Decision tree-augmented stochastic frontier analysis

Authors

• Ming-Yu Deng, Anhui University
• Levent Kutlu, The University of Texas Rio Grande ValleyFollow
• Xi Mao, The University of Texas Rio Grande ValleyFollow

Document Type

Article

Publication Date

5-2026

Abstract

We propose a decision tree-augmented stochastic frontier analysis (DT-SFA) strategy that embeds an SFA model-based recursive partitioning algorithm into the conventional classification and regression tree. Our methodology accounts for heterogeneity across subpopulations by generating segmented (local) SFA models, and makes it effectively visualized for analyzing and interpreting technical efficiency. Our empirical application to Chinese industrial enterprises illustrates that DT-SFA outperforms a standard single-frontier SFA in fit, which is consistent with substantial technological heterogeneity among firms. The approach preserves a parametric structure that supports statistical inference.

Comments

Original published version available at https://doi.org/10.1016/j.econlet.2026.112943

Recommended Citation

Deng, M.Y., Kutlu, L. and Mao, X., 2026. Decision tree-augmented stochastic frontier analysis. Economics Letters, p.112943. https://doi.org/10.1016/j.econlet.2026.112943

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.

Publication Title

Economics Letters

DOI

10.1016/j.econlet.2026.112943