Abstract
Digital transformation of public administration increasingly intersects with business freedom within institutional environments shaped by transparency, responsible governance, institutional accountability, and the quality of state–business interaction. Previous research has extensively examined digital government in relation to administrative efficiency, innovation, and economic growth yet has provided limited cross-country evidence on how the transparency and accountability of multidimensional digital governance are associated with the institutional conditions under which businesses operate. This study assesses the relationship between digital government maturity and business freedom and determines which dimensions of government technology have the strongest statistical and predictive relevance. The empirical analysis covers 182 countries in 2020, 2022, and 2025, combining the Business Freedom Index with 43 World Bank government technology indicators aggregated into 23 composite dimensions. The methodology integrates descriptive statistics, correlation and multicollinearity analysis, the Kruskal–Wallis test with pairwise comparisons, fixed-effects panel regression with Arellano robust standard errors, eight predictive algorithms, permutation feature importance, Shapley additive explanations, and regional analysis implemented in Python and Gretl. The final fixed-effects model identifies Digital Policy and Regulation as exhibiting the strongest positive within-country association, with a coefficient of 3.271 and a probability value below 0.001, while Debt Management and Digital Identification have coefficients of −1.991 and −1.102, indicating that the regulatory and institutional dimensions of digitalization are particularly relevant to variation in business freedom. Support vector regression achieves a test coefficient of determination of 0.497 with a train–test R² gap of only 0.031, whereas the inclusion of regional affiliation raises the test coefficient of determination to 0.603 and reduces the test root mean squared error from 12.55 to 11.14, indicating that broader regional context provides substantial additional predictive information. Permutation and Shapley-based interpretation identify Online Public Services, Digital Tax, Emerging Technologies, Government Architecture, Employment Services, and Open Government among the most influential predictors, while Europe reaches a permutation importance of 0.241, substantially exceeding that of any individual digital-government component. The findings provide an empirical basis for strengthening transparency, responsible governance, and institutional accountability in digital public administration and for future research on how ethically grounded state–business interaction can translate government technology maturity into more favorable conditions for business freedom.