. Social implications –The results serve as an alternative for countries that have yet to be considered in international tourism competitiveness comparisons. Originality/value –A better explanatory power of the proposed index is given, thanks to their decomposition capacity and the reduction of the limitations of the original TTCI. Moreover, the proposals facilitate the inclusion of external information or the execution of a completely objective methodology. Keywords Competitiveness, Central America and the Caribbean, Composite indicators, Distance-based methods, Multicriteria, Data envelopment analysis Paper type Research paper 旅游竞争力测量-中美洲和加勒比海目的地的視角 目的 :本研究旨在为衡量旅游目的地竞争力提出多样化的建议,并作为旅行和旅游竞争力指数的替代方 案。 设计/方法/方法:该提案包括主成分分析、DP 2 距离方法、目标规划、数据包络分析和Borda 计数。该 研究评估了中美洲和加勒比地区的17 个目的地。 调查结果:结果包括这些方法提供可靠的竞争力排名的可行性,以及由于统计方法的优势而使用较少信息 的可能性。国际旅游人数、国际旅游收入以及旅行和旅游对GDP 的贡献可以用作旅游目的地竞争力的 近似值。 研究局限/影响:主要局限是该地区没有竞争激烈的主要目的地。 实际意义:新的聚合方法可以为竞争力测量建立综合指标,并以更易于理解的方式呈现。 社会影响:结果可作为国际旅游竞争力比较中,衡量尚未考虑国家的替代方案。 原创性/价值:由于其分解能力和原始TTCI 限制的减少,所提出的指数具有更好的解释力。此外,这些建 议有助于纳入外部信息及执行完全客观的方法。 关键词竞争力,中美洲和加勒比,综合指标,基于距离的方法,多标准,数据包络分析 文章类型研究型论文 (Information about the authors can be found at the end of this article.) Received 7 March 2022 Revised 14 May 2022 31 May 2022 Accepted 1 June 2022 ©Vı ´ctor Ernesto Pe ´rez Le on, Flor M a Guerrero and Rafael Caballero. Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at: http:// creativecommons.org/licences/ by/4.0/legalcode This work was supported by the Spanish Ministry of Economics and Competitiveness, Grant Number: PID2019-104263RB- C41/C42. Also, by the Andalusian Agency of Innovation and Development, Grant Number: P18-RT-1566, UMA18-FEDERJA-065. DOI 10.1108/TR-03-2022-0119 VOL. 77 NO. 6 2022, pp. 1401-1417, Emerald Publishing Limited , ISSN 1660-5373 jTOURISM REVIEW jPAGE 1401 Medici  on de la competitividad turı ´ stica. Una perspectiva desde los destinos de centro Ame ´rica y El Caribe Resumen Prop osito : El presente estudio busca presentar diversas metodologı´ as para medir la competitividad de los destinos turı´ sticos, de modo que sirvan como alternativa al I´ ndicedeCompetitividaddeViajesyTurismo. Diseño/metodología/enfoque : La propuesta incluye An alisis de Componentes Principales, el me ´todo de distancia DP 2 , Programaci on por Metas, An alisis Envolvente de Datos y el Recuento de Borda. Se analizan 17 destinos de Centro Ame ´rica y el Caribe. Hallazgos : Estos incluyen la validez de las metodologı ´ as para obtener rankings de competitividad fiables y la posibilidad de emplear menor cantidad de informaci on, dadas las fortalezas de los procedimientos estadı ´ sticos propuestos. Las Llegadas de Turistas Internacionales, los Ingresos por Turismo Internacional, y la Contribuci on del Turismo al PIB podrı ´ an ser buenas aproximaciones para medir competitividad turı ´ stica Limitaciones/implicaciones : La principal limitaci on es la ausencia de destinos importantes de la regi on, que se consideran importantes competidores. Implicaciones pr acticas : Novedosos procedimientos de agregaci on para crear indicadores sinte ´ticos para medir la competitividad turı ´ stica y su presentaci on de un modo m as comprensible. Implicaciones sociales : Los resultados sirven como alternativa para otros destinos que a un no han sido considerados en comparaciones internacionales de competitividad turı ´ stica. Originalidad : Un mejor poder explicativo de los ı ´ ndices propuestos, gracias a su capacidad de descomposici on, y la reducci on de las limitaciones del ı ´ ndice del WEF. Adem as, las propuestas facilitan la inclusi on de informaci on externa o la ejecuci on de un me ´todo completamente objetivo. Palabras clave Competitividad, Centro ame ´rica y el caribe, Indicadores sinte ´ticos Me ´todos basados en distancia, Multicriterio an alisis envolvente de datos Tipo de papel Trabajo de investigaci on 1. Introduction To assess destination competitiveness, researchers have diagnosed the competitive positions of a specific destination or groups of destinations using a wide range of approaches, tools and simple and specific indicators (Abreu-Novais et al.,2016). In addition, the literature reveals the existence of several studies dedicated to this end (Carayannis et al., 2018;Croes, 2011;Croes and Kubickova, 2013;Dwyer et al.,2000; G omez-Vega and Picazo-Tadeo, 2019;Kayar and Kozak, 2010;Knez ˇevi c Cvelbar et al., 2016;Kunst and Ivandi c, 2021;Ritchie and Crouch, 2010;Rodrı ´guez-Dı ´az and Pulido- Fern andez, 2021;Uyar et al.,2022). Nevertheless, the progress presented to date reveals, among other factors, certain limitations regarding the selection of evaluation variables and the calculation