Introducing a tourist tax in a mature destination: Evidence from a beta-binomial model

Abstract

This study analyzes tourists’ stated willingness to pay (WTP) for a tourist tax using a beta-binomial model. WTP responses take discrete bounded values between €0 and €4 per night, which standard count models with unbounded support cannot accommodate. Our framework allows flexible dispersion patterns and incorporates covariates through a bounded logistic link. Using survey data from 35,377 tourists departing a mature sun-and-beach destination (2024Q2–2025Q1), we find that income, country of residence, accommodation type, length of stay, and satisfaction explain WTP heterogeneity. Out-of-sample validation on 17,010 observations from 2025Q2–Q3 shows well-calibrated probabilistic predictions, yielding a 35% Ranked Probability Score gain over a naïve baseline. Translating the fitted WTP distribution to a universe of 15.0 million tourists per year, a voluntary €2/night tax yields €131 million annually with 52% acceptance, while mandatory rates above €2 raise gross revenue at the cost of a rapidly rising burden on tourists beyond their stated tolerance.