Abstract
Standard regression quantiles, which are commonly used in heteroscedastic regression models, are highly vulnerable with respect to the presence of leverage points in the data. The aim of this paper is to propose a novel robust version of regression quantiles, which are based on the idea to assign weights to individual observations.
The novel method denoted as least weighted squares quantiles (LWSQ) is applied to a world tourism dataset, where the number of international arrivals is modeled for 140 countries of the world as a response of 14 pillars (indicators) of the Travel and Tourism Competitiveness Index (TTCI). Here, the economic motivation is to investigate whether tourism competitiveness promotes tourism performance.
The data analysis reveals the advantages of LWSQ. Particularly, LWSQ is able to clearly outperform standard regression quantiles in several artificially contaminated versions of the tourism dataset.
From the economic point of view, the study determines countries which are not effective in transforming their competitiveness to higher levels of tourist arrivals.