of their respective weights (Carayannis et al., 2018), the methodology used to aggregate the information and the explanatory power of the results. Amongst the diverse initiatives developed to measure destination competitiveness, there is the travel and tourism competitiveness index (TTCI) developed by the World Economic Forum (WEF) (WEF,2015, 2017, 2019), which constitutes the most noteworthy contribution. This index has been launched biannually since 2007 and serves as a comprehensive strategic tool to measure the factors and policies that make the development of the tourism sector attractive in various countries, by enabling all stakeholders to work jointly to improve the competitiveness of the tourism industry in their national economies, thereby contributing towards growth and national prosperity (WEF, 2019). The TTCI is composed of 14 “pillars” comprising a set of qualitative and quantitative variables. Each of the pillars is calculated as an unweighted average of the individual component variables. The sub-indices are then calculated as unweighted averages of the pillars included, and this process has remained invariable since its first publication (WEF, 2019). This is one of the most commonly used and feasible indices, thanks to its credibility, data accuracy (Abreu-Novais et al., 2016) and the desirable combination of hard and soft data, which is narrowly limited to a small number of initiatives. The index is a valuable comparability tool for the demonstration of destination strengths (Pe ´rez Le on, et al., 2021a) PAGE 1402 jTOURISM REVIEW jVOL. 77 NO. 6 2022 and support of their visibility. Consequently, following the findings of Uyar et al. (2022), diverse studies assess destination competitiveness using the TTCI in global analysis (Rodrı ´guez-Dı ´az and Pulido-Fern andez, 2021;Salinas et al.,2022) and evaluate different destinations according to their general behaviour (Salinas et al., 2020) or compare regional destinations, including the Mediterranean (Kunst and Ivandi c, 2021), Middle Eastern destinations (Leung and Baloglu, 2013), European Union countries (Kayar and Kozak, 2010) and Caribbean destinations (Pe ´rez Le on et al., 2021a), among others. This index is one of the most highly criticised initiatives in the measurement of destination competitiveness, due to its intense use. The criticism involves methodological issues (Croes and Kubickova, 2013), the arbitrary weighting of the variables (Pulido-Fern andez and Rodrı ´guez-Dı ´az, 2016;Salinas et al., 2020), the number of indicators within each pillar (G omez-Vega and Picazo-Tadeo, 2019), the components of the index that most influence destination competitiveness (Kubickova and Martin, 2020;Uyar et al.,2022), its viability as a reliable measure of destination competitiveness (Kunst and Ivandi c, 2021) and the amount of information required for its creation (Mendola and Volo, 2017), among other issues. The latter consideration is the main reason why various countries have been omitted from certain editions, as is the case with several destinations in the Caribbean region. While most developed countries succeed in collecting reliable tourism data, less developed countries struggle to provide accurate and timely statistics (Mendola and Volo, 2017). Alternatives are therefore needed that allow tourism competitiveness to be measured with a smaller number of indicators, whose degree of reliability and understanding is at least as high as that of the TTCI. Along these lines, this study aims to introduce various proposals for the measurement of tourism destination competitiveness (TDC) that serves as an alternative to the TTCI, which reduces negative aspects such as the amount of information needed for its creation, includes the possibility of introducing external information and provides ease in interpreting the results, thereby revealing the strengths and weaknesses of the destinations analysed, and identifying the contribution of the subindices to the global competitiveness measure. This research includes the achievement of a competitiveness ranking using different methods, such as the DP 2 -distance, principal component analysis, goal programming and data envelopment analysis (DEA), and the study of their differences according to the weights and aggregation processes. Additionally, a meta-index is obtained by means of the Borda count method through allowing decision-makers to achieve a global ranking representative of the overall degree of competitiveness for compared destinations, starting from the results of different aggregation methods. This is an innovative approach in the achievement of meta-indices as it enables