Using quantile regression to understand visitor spending

A common approach to assessing visitor expenditures is to use least-squares regression analysis to determine statistically significant variables upon which key market segments are identified for marketing purposes. This was done by Wang (2004) for survey data based on expenditures by Mainland Chinese visitors to Hong Kong. In this research note we use this same dataset to demonstrate the benefits of using quantile regression analysis to better identify tourist spending patterns and market segments. The quantile regression method measures tourist spending in different categories against a fixed range of dependent variables, which distinguishes between lower, medium, and higher spenders. The results show that quantile regression is less susceptible to influence by outlier values and is better able to target finer tourist spending market segments