Tourism and economic growthcasuality, persistence and asymmetry

Abstract

This dissertation provided new empirical evidence on the relationship between tourism and economic growth extending previous literature thanks to the use of recent and alternative econometric approaches including the use of structural breaks and nonlinear models in both the short and longer term. Furthermore, this work provided two empirical characterizations both for the supply and for the influence of the national and international business cycle of the Spanish Tourism Industry in order to contribute to a better knowledge about the demand and the regional distribution of this industry. This dissertation is organised around four empirical essays. Two of them where the relationship between tourism and economic growth is studied in both the short and longer term, and two others where the analysis of the Spanish Tourism Industry works as nexus among them. Based on these analyses the results of the first work point to the existence of three different stages in the last four decades. The main breaks seem to take place at the beginning of 1985 and during the second half of the �90s. The results suggest that the relationship between these two variables can be likened to a sowing and reaping process in which Spain collected the benefits of having invested in tourism specialization during many years -before the first structural break-. From 2000 onwards there exists causality on both directions meaning the influence between the economic growth and the tourism development is reciprocal. Turning our attention to the exploration of the long term, this thesis also provided evidence of two different regimes separated by a threshold suggesting the existence of persistence in both regimes. Moreover, the parameter indicating that the tourist activity is relevant for the output growth turns to be significant, thus confirming the Tourism-Led Growth hypothesis again. These set of results have powerful policy implications. In the first place, shocks affecting the Spanish tourist sector will cause not only temporary effects but also permanent effects in the long run, implying that any measures adopted to modify the situation of this sector will have long-lasting effects and will remain active in the future. Plus, the fact that there is persistence in its behaviour can be regarded as a sign of stability and of absence of volatility in the market. Furthermore, as I mentioned above, the second part of this disseitation tried to characterize the Spanish tourist sector by means of statistical tools. First, we examined the hypothesis of full convergence and club convergence amongst twelve different Spain�s tourist source markets. As expected, the hypothesis of full convergence amongst the different Spain�s source markets was not accepted. However, we were able to identify three different convergence clubs. One group comprises Germany, UK, Portugal, Italy, Netherlands and Belgium. Another group is formed by France, Switzerland and the rest of Europe. The last one contains USA, the rest of America and the rest of the World. Our results also indicates a lower dependence on the classic source markets -particularly Germany and UK-, that is, there has been a diversification process in the Spanish tourist sector. However, this diversification is still slow and is mostly due to a higher degree of participation of tourists from France and other European countries rather than an increase on the relative weight of the American tourists. We can then conclude that the marketing actions and policy measures performed do not seem to be as effective as one might expect, perhaps derived from an undifferentiated approach. Policy makers may this convergence clubs identified to detect common features in the countries of each cluster in order to redirect their efforts in a more precise manner, with higher levels of differentiation. The last chapter helped to have a better understanding about the relationship of interdependence between arrivals, and the business cycle in both in Spain and in the country of origin. By using a forecast error variance decomposition technique the work identified the relative importance of one variable's shock in explaining the predictive error variability of another at different time leads. To conclude, we would like to highlight some limitations concerning our research. First, the use of longer data set or even panel data sets would have been of interest in order to identify previous structural breaks in earlier decades or nonlinear models for panel data. Moreover, it would be appropriate to follow the same procedure that has been applied through the present document incorporating new variables such as tourism income or tourism arrivals to test the consistency of the results. In that regard, the structural break happening in 2008 should be re-examined with those other variables to test whether the financial crisis made an impact on the causality relationship between economic growth and tourism in Spain, since it is possible that the number of nights spent does not help to detect further effects. Second, due to a lack of differentiation in the data concerning different source markets of America, we have not been able to detect further convergence-clubs amongst American countries. Further research should be developed in order to solve this issue, since it is not reasonable to develop a similar strategy for such a diverse and large group of countries. Third, as this is the first time that the issue of convergence amongst Spain�s tourism markets has been addressed, we cannot cany out any comparisons with previous findings in the literature, which makes later investigations necessary to test the robustness of our results. Nonetheless, our contribution should help to pave the way for future works about convergence in Spain and also enables replication studies for the case of other countries� tourism markets.