Stochastic Frontier Analysis
摘要
This chapter reviews the theoretical foundation of SFA and the shortcomings of traditional deterministic frontiers (which neglect random factors). It then discusses the application of SFA to measure energy efficiency and shadow prices, before extending to Meta-SFA for addressing technological heterogeneity. The chapter further covers the development of spatial stochastic frontier models (SAR-SF, CSSF, SDF-CSD, SDF-STE, endogenous spatial SFA), which integrate spatial econometric techniques to capture efficiency spillover effects under geographic or economic linkages. These models address the biases caused by ignoring spatial dependence and endogeneity in traditional models, thereby expanding the applicability and explanatory power of efficiency analysis.