the strengths of the composite indicators to be taken into account while striving to reduce their weaknesses. In contrast to other studies that use similar methods (Salinas et al.,2020, 2022), the DP 2 -distance proposes the identification of those indicators that measure tourism competitiveness without having to use all the information required in the TTCI and/or in the distance-principal component (DPC) indicator. Moreover, in contrast to G omez-Vega and Picazo-Tadeo (2019), our proposal uses goal programming to create the dimensional indicators with all the information available, together with the consideration of both internal and external information. Furthermore, the use of DEA is proposed to obtain the global competitiveness index so that it could be possible to identify the contribution of each dimension to the global measure. Finally, the use of the Borda count method is proposed to merge the rankings obtained and to solve the problem of their differences. Additionally, the study includes the comparison of the rankings obtained with the rankings from the WEF, both for each sub-index and globally to validate the feasibility of the proposed approaches. Comparison to other indicators related to TDC is also made to evaluate their possibility of being representative of a certain degree of competitiveness. VOL. 77 NO. 6 2022 jTOURISM REVIEW jPAGE 1403 This research involves the measurement of destination competitiveness in various destinations from Central America and the Caribbean region, using the country level. Notwithstanding, this topic has been addressed at different levels: resorts (Claver-Corte ´s et al., 2007), tour operator and hotel companies (Assaf, 2012), cities (Enright and Newton, 2005), regions (Cracolici and Nijkamp, 2009) and countries (Salinas et al., 2020). The approaches presented in this study are useful for all destination sizes and depend on the scope of the indicators and the information used. The paper is structured as follows. First, after the presentation of the research gap and of the objectives in the introduction, the proposals for the measurement of destination competitiveness are described in detail. The region case study and the data used in the verification of the suitability of the proposed methods are then presented. The results are given, both per dimension and globally, and include their relationship to other nonpreviously used indicators. Lastly, the conclusion section reveals the implications and proposes further research. 2. Methods 2.1 The DP 2 -distance indicator The first method, called the DP 2 -distance indicator, was initially developed to measure the evolution of social welfare (Pena, 1978;Zarzosa and Somarriba, 2013). This method is objective and eliminates the problems related to duplicity of information. It has also been used as an alternative in the measurement of TDC by Salinas et al. (2020,2022) to solve the problems arising from the aggregation of variables with different measurements and the assignation of arbitrary weights. The DP 2 -distance for a destination is defined as: DP2¼X n i¼1 di s i 1R2 i;i1;i2;...;1  with R2 1¼0 For i= 1, ... ,n,d i is the distance between the observed unit and the reference situation for the ith indicator, and s i is the standard deviation of the ith indicator. The d i dividing the standard deviation of each indicator eliminates the problems associated with the units of measure. R2 i;i1...1is the determination coefficient, and the term 1 R2 i;i1...1is the correction factor that represents the variability percentage of the ith indicator that is not lineally explained for the previous i1 indicators. In this way, the problem of information duplicity is solved because this coefficient eliminates the information contained in the ith indicator contributed in the i1 previously added indicators. This procedure contains certain advantages, such as its objectivity, its independence from normalisation processes and the fact that its weights are determined endogenously; therefore, any duplicity of information is eliminated. 2.2 Distance-principal component indicator (DPC) This indicator combines principal component analysis with the concept of distance to a reference point based on multi-criteria decision-making philosophy and is defined as follows: DPCi¼X q j¼1 VEjX p k¼1 INik jCorrjk j !"