Abstract
Tourism demand is one of the major areas of tourism economics research. The current research studies the interdependencies of international tourism demand across 24 major countries around the world. To this end, it proposes to develop a tourism demand model using an innovative approach, called the global vector autoregressive (GVAR) model. While existing tourism demand models are successful in measuring the causal effects of economic variables on tourism demand for a single origin-destination pair, they tend to miss the spillover effects onto other countries. In the era of globalisation, tourism destinations become interdependent on each other. Impacts of a distant event can be transmitted across borders and be felt globally. Hence, modelling international tourism demand requires one to go beyond a particular origin-destination pair, and take into account the interdependencies across multiple countries. The proposed approach overcomes the ‘curse of dimensionality’ when modelling a large set of endogenous variables. The empirical results show that, to different extents, co-movements of international tourism demand and of macroeconomic variables are observed across all the 24 countries. In the event of a negative shock to China’s real income level and that to China’s own price level, it is found that in the short run, almost all countries will face fluctuations in their international tourism demand and their own price. But in the long run the shocks will impact on developing countries and China’s neighbouring countries more deeply than on developed countries in the West. The current research contributes to the knowledge on tourism demand. It models tourism demand in the setting of globalisation and quantifies the interdependencies across major countries. On the practical front, tourism policy makers and business practitioners can make use of the model and the results to gauge the scale of impacts of unexpected events on the international tourism demand of their native markets.