This paper analyses the determinants of flows of domestic tourism in Uruguay in the context of an extended gravity model during the period 2010-2012. This analysis is done at a disaggregated level, considering the bilateral flows between two regions (department) of the country, one as source of tourism and the other as the destination. Gravity models suggest that tourist flows depend positively on the size of each region (department in Uruguay) and negatively on the distance between them. The extended version includes explanatory variables to characterize the supply and demand for domestic tourism.
Three models are estimated, one for each year, which allows to compare and to estimate the robustness of the results. The methodology of estimating applied (following Santos, Silva and Teynero (2006)) was Pseudo-Poisson Maximum Likelihood with cross-section data which has been proven, is one of the most appropriate for estimating gravity models. Domestic tourist flows are represented from the number of trips from one region (department) to another, for the 19 departments of the country within a calendar year.
Results obtained in this paper constitute a first contribution to the analysis of the determinants of domestic tourist movements, and provides potentially valuable information for decision-making of public and private sectors (planning strategies, policy, marketing, communication).
The study shows that tourist flows depend positively on the size of population of each department and negatively of the distance that separates them. This result is in line with the general gravity models and also with the results recently found in another countries (Gálvez, Muro and Such, 2014; Massida and Etzo, 2012; Marrocu and Pacci, 2013). The regions (departments) with higher income and the capital of the country (Montevideo) are the main sources of domestic tourists. The research also shows that the departments that share a border have greater tourist flows between each other. In addition, departments with ocean coasts or good quality accommodation have a significant comparative advantage over the others.
On the demand side, income earnings of people, as well as being determinant of the number of tourists that a department emits, has an elasticity greater than unity, showing that domestic tourism behaves as a luxury good. It is important to highlight the role that the capital as an issuer of tourists. Montevideo is the city where the main terminals of public transport are located and begins the national road network, allowing direct connection of the city with any department of the interior. The results suggest also that strategies of communication and promotion of tourism products and destinations, both public and private utilities, are oriented towards/from the capital (Montevideo) and the regions with higher levels of population and income. In this regard, it is important to think of better transport links between the different departmental capitals that, although have transport infrastructure (terminals and road network), in some cases, do not have direct mobility to all departments if no prior connection to Montevideo.
From the supply-side point of view, the Atlantic Ocean beaches are the main comparative advantage of the departments as a tourist destination. This is the reason that explains the fact that the departments of Maldonado and Rocha concentrate much of the tourism of sun and beach during the summer months. One of the unexpected results of this work was the negative impact on incoming tourism flows of the departments that have coasts on the Rio de la Plata. This result is closely related to the concentration of domestic tourism in the summer season where the preferences of tourists are sun and beach, and the quality of water and sand is better on the Atlantic coast.
Additionally, it was found that the existence of good quality accommodation (3-5 stars) causes a differential effect on the decision about which department tourists choose to vacation. Controlled by the other factors, the existence of differential lodging has a positive and differential effect on the inflow of tourists.
Finally, there is a negative effect on domestic tourism flows if departments share border with Argentina. In these regions, the border effect is clearly unfavourable. Therefore, in these cases should be maximized efforts to improve the competitiveness of services, from the quality of services provided and tourism products offered.
In more general terms, the results show the rationality of that communication strategies and promotion of tourism products and destinations, are geared towards Montevideo and the departments with major population and income per capita. It is also important to address these efforts to neighbours departments, as empirical evidence shows a positive relationship between tourist flows and the fact that departments share administrative boundaries.
In terms of the policy implications of these results, it would be interesting to think of departmental or regional agreements for the creation of a network of promotion, where the adjoining departments can benefit from the implementation of joint tourist promotion strategies.
Finally, thinking of an efficient marketing, this work provides relevant information on the system components of internal tourism in Uruguay: the peculiarities of the source market of tourists in the country and also provides information on the competitive position of destinations. Taking into account these information would help to attract and retain domestic tourists.
Thinking in extensions of this work, a first one that emerges is the incorporation of the information for the years 2013 onwards (not available at the date of preparation of this paper). For example, extending the information a couple of years would enable applying other estimation techniques (pool cross section and panel data) that would enrich the analysis. In turn, it would make it possible to analyse the robustness of the results obtained by applying alternative estimation methods (Models Zero Inflated, etc.) and to explore a better way to capture the effect of the (not significant as these results) multilateral resistance.
Moreover, the desegregation of analysis between tourist flows corresponding to regular trips, no regular trips and excursions, could yield to relevant results for public policy. Regular trips have different characteristics than the other (more associated with vacation travel) and it is important to considering them separately.
Finally, another possible variant of this study is to consider a different regional disaggregation, for example using the regionalization criteria defined by the Ministry of Tourism and Sports, grouping them in six tourist areas (Montevideo, Southeast, Central, South west, Coast, North). Classification-JEL: R1 Keywords: Turismo Interno, Modelo Gravitatorio, Pseudo Máxima Verosimilitud de Poisson, Uruguay, Domestic tourism, Gravity model, Poisson Model Pages: 43-78 Volume: 1 Year: 2017 File-URL: http://www.revistaestudiosregionales.com/documentos/articulos/pdf-articulo-2512.pdf File-Format: Application/pdf Handle: RePEc:rer:articu:v:1:y:2017:p:43-78