Predicting the Quality of Local Governance: An Artificial Intelligence Approach Applied to the 77 Municipalities of Benin

Abstract

Background: Digital transformation is a key catalyst for improving public administration, yet empirical evidence quantifying its impact on local governance in Sub-Saharan Africa is scarce. This study bridges this gap by developing a predictive framework to assess the relationship between digital maturity and governance performance. Methods: Covering all 77 municipalities of Benin from 2016–2021, this study constructed two composite indices: the Communal Digital Maturity Index (CDMI) and the Local Governance Index (LGI). Data was triangulated from administrative reports, surveys, and public databases. The analytical approach combined multiple linear regression with machine learning algorithms, specifically Random Forest and Neural Networks, to model and predict governance outcomes based on digital indicators. Results: A strong positive correlation (r ≈ 0.71) was found between digital maturity and governance quality. The regression model explained 69% of the variance (Adjusted R² = 0.69), while the Neural Network model achieved a predictive accuracy of 85%. The analysis also revealed a significant digital divide between urban and rural municipalities, a key factor influencing governance performance. Conclusion: Digital transformation is a significant predictor of local governance quality in Benin. The integration of AI provides a robust, actionable tool for policymakers to anticipate the impacts of digital investments, highlighting the need for context-specific strategies to strengthen local governance.