# for i = 1, 2, ...,n, where nis the number of observations, pis the number of original indicators, qis the number of components selected, VE j is the variance explained by the jth PAGE 1404 jTOURISM REVIEW jVOL. 77 NO. 6 2022 component and Corr jk is the correlation between the jth component and the kth indicator. IN ik is the normalised value of the ith observation in the kth indicator, which is needed for the normalisation of the data such that the measuring units used for each indicator exert no effect on the final result. This procedure involves dividing the distance to the anti-ideal point by the difference between the maximum and the minimum values: INik ¼Iik Min Max Min where I ik is the value of the ith observation in the kth indicator. The minimum value of each indicator is taken as the reference point while bearing in mind that higher values indicate that the destination is assumed to be more competitive. This approach enjoys certain advantages, such as the ease in interpreting the results, as the values of the initial indicators are defined according to their distance to a fixed reference value such that the synthetic indicator is a linear combination of these distances and not of the principal components. Moreover, weights are determined endogenously. The DP 2 and the DPC are valid for those analyses in which the destinations have not gathered all the information requested by the WEF for the creation of the TTCI. In this way, the procedure may help to identify the indicators required to measure destination competitiveness. Hence, only the data concerning the indicators resulting from the initial selection process should be collected, i.e. those indicators that contribute with a higher level of information to the competitiveness measure (DP 2 ) or those obtained from the prior application of the principal component analysis (DPC). Furthermore, their use is proposed when there is no information regarding the level of importance of the indicators. 2.3 The goal programming synthetic index (GPSI) The goal programming synthetic index (GPSI) is encouraged in the procedure of Blancas et al. (2010), whereby a set of minitial indicators (I j with j=1,2,...,m) is considered, for n units (U i, with i=1,2,...,n), where X ij represents the value of the ith unit valued in the jth indicator with 1 inand 1 jm. Firstly, it is necessary to differentiate between positive (Iþ ij ) and negative (I ik ) indicators, depending on the direction of improvement: “more is better” or “less is better”, respectively. In this way, Xþ ij represents the value for the ith unit in the jth positive indicator, with j[J,(J, positive indicators) and X ik is the value of the ith unit in the kth negative indicator, with k [K,(K, negative indicators). The achievement levels or the target for each indicator can therefore be determined: uþ jfor the positive and u kfor the negative. Subsequently, goals are created by introducing the deviation variables to measure the difference between the indicator value and the target: For positive indicators: Iþ ij þnþ ij pþ ij ¼uþ jwith nþ ij ;pþ ij 0;nþ ij pþ ij ¼0 For negative indicators: I ik þn ik p ik ¼u kwith n ik ;p ik 0;n ik p ik ¼0 where nþ ij is the undesirable variable for positive indicators, and p ik is the undesirable variable for the negative indicators. Values higher than these variables reveal an absence of competitiveness. This procedure enables several indices to be obtained and the net GPSI, GPSI N , is selected for its compensatory character between the strengths and weaknesses for each unit under evaluation. The GPSI N for a unit is defined as: GPSIN i¼X j2J wþ jpþ ij nþ ij  uþ j þX k2K w kn ik p ik ðÞ u k ;8i21;2;...;nfg where wþ jand w kare the weights for positive and negative indicators, respectively. The first sum shows the difference between the strengths and weaknesses for positive indicators, and similarly, the second sum shows this difference for the negative indicators. VOL. 77 NO. 6 2022 jTOURISM REVIEW jPAGE 1405 The contribution of this proposal in measuring TDC involves the possibility of establishing a lower bound for the indicators in such a way that a destination could be considered competitive with respect to this target value in comparison with its competitors. Moreover, there is the facility of interpreting the results through the identification of the strengths and weaknesses of the destinations under comparison in a more comprehensible way than when the TTCI is used. This procedure can be used both for those destinations that hold all the information available and for those that lack some data. This enables the inclusion of weights obtained externally. Once the dimensional indicators are obtained through the proposed methods, the second stage then involves the use of DEA to generate a global index, as described below. 