Analyzing the Influence of Transportation and Macroeconomic Determinants on Chinese Inbound Tourism: A Markov Switching Model Using Ridge and Lasso Estimation

Abstract

This empirical study examines the influence of transportation and Macroeconomic determinants on Chinese inbound tourism. The purposes of this study are: (1) To measure the cycle of Chinese inbound tourism and find the high-growth stage and low-growth stage; (2) To investigate the impact of transportation and Macroeconomic factors on Chinese inbound tourism. All data used in this paper are secondary annual data from 1995 to 2018. The tourism demand equations are estimated by the Markov switching model based on the Ridge and Lasso estimation. The innovations of this paper are: (1) introducing the transportation factors into tourism demand; (2) because the data samples used in this paper are small, it is impossible to use the traditional Markov switching model for regression. Therefore, innovation is made in the model, namely the Markov switching model based on the Ridge and Lasso estimation method. The results are as follows: (1) the nonlinear model is more suitable for analyzing the tourism business cycle than the linear model. (2) Feature selection screened out the main factors that affect China’s inbound tourism. They are lnHWL (China highway length), lnIAPT (China international air passenger traffic), lnNLR (China navigable length of the river), lnADU (China aircraft daily utilization and UNPR (China’s unemployment rate). (3) For China’s inbound tourism business cycle, the possibility of being in a high-growth stage is large, with long persistence and stability. The possibility of being in a low-growth stage is small, with short persistence and variability.