From impact to action: Enhancing international tourism resilience through counterfactual explanations
METADATA ONLY
Loading...
Author / Producer
Date
2026-06
Publication Type
Journal Article
ETH Bibliography
yes
Citations
Scopus:
Altmetric
METADATA ONLY
Data
Rights / License
Abstract
COVID-19, as an unexpected global shock, had a profound impact on international tourism and created a unique testing ground for empirical research on external shocks. However, existing studies lack a systematic analysis of the heterogeneous impact of COVID-19 on tourism recovery patterns across countries, providing limited insights into the actionable strategies to enhance recovery. This study quantitatively assesses the impact of COVID-19 on international tourism demand across 53 countries and provides actionable advice for counterfactual scenarios through a three-stage analytical framework. The first stage revealed that resilience varies considerably between countries, with the recovery trajectories falling into three distinct categories: V-, U-, and W-shaped patterns. Conceptualizing resilience as a process involving the resistance, adaptation, and recovery phases, the second stage employs econometric modeling within the framework of regional economic resilience. It identifies key factors significantly influencing the recovery stage. Finally, the third stage employs an interpretable machine learning method, counterfactual explanations, to assess how changes in the important factors might improve recovery outcomes. These analyses provide a comprehensive, data-driven perspective for examining the heterogeneity and mechanisms of international tourism resilience across countries in the aftermath of COVID-19.
Permanent link
Publication status
published
External links
Editor
Book title
Journal / series
Volume
114
Pages / Article No.
105356
Publisher
Elsevier
Event
Edition / version
Methods
Software
Geographic location
Date collected
Date created
Subject
Resilience; Counterfactual explanations; International tourism; Covid-19; Tourism demand