2.4 Data envelopment analysis (DEA) DEA is a non-parametric technique used for the construction of composite indicators (G omez-Vega and Picazo-Tadeo, 2019). DEA models possess the advantage of displaying unit invariance, which renders the normalisation stage redundant. For this stage, the initial information was previously obtained from the dimensional indicators for each destination. A single dummy input with value unity for each destination can be used. This model is formally equivalent to the original input-oriented, constant-returns-to-scale DEA model presented (Charnes et al., 1978). The global synthetic index for the i 0 observation is obtained by solving the following the linear programming problem: DEAi0¼MaxwX d j¼1 wi0 jDIi0j subject to: X d j¼1 wi0 jDIij 18i¼1;...;nnormalisation constraint ðÞ wi0 jDIij  v 8i¼1;...;n;8j¼1;...;dvirtual output constraint ðÞ wi0 j08j¼1;...;dnon negativity constraint ðÞ where wi0 jare the weights for the observation i 0 ,DI represents the jth dimension indicator for the ith observation, which would be the DPC if the global index refers to DEA after distanceprincipal component (DEAPC) or the GPSI is used if the global measurement represents DEA after goal programming (DEAGP); dis the number of dimensions considered (the subindices held in the TTCI) and v is a real number that represents the minimum value allowed for the jth virtual output for the ith observation. The virtual output constraint involves the implication of all the dimensions in the global composite index. The objective function chooses the weights that maximise the value of the composite index for observation i 0 . In the best situation, the index takes a value of 1, which implies that the destination has a performance equal to its reference unit. The 0 value represents the worst situation. The [0,1] range is a characteristic of the input-oriented model, which numerically renders results more comprehensible and guarantees results with a higher explanatory power: this is a desirable characteristic of composite indicators. The virtual output constraint has been introduced to guarantee the presence of all dimensions in the composite index with a minimum value of v . Its use in the second phase of aggregation enables the identification of the contribution of each dimension towards the global index. PAGE 1406 jTOURISM REVIEW jVOL. 77 NO. 6 2022 2.5 The Borda count method The Borda count method uses mapping from a set of individual rankings to create a combined ranking that leads to the most relevant decision (Lumini and Nanni, 2006). In Borda count, a voter ranks all candidates in a strict order by assigning different points according to the ranking (Vainikainen et al.,2008). This method assigns zero points to a voter’s least preferred option, 1 point for the next option and (n–1) points for the most preferred (where nis the number of alternatives). However, this way of assigning zero points to the least preferred candidate is unfavourable for the implementation of the analytical calculation (Lawrence et al.,2012). The Borda ranking is therefore determined by placing the Borda scores in order. This approach is useful in those cases where the decisionmakers have attained different rankings due to the use of diverse aggregation methods. 3. Geographical context and dataset The study comprises 17 destinations from Central America and the Caribbean, which is the highest number of countries to have been included in an edition of the TTCI. These are underdeveloped countries in close geographical proximity, in the most tourism-dependent region worldwide, according to the World Travel & Tourism Council (WTTC) (2020,2021). These destinations compete within the same tourist market, and they offer a similar tourist experience: predominantly sun-and-sand tourism, with emphasis on cruise tourism, which has become big business, with the Caribbean accounting for more than 35% of all such vacations globally (Caribbean Council, 2019). Moreover, while the number of countries included in the TTCI had been steadily rising, the number of Caribbean countries included has decreased in the latest editions, thereby passing from 17 to 13 destinations in the space of only four years. Their absence was caused due to difficulties in providing all the information required. Consequently, some of the proposed approaches serve as alternatives to these destinations, due to the possibility of using less information. The data used correspond to the 2015 edition of the TTCI, which is the year that included the most destinations from the region. It comprises 90 indicators distributed into 14 pillars grouped into four sub-indices (WEF, 2015). Those indicators with more than three missing values are excluded. For those indicators with three or fewer missing values, their scores are substituted with the minimum. This substitution guarantees the presence of those indicators in the composite measure and, therefore, its representativeness. Moreover, the scores are not influenced, thanks to the proposed method. Consequently, the data set comprises 86 indicators: 40, 22, 14 and 10 for sub-indices A, B, C and D, respectively, of which 30 are subjective. In addition, all the pillars are presented in the study as follows: Pillar A.01 (12 indicators), A.02 (5), A.03 (6), A.04 (9), A.05 (8), B.06 (6), B.07 (3), B.08 (4), B.09 (10), C.10 (5), C.11 (4), C.12 (4), D.13 (5) and D.14 (5). 