Identifying the Frontier Structural Function and Bounding Mean Deviations

2025-04-28

AI Summary

This paper addresses key limitations of stochastic frontier analysis (SFA)—namely, its reliance on instrumental variables or strong parametric assumptions regarding unobserved inefficiency. We propose a nonparametric identification strategy for the frontier structural function that requires no instruments and does not assume independence between inefficiency and inputs. The model specifies the outcome as the frontier function minus a nonnegative inefficiency term, whose distribution may vary with inputs. By imposing inequality constraints on conditional moments—specifically, variance and skewness—we achieve pointwise identification of the frontier function for the first time and derive a tight lower bound on the mean inefficiency. This bound is input-adaptive, relaxing standard SFA assumptions of inefficiency independence and parametric distributional form. Empirically, the method accurately disentangles the technological frontier from the efficiency distribution, and exhibits markedly superior finite-sample robustness compared to constant-shift estimators.

Technology Category

Reasoning under Uncertainty: Stochastic OptimizationSearch and Optimization: Non-convex OptimizationMachine Learning: Structured Learning

Application Category

Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphsSearch and Retrieval-Augmented AI: Efficiency and scalability of Web search enginesSemantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics

Abstract

This paper analyzes a model in which an outcome variable equals the difference between a frontier function of inputs and a nonnegative unobserved deviation. If zero is in the support of the deviation at a given input value, then the frontier function is identified by the maximum outcome there. This obviates the need for instrumental variables. Implementation requires allowing for the distribution of deviations to depend on inputs, thus not ruling out endogenous inputs and ensuring the estimated frontier is not merely a constant shift of a biased conditional expectation. Including random errors results in a stochastic frontier analysis model generalized to allow the joint distribution of deviations and errors to depend on inputs. If the minimum deviation is a function of inputs, then we derive a lower bound for the mean deviation using variance and skewness, without making parametric distributional assumptions. We apply our results to a frontier production function, with deviations representing inefficiencies.

Problem

• Identifies frontier function without instrumental variables
• Allows deviation distribution to depend on inputs
• Derives mean deviation bounds using variance and skewness

Innovation

• Frontier function identified by maximum outcome
• Distribution of deviations depends on inputs
• Lower bound for mean deviation using variance