Estimating Cultural Tourists in Spanish Regions Through Machine Learning

Cultural tourism is receiving increasing attention and plays a significant role in shaping the global tourism landscape. However, defining cultural tourism remains ambiguous, and there are no standardized methods to measure its extent. This study introduces a novel methodology to quantify cultural tourists in Spanish regions. The proposed approach involves calculating the elasticity of tourist flows in a region concerning the dimension of the cultural and creative sectors. To achieve this, machine learning techniques, in particular Causal Forest, are applied, employing a comprehensive database that gathers information from European regions spanning the years 2008 to 2019. A counterfactual scenario is simulated, assuming the absence of cultural and creative workers in each region in order to identify the number of overnight stays attributed to cultural tourism. The results indicate that cultural tourism accounts for 18.6% of total overnight stays in Spain, above the traditional method that accounted for 10.1% of overnight stays. Though there are important differences between regions, ranging from 65% in Galicia to almost negligible in the Canary Islands or Cantabria. The proposed methodology is promising and could become an acceptable and comparable standard for other regions.