4. Results and discussion The aggregation process is developed in the same way as that proposed by the WEF to create the TTCI (Figure 1). Firstly, the indicators are grouped into their pillars, and the pillars are then used to create the dimensional indicators. Lastly, a global index is built by grouping the sub-indices. The dimensional indicators are created through the DP 2 , the DPC and the GPSI approaches. The global competitiveness index is subsequently built using the DP 2 and DEA methods. The latter is used for the two global indices proposed: 1. the DEAPC; and 2. the DEAGP methods. VOL. 77 NO. 6 2022 jTOURISM REVIEW jPAGE 1407 As a result, three alternative methods are presented for the creation of the dimensional and global indicators. Despite the unfeasibility of the DPC and DP 2 indicators embracing more indicators than destinations, the proposed steps allow the inclusion of all the information. Moreover, to attain a process as close as possible to the WEF proposal, the aspiration level used for the GPSI is zero. In this regard, under this approach, all the destinations only evaluate their strengths. The denominators are omitted from the GPSI, and therefore, the weaknesses are not included. The weights are achieved in a different way for each methodology. The DP 2 and the DPC methods calculate their weights endogenously. On average, the DP 2 assigns the highest weight to Sub-index B “T&T Policy and Enabling Conditions”, and the DPC gives more weight to Sub-index C “Natural and Cultural Resources”, while least importance is assigned to Sub-index A “Enabling Environment”. These assignations are consistent with the conditions of the region, comprised of underdeveloped countries with lower scores on safety, health, information and communication technology (ICT) readiness (WEF,2015, 2017) and the demonstrated efforts made by the governments towards the development of the tourism sector in the region (Pe ´rez Le on et al., 2021b), due to their dependence on this activity. For the GPSI, however, weights should be assigned. For the latter procedure (GPSI), the same importance is given to all the indicators contained in each pillar. For subdimensional indicators, the same importance is given to each pillar within the sub-indices. Lastly, all sub-indices receive the same importance in order to calculate the global indicator. 4.1 Dimensional results The results for the pillars appear in Table A1, while the dimensional results are shown on Table 1. The dimensional results reveal great stability amongst the rankings, including the comparison with the results attained with those from the WEF. The five most and least competitive destinations coincide in all the rankings created. This is a great achievement because, despite the differences between the procedures, the results seem to present major similarity. The results from the GPSI approach are those more closely related to the other methods and, compared to those of the TTCI, the DP 2 and the GPSI are those closest to the WEF outputs. This is a great advantage for the DP 2 methodology because it reveals its capacity for englobing the relevant information under the statistical methods comprised in the procedure. The similarities can be corroborated statistically. The Pearson correlation between scores demonstrates their similarity with all values higher than 0.871 and significant at the 0.01 Figure 1 Aggregation procedure PAGE 1408 jTOURISM REVIEW jVOL. 77 NO. 6 2022 Table 1 Dimensional results Destinations TTCI_A DP 2 _A DPC_A GPSI_A TTCI_B DP 2 _B DPC_B GPSI_B Score Rank Score Rank Score Rank Score Rank Score Rank Score Rank Score Rank Score Rank Barbados 5.25 1 10.256 1 3.365 1 5.246 1 4.24 8 6.378 14 1.086 13 3.974 13 Colombia 4.08 13 5.588 11 1.919 14 4.152 13 4.24 9 7.795 8 1.517 8 4.269 8 Costa Rica 4.84 2 9.197 2 3.272 2 4.921 2 4.47 3 9.735 3 1.722 5 4.514 3 Dominican Republic 4.21 9 5.514 12 2.221 10 4.264 9 4.07 12 6.876 12 1.380 10 4.131 10 El Salvador 4.13 11 5.672 10 1.999 13 4.174 12 4.4 4 8.758 5 1.768 3 4.440 