A Nonparametric Multivariate Framework With Stochastic Forecasting: Evaluating Public Pension Service Quality

Evaluation of service quality in public systems represents a critical challenge in the context of increasing demographic pressure and economic uncertainty. Pension systems, such as the Italian system managed by INPS, can be interpreted as complex service infrastructures whose quality depends on financial sustainability, efficiency, and accessibility. This paper proposes an integrated statistical framework combining nonparametric multivariate analysis and stochastic forecasting to assess the quality and sustainability of pension services. The empirical analysis is based on official data from INPS and ISTAT and is conducted across Italian macro‐regions (North, Center, South), capturing territorial heterogeneity. The methodology integrates Spearman rank correlation, variable selection based on collinearity thresholds, and ARIMA time series models, providing a unified analytical pipeline. Digitalization is considered an institutional context that may affect the interpretation of empirical findings; however, it is not explicitly incorporated into the model due to the lack of consistent historical indicators. The results reveal strong dependence relationships among key variables, significant territorial disparities, and increasing uncertainty in future pension dynamics, particularly in Southern Italy. The proposed framework provides a robust and flexible tool for evaluating public service quality and offers relevant insights for policy‐making.