Explainable Machine Learning Financial Econometrics for Digital Inclusive Finance Impact on Rural Labor Market

Huanhao Chen, Yong Chen, Jiaxuan Wu, Xiaofei Du

The research examines how digital inclusive finance reshapes the rural labor market using an auditable index system and an interpretable learning pipeline. We construct a four-pillar framework for the rural labor market covering labor behavior, labor structure, security and fairness, and sustainability, and compute county-level scores with an Attribute Hierarchy Model plus Fuzzy Comprehensive Evaluation (AHM–FCE). Using data for 58 counties in Jiangsu from 2014 to 2023, we estimate nonlinear links from overall and sub-dimensional digital finance to labor market outcomes with a random forest optimized by Particle Swarm Optimization plus Genetic Algorithm (PSO-GA-RF). Theoretical contribution: we provide a measurement-based bridge from digital inclusive finance to rural labor markets by aligning access, usage, and service quality with the four pillars of the rural labor market index, which yields testable county level predictions on participation, job quality, equity, and persistence of gains. Maps show heterogeneity, with higher behavior scores, lagging sustainability, and a north–south gradient. Empirically, stronger digital finance is associated with higher non-agricultural employment, better job quality, narrower urban–rural gaps, and stronger protection mechanisms, with larger effects where rural population shares and policy support are higher. Findings are robust to variable transforms, bandwidth choices, and tuning.