Ivan Sever (p), Ana Perišić, Neven Ivandić

Annalisa Stacchini

This paper contributes to the ongoing efforts for assessing tourism sustainability by focusing on the analysis and interpretation of the perceptions of local residents in relation to the thresholds of social carrying capacity. More specifically, we propose a statistical framework that supports identification of risk factors contributing to negative perceptions of tourism, rather than focusing on identification of a clear threshold for a given sustainability indicator. This framework is based on explanatory and predictive modeling and adopts multiple regression analysis, dominance analysis and random forest method to identify the risk factors and their relative importance. Estimating the relative importance of each risk factor provides a means to prioritize management and monitoring of sustainability indicators. The proposed statistical framework is intuitive and its usefulness is demonstrated on the case of the city of Split, one of the major tourism destinations in Croatia. The findings demonstrate that apartmentization, and perceived changes in city appearance and its authenticity are the key risk factors that affect overall perception of tourism development in Split.

Henri L.F. de Groot (p), Thomas de Graaff, Elisa Panzera

Ivan Sever

There is a large and growing body of literature modeling tourism flows using gravity models. This meta-analysis summarizes and explains the variation in the deterring effect of distance on tourism flows by analysing 762 estimates from 119 primary studies utilising data covering the last two decades. We find substantial heterogeneity amongst studies that mostly correlates with (unobserved) study characteristics, estimation methods, and locations of origin and destination. We make the following five contributions to the literature. First, we confirm previous findings that the mean distance-decay effect is close to unit elasticity in absolute value (−0.92). Second, we argue that this is a total effect as we find that, controlling for mediator variables, the direct effect between distance and tourism flows is substantially lower (−0.68). Third, we document a wide range of mediator variables yielding significant associations with the total effect of distance, such as adjacency, world heritage sites, exchange rates and island destinations. Fourth, we do not find changes in the distance-decay effect over the last two decades. And, finally, we point out that our findings indicate a positive relation between distance and the total amount of tourists.

Hans Kremers (p), Mario Fortuna

Henri L.F. de Groot

The breakdown in the global supply chains following the recent pandemic have had serious implications on international travel leading up to a subsequent decrease in international tourism demand in the Azores. We apply a regional static single country computable general equilibrium model to quantify the net economic cost of the COVID19 pandemic under three different counterfactual equilibria. The model is an improvement of an existing computable general equilibrium model in that it is calibrated on a Social Accounting Matrix (SAM) adjusted to satisfy recommendations with respect to tourism statistics formulated by the WTO and other international organisations. Furthermore, it provides a proper inclusion of travel induced rationing in export demand of tourism related products. The main idea of the paper is to initialise further application of computable general equilibrium modelling in the area of tourism.

Annalisa Stacchini (p)

Hans Kremers

This study offers an innovative contribution to the transport and tourism research, by proposing a new strategy for estimating the environmental impacts of tourism-related transport, skipping the difficulties caused by the unavailability of microdata relevant to accurately quantify emissions intensities. Instead of computing the quantity of emissions deriving from tourist mobility, we estimate the percentage change in overall emissions for a 1% increase in tourist transport-related variables. Since the presented strategy is data-driven, the need for assumptions and subjective decisions is minimized, ensuring a high degree of transparency. The assumptions required regard only the variables distribution, but they can be tested inferentially. Contrarily to extant methods, our strategy allows us to detect possible energy efficiency and emission efficiency gains (or losses) occurred over time.
While most extant literature has considered carbon emissions only, we analyze a multiplicity of environmental quality indicators and exploit the correlations among them in a two-fold imputation procedure based on Random Forest. This way, a balanced panel dataset can be obtained from official environmental data with many missing values. The proposed strategy permits analysts to also estimate mitigating effects hopefully introduced by innovative technology, sustainable behaviors or policy interventions. These are normally carried out at a sub-national level. Hence, when research is aimed at assessing mitigating effects, we include in our strategy small area estimation methods to increase estimates’ accuracy and precision. In particular, we specify autoregressive normal-normal and beta-logistic hierarchical Bayesian area-level models. Compared to extant environmental impact assessments at the destination level, our method allows us to quantify the overall reliability of results, besides providing a more holistic representation of environmental impacts.
We illustrate the propounded strategy through an empirical application to Italian destinations. In Italy the data availability issue is very serious. Not even the destination of domestic tourists is disclosed. Environmental data, full of missing values, are published only for 25 provinces (NUTS3 level) out of 110. Consequently, the obtained results are completely unreliable, despite our analytical effort. This highlights that, even with a methodology that maximizes the exploitation of the available information, a definitely greater data availability is required for modelling a complex problem, such as the environmental impacts of tourism-related transport. Hence, our empirical application represents a call to Italian institutions to improve the collection and publication environmental data, by ensuring consistency over time in the definition of variables, territorial comprehensiveness and transparency.