Modelling expenditure in tourism using the log-skew normal distribution

Abstract

Many factors are involved in tourist decision expenses. Such circumstances may give rise to some asymmetry in the distribution of tourism expenditure. We propose in this paper a reparameterization of the three-parameter log-skew normal distribution for modelling the expenditure at the country of origin, at destination, and total expenditure in a tourism setting. This distribution seems to fit the expenditure data satisfactorily in all the parts of the empirical distribution. In particular, the proposed model is well suited to capture the skewness and kurtosis that may be present and the long tail to the right that the three variables mentioned above tend to present in practice.