This study focuses on the underpinnings of the households’ tourism expenditure decisions during the global economic crisis in 2009. In particular, this study tests if during an economic crisis, decisions on tourism expenditure depend on climate conditions of the place of origin, GDP and GDP growth, among other well-known determinants. It should be noted that cutback decisions on tourism expenditure are not independent of destination choice, and for that reason the model requires the estimation of both decisions simultaneously.
The methodology proposed in this paper represents a new way of analyzing the impacts of an economic crisis on tourism expenditure. Two levels of analysis can be considered. On the one hand, macroeconomic data of tourism expenditure is usually explored. On the other hand, the microeconomic analysis of the household and regional variables of their environment that may enrich the analysis. If the econometric model takes into account all these variables simultaneously, then the linkage between GDP changes and tourists´ behavior is enriched and it may be estimated more accurately. As far as we know, this paper is the first study that models the cutback decision on tourism expenditure. Modeling such decision is a challenge because it is not independent of the destination choice. For instance, households that travel domestically may not be as sensitive to the crisis as those who travel abroad. For this purpose, the econometric model employed is a simultaneous system of cutback decision and destination choice. More precisely, Simultaneous Semi-Ordered Bivariate Probit has proved to be the most useful econometric model for the estimation because it deals with the simultaneity of the cutback and destination choice decisions as well as the endogeneity.
This research has proved that during an economic crisis, households react cutting back their tourism expenditure depending on GDP, GDP growth, and climate in their place of origin. In that sense, there are marked differences between North-European and Mediterranean regions. It is interesting to note that regions with bad climate are less likely to cut back than those households located in regions with good climate.