5 Guatemala 4.14 10 5.986 8 2.085 12 4.209 11 4.32 6 8.55 6 1.713 6 4.376 6 Guyana 4.11 12 4.734 14 2.276 8 4.22 10 4.3 7 7.103 10 1.186 12 3.986 12 Haiti 3.42 17 1.319 17 1.569 16 3.42 17 3.98 14 4.341 15 1.046 15 3.695 15 Honduras 3.92 15 4.688 15 1.851 15 3.998 15 4.5 2 10.039 2 1.820 2 4.548 2 Jamaica 4.26 8 5.924 9 2.092 11 4.339 8 4.23 10 7.964 7 1.437 9 4.254 9 Mexico 4.34 7 6.117 7 2.249 9 4.392 7 4.22 11 7.3 9 1.519 7 4.343 7 Nicaragua 4.06 14 5.275 13 2.328 7 4.085 14 4.36 5 8.812 4 1.757 4 4.461 4 Panama 4.7 4 8.239 4 2.893 4 4.766 4 4.69 1 10.956 1 1.854 1 4.729 1 Puerto Rico 4.73 3 8.459 3 2.951 3 4.794 3 4.06 13 6.556 13 1.070 14 3.749 14 Suriname 4.42 6 6.356 6 2.577 5 4.411 6 3.67 16 3.785 16 0.842 16 3.456 17 Trinidad and Tobago 4.53 5 7.062 5 2.358 6 4.584 5 3.96 15 7.025 11 1.338 11 4.021 11 Venezuela 3.58 16 3.088 16 1.252 17 3.643 16 3.4 17 2.99 17 0.756 17 3.479 16 Destinations TTCI_C DP 2 _C DPC_C GPSI_C TTCI_D DP 2 _D DPC_D GPSI_D Score Rank Score Rank Score Rank Score Rank Score Rank Score Rank Score Rank Score Rank Barbados 5.18 1 6.224 1 2.245 1 4.990 1 1.65 16 6.961 4 0.796 4 1.645 16 Colombia 2.92 13 2.25 11 1.017 10 3.121 10 3.67 2 6.608 6 0.742 6 3.666 2 Costa Rica 3.7 7 3.397 9 1.484 7 3.901 7 3.39 3 8.638 2 0.959 2 3.348 3 Dominican Republic 3.68 8 3.667 7 1.480 8 3.676 8 2.05 10 4.768 13 0.607 11 2.046 10 El Salvador 3.32 9 3.641 8 1.256 9 3.316 9 1.78 15 5.202 11 0.526 15 1.752 15 Guatemala 2.95 12 2.449 10 0.960 12 2.950 12 2.64 6 5.745 7 0.638 7 2.659 6 Guyana 2.84 14 1.974 14 0.897 13 2.840 13 1.78 14 3.824 14 0.553 13 1.770 13 Haiti 2.29 17 0.337 17 0.372 17 2.245 17 1.3 17 0.523 17 0.343 17 1.290 17 Honduras 3 11 2.22 12 0.965 11 2.998 11 2.24 8 5.375 9 0.542 14 2.215 8 Jamaica 3.93 5 4.051 6 1.530 6 3.927 6 1.95 12 5.362 10 0.57 12 1.928 12 Mexico 3.83 6 4.7 4 1.705 5 4.026 5 5.05 1 8.458 3 1 1 5.006 1 Nicaragua 2.8 15 1.904 15 0.779 15 2.741 15 2.28 7 5.076 12 0.631 8 2.301 7 Panama 4.72 2 5.43 3 2.229 2 4.742 2 3.02 5 8.753 1 0.876 3 2.963 5 Puerto Rico 4.64 3 5.882 2 2.051 3 4.645 3 2.22 9 6.684 5 0.777 5 2.171 9 Suriname 3.01 10 2.049 13 0.839 14 2.785 14 2.01 11 3.567 15 0.626 9 2.012 11 Trinidad and Tobago 4.57 4 4.682 5 1.942 4 4.566 4 1.8 13 5.42 8 0.613 10 1.767 14 Venezuela 2.43 16 0.727 16 0.5556 16 25.585 16 3.31 4 2.157 16 0.511 16 3.276 4 VOL. 77 NO. 6 2022 jTOURISM REVIEW jPAGE 1409 Appendix Table A1 Results for the pillars Destinations/pillars A.01 A.02 A.03 A.04 A.05 DP 2 DPC GPSI DP 2 DPC GPSI DP 2 DPC GPSI DP 2 DPC GPSI DP 2 DPC GPSI Barbados 11.508 3.600 0.925 9.438 2.678 1.151 12.759 2.527 1.227 12.683 2.559 0.962 10.878 2.489 0.981 Colombia 7.639 2.644 0.845 2.418 0.708 0.564 10.131 1.990 1.021 10.088 2.272 0.875 7.923 1.998 0.847 Costa Rica 11.293 3.700 0.958 8.978 2.546 1.117 10.566 2.145 1.054 13.592 2.836 0.941 9.188 2.287 0.851 Dominican Republic 8.917 2.805 0.859 5.503 1.561 0.86 9.861 1.838 0.992 8.777 2.002 0.834 4.942 1.426 0.72 El Salvador 9.002 2.917 0.857 4.053 1.011 0.723 9.672 1.884 0.99 9.190 1.984 0.838 6.581 1.735 0.766 Guatemala 9.167 3.226 0.858 3.918 0.946 0.715 8.903 1.807 0.981 11.075 2.325 0.865 7.110 1.819 0.79 Guyana 10.346 3.398 0.914 6.145 1.784 0.914 4.923 1.259 0.964 8.864 2.069 0.79 3.580 0.985 0.638 Haiti 3.632 1.507 0.594 5.848 1.824 0.949 2.899 0.420 0.697 5.087 1.360 0.759 1.663 0.034 0.42 Honduras 7.728 2.586 0.834 4.186 1.014 0.727 8.284 1.694 0.944 9.816 2.058 0.833 4.667 1.270 0.66 Jamaica 9.919 3.168 0.92 4.78 1.146 0.77 7.323 1.496 0.973 10.753 2.301 0.911 6.629 1.591 0.764 Mexico 7.794 2.742 0.843 4.612 1.373 0.819 11.763 2.303 1.081 9.918 2.156 0.868 6.792 1.633 0.78 Nicaragua 7.265 2.219 0.725 7.926 2.270 1.028 7.446 1.503 0.871 7.870 1.791 0.807 4.409 0.894 0.655 Panama 12.221 4.107 1.016 7.674 2.138 1.005 10.142 2.007 1.04 9.562 1.892 0.827 8.424 2.293 0.879 Puerto Rico 12.23 3.853 0.971 7.51 1.952 0.963 13.32 2.565 1.189 12.387 2.590 0.878 7.152 2.132 0.792 Suriname 8.289 2.374 0.718 9.067 2.594 1.122 9.38 1.895 1.05 6.691 1.384 0.769 5.647 1.301 0.751 Trinidad and Tobago 11.365 3.788 0.937 4.853 1.325 0.82 9.404 1.855 1.069 9.964 1.997 0.863 8.646 2.058 0.895 Venezuela 2.932 0.715 0.502 3.238 0.774 0.672 10.856 2.114 1.057 4.452 0.860 0.709 4.486 1.245 0.704 Destinations/pillars B.06 B.07 B.08 B.09 DP 2 DPC GPSI DP 2 DPC GPSI DP 2 DPC GPSI DP 2 DPC GPSI Barbados 10.312 2.218 1.391 2.378 0.411 0.712 4.129 0.512 0.896 13.600 2.032 0.975 Colombia 7.138 1.422 1.095 5.279 1.012 1.075 4.733 1.121 1.119 12.899 2.047 0.981 Costa Rica 10.468 2.078 1.324 5.063 1.105 0.996 5.203 1.073 1.1 15.373 2.467 1.094 Dominican Republic 11.069 2.334 1.455 3.657 0.699 0.809 3.355 0.965 1.006 10.810 1.646 0.861 El Salvador 8.869 1.595 1.142 6.456 1.307 1.132 6.245 1.408 1.228 10.400 1.697 0.938 Guatemala 6.969 1.436 1.108 5.115 1.084 0.986 7.864 1.634 1.337 11.623 1.785 0.945 Guyana 6.123 1.076 0.921 3.893 0.611 0.914 7.677 0.972 1.051 13.331 2.102 1.101 Haiti 5.198 1.085 0.946 4.259 0.593 0.97 7.630 0.989 1.031 6.385 1.132 0.748 Honduras 8.794 1.718 1.201 5.444 1.148 1.012 7.537 1.528 1.286 14.293 2.222 1.05 Jamaica 10.130 2.173 1.382 4.449 0.782 0.899 5.991 1.029 1.073 11.161 1.737 0.901 Mexico 10.699 2.058 1.315 3.427 0.814 0.87 6.089 1.297 1.221 9.930 1.628 0.936 Nicaragua 6.761 1.481 1.135 5.311 1.119 0.991 7.877 1.624 1.314 12.543 1.929 1.019 Panama 10.908 2.088 1.323 5.000 1.114 1.025 7.225 1.481 1.287 15.667 2.280 1.094 Puerto Rico 9.907 1.873 1.217 2.960 0.244 0.461 7.287 1.161 1.204 17.358 2.254 0.979 Suriname 4.120 0.815 0.861 1.210 0.236 0.581 5.689 0.711 0.906 13.402 2.151 1.108 Trinidad and Tobago 3.997 0.830 0.877 4.063 0.715 0.875 8.456 1.497 1.336 10.854 1.686 0.932 Venezuela 5.531 0.835 0.852 0.763 0.112 0.56 7.456 1.003 1.158 7.436 1.257 0.908 Destinations/pillars C.10 C.11 C.12 D.13 D.14 DP 2 DPC GPSI DP 2 DPC GPSI DP 2 DPC GPSI DP 2 DPC GPSI DP 2 DPC GPSI Barbados 5.636 1.110 1.124 7.863 2.450 1.892 9.83 1.937 1.975 3.134 0.430 1.085 0.197 0.109 0.56 Colombia 3.824 0.729 0.887 1.960 0.710 1.039 5.29 0.949 1.195 5.393 1.387 2.066 4.249 2.288 1.6 Costa Rica 4.146 0.895 0.986 2.816 0.877 1.158 8.95 1.701 1.758 8.346 1.829 2.546 1.312 0.680 0.803 Dominican Republic 4.219 0.898 0.941 3.917 1.230 1.192 7.38 1.403 1.542 2.864 0.596 1.293 0.936 0.556 0.753 El Salvador 3.220 0.537 0.735 5.172 1.620 1.356 5.28 0.976 1.225 1.751 0.422 1.091 0.491 0.298 0.66 Guatemala 1.864 0.344 0.631 3.046 0.980 1.085 5.88 1.029 1.234 4.425 1.126 1.837 1.043 0.711 0.822 Guyana 3.766 0.943 1.077 2.674 0.877 1.034 1.94 0.299 0.728 2.514 0.569 1.218 0.064 0.060 0.553 Haiti 0.862 0.141 0.538 0.104 0.030 0.709 3.79 0.673 0.998 0.180 0.069 0.732 0.149 0.105 0.558 Honduras 2.031 0.428 0.698 2.697 0.928 1.067 5.37 0.990 1.233 3.544 0.808 1.487 0.630 0.445 0.728 Jamaica 3.436 0.675 0.815 5.164 1.714 1.582 7.19 1.376 1.531 3.280 0.572 1.249 0.524 0.365 0.68 Mexico 7.280 1.249 1.213 4.781 1.469 1.323 7.08 1.335 1.49 7.914 1.891 2.59 6.762 3.931 2.415 Nicaragua 1.694 0.237 0.577 2.311 0.670 0.955 5.57 0.984 1.208 3.546 0.899 1.605 0.518 0.418 0.697 Panama 8.419 1.739 1.515 5.094 1.750 1.406 9.06 1.764 1.821 6.549 1.473 2.169 1.471 0.622 0.795 Puerto Rico 5.318 0.976 1.015 7.419 2.296 1.829 9.28 1.766 1.801 4.268 0.768 1.444 1.104 0.529 0.726 Suriname 2.001 0.423 0.719 2.901 0.947 1.06 4.77 0.689 1.006 4.066 0.811 1.48 0.048 0.035 0.532 Trinidad and Tobago 6.494 1.475 1.392 4.898 1.521 1.456 7.85 1.585 1.717 2.399 0.469 1.139 0.381 0.223 0.629 Venezuela 1.450 0.419 0.712 0.907 0.253 0.821 2.62 0.604 1.025 5.140 1.518 2.243 1.214 0.989 1.032 PAGE 1416 jTOURISM REVIEW jVOL. 77 NO. 6 2022 About the authors Vı ´ctor Ernesto Pe ´rez Le on is a Lecturer in Statistics at the Department of Applied Economics II, University of Seville. PhD in Economic Sciences (University of Pinar del Rı ´o) and PhD in Business Administration (Pablo de Olavide University). Experience in Operations Research acting on composite indicators applied to tourism sustainability and competitiveness. Contribution to the article: data collection, data analysis and interpretation, drafting the article. Vı ´ctor Ernesto Pe ´rez Le on is the corresponding author and can be contacted at: [email protected] Flor M a Guerrero is a Professor of Mathematics Methods for Economy and Business at Pablo de Olavide University. Her research activity focuses on mathematical applications in Business Administration. In particular, sustainable development, tourist activity, social welfare, through simple and synthetic indicators, multidimensional indicators and multiobjective optimisation. Contribution to the article: conception or design of the work, critical revision of the article, final approval of the version to be published. Rafael Caballero is a Professor at Department of Applied Economics (Mathematics), University of M alaga, Spain. He is interested in the field of multiple objective programming. Presently, his research is in metaheuristics methods and application to problems in economy and business and forestry sciences. Contribution to the article: conception or design of the work, critical revision of the article. For instructions on how to order reprints of this article, please visit our website: www.emeraldgrouppublishing.com/licensing/reprints.htm Or contact us for further details: [email protected] VOL. 77 NO. 6 2022 jTOURISM REVIEW jPAGE 1417