Fiscal Sustainability Investigation Based on Cluster Analysis and Panel CS-ARDL Models

This paper examines the hypothesis of fiscal sustainability for 88 countries by analyzing governments’ responses to increases in public debt from 2000 to 2024. To address the challenges of heterogeneity in large panels, we propose a clustering technique based on fiscal indicators to group countries and estimate group-specific CS-ARDL models. The clustering approach allowed us to identify which variables played the most relevant role in classifying the groups with regard to fiscal sustainability, providing valuable information to guide corrective measures. Additionally, we analyzed the impact of COVID-19 and found that the economies with stronger payment capacity only experienced slight reductions in their fiscal responsiveness in the post-pandemic period, whereas some countries with weaker fundamentals transitioned to a fiscal unsustainability classification.