The relationships among the demand for travelling, income of origin country, tourism prices, transportation costs and exchange rates have received considerable attention in empirical tourism research (Dritsakis & Papanastasiou, 1998; Lim, 1999). Even though it is well known empirically that many macroeconomic time series are nonstationary, most published tourism research has estimated static models in logarithmic levels using ordinary least squares. This practice gives rise to invalid inferences, so there is little informational content in examining the alleged significance of the estimated coefficients.