Abstract
<jats:p>Problem. The relevance of the study is caused by the need for empirical verification of G. Esping-Andersen’s classical typology of welfare state systems in the context of the global transformation of public financial regulation models. Aim. The aim of this work is to test the stability of the three ideal types (social-democratic, conservative-corporatist, liberal) based on outcome indicators of poverty and income inequality, as well as the standard of living, using a global sample of countries. Methods. The methodological framework consisted of principal component analysis (PCA) and k-means clustering (optimal number of clusters k=9, silhouette coefficient 0.405) based on World Bank data for 2019– 2024 (142 countries), followed by verification using instrumental characteristics of tax and transfer mechanisms. Results. It was established that the classical Esping-Andersen models do not form separate clusters based on outcome indicators but are united into a single cluster, demonstrating outcome convergence (poverty rate 0.6–1.3 %, income Gini coefficient 0.280–0.317). At the same time, instrumental divergence persists, manifested in differences in tax progressivity, the structure of social insurance contributions, and the volume of social transfers. Russia is classified within the subgroup of post-socialist hybrid models with a limited redistributive function and the potential for increasing the progressivity of personal income taxation. A structural division was identified between countries with institutionalised social protection systems and states where informal solidarity mechanisms dominate (clusters 6–9). Conclusions. The conclusion is drawn about the necessity of a two-level typology of welfare state financial models: (1) institutional for upper-middle and high-income countries and (2) oriented towards informal social protection models for low-income economies with underdeveloped welfare systems.</jats:p>