Integrating behavioral insights into cleaner production efficiency: a prospect theory-enhanced DEA analysis of the automotive industry

This study evaluates the operational cleaner-production performance of automotive vehicle segments using a Prospect Theory-enhanced Data Envelopment Analysis (PTE-DEA) framework. The unit of analysis is defined from manufacturer-level decision-making units to manufacturer–vehicle segment-level decision-making units to capture heterogeneity across vehicle portfolios. Using the 2024 dataset, 50 decision-making units (DMUs) are considered, each of them represents a specific combination of manufacturer, regulatory class, and vehicle type. The selected inputs are vehicle weight, horsepower, and footprint, representing operational vehicle burdens, while the selected outputs are real-world MPG, real-world MPG city, and real-world MPG highway, representing fuel-efficiency-based cleaner-performance outcomes. The baseline PTE-DEA results indicate that Mazda (Sedan/Wagon), Subaru (Sedan/Wagon), Stellantis (SUV), Volkswagen (Sedan/Wagon), and Kia (Sedan/Wagon) form the leading segment-level group. Production-weighted manufacturer-level aggregation further shows that Subaru, Volkswagen, Nissan, Honda, Mazda, Toyota, and Kia demonstrate relatively strong portfolio-level cleaner-performance outcomes. Robustness analyses, including 600 behavioral parameter scenarios, rank-correlation assessment, and Monte Carlo perturbation, confirm that the leading segment-level patterns are generally stable. The study contributes to cleaner-production benchmarking by demonstrating how segment-level heterogeneity, behavioral assumptions, and fuel-efficiency-based operational indicators can be integrated into a more discriminative PTE-DEA framework. The findings provide a scenario-based decision-support perspective for automotive manufacturers and policymakers seeking to evaluate operational cleaner performance across heterogeneous vehicle segments.