Abstract
This article examines the time series behavior of two important tourism statistical variables: ITR (International Tourism Receipts) and NITAC (Number of International Tourist Arrivals to China) over the past two decades (1984-2004). Our unit root tests reveal that both series are integrated of order one, with the shocks having permanent effects. Information criteria used for the unit root tests suggest ARIMA (2, 1, and 2) and ARIMA (1, 1, and 0) models for series ITR and NITAC, respectively. No tests are statistically significant at the 5 percent or higher confidence levels, which indicates that the data are completely random and the models are adequate. The models can best describe the behavior of tourism growth variables.