Spatial tourism supply: The case of ASEAN-5 countries

Abstract

This article is the applied econometric analyses in estimating the spatial stochastic frontier (SSF) model. The heart of the modelling inference is Bayesian statistics with Gibbs sampling, which can provide flexible parametric distributions for capturing abnormal distributional observations. Yearly panel data between 2008 and 2018 regarding tourism supplies such as numbers of hotels, service rooms, and labour forces in ASEAN-5 countries (Vietnam, Indonesia, Malaysian, Singapore, and Thailand) is observed. The empirical results are divided into two major parts. The first section is the understanding of spatial effects which causally relate to the way for expanding tourism supplies is explained by the model validation using deviance information criterion (DIC). The panel spatial-error stochastic frontier (random-effect model) contains the minimum DIC which implies the parameters of locational effects is the minor, but it should be still considered for stimulating tourism supply sides in these five countries. The second section is the estimation of random parameters are processed to monitor technical efficiencies. Spatial effects are a specific impact on the group of connected borderline countries. On the other hand, this impact is dependently reduced when a country contains a remote area condition. Ultimately, this paper is the alternative to better analyse supply-chain management in the tourism sector without the restrictions of P-value, asymptotic struggling, and multicollinear problems. For further researches, subjective statistics (Bayesian) is the main issue that applies for modern econometrics.