Recent Advances in the Construction