A Spatial Analysis of International Tourism Demand Model: The Exploration of ASEAN Countries

Abstract

© 2020, Springer Nature Switzerland AG. The main objective of the paper is to apply Bayesian statistics to the panel linear regression models for understanding the tourism demand function in 7 countries of South East Asia (Brunei, Indonesia, Malaysia, Singapore, Thailand, Vietnam, and the Philippines) regarding the spatial effect. The observed panel data is an annual range between 2013 and 2019. The dependent variable is the number of international tourists. The independent variables are world gross domestic products, world prices for jet fuel, domestic hotel rental prices, exchange rates, average annual temperature, and visibility. In the first methodological part, exogenous variables are investigated by the least absolute shrinkage and selection operator (LASSO) regression for validating the set of predictable variables. For the second section which is the highlight, three types of linear panel regression models such as pooled regression, spatial lag regression, and spatial errors regression are used for Bayesian approach. With comparing by deviance information criterion (DIC), the spatial lag regression (pure space-recursive model) is the most appropriate estimation can be proceeded to decide tourism policies for this equator continent.