Economic crises and market performance—A machine learning approach

Abstract

This note analyzes the relationship between economic crises and tourism performance in Spain during the period 1970–2013 using machine learning techniques. Specifically, a regression tree is estimated to confirm that, although the dynamics of Spanish tourism performance is influenced by the general variables established by the literature, the crisis periods disrupt the natural functioning of these dynamics, provoking disturbances that affect the tourism market position of destinations to a greater extent than expected. Conversely, to other econometric techniques, machine learning approach allows us to achieve greater flexibility and enriches the information, estimating the interrelations and thresholds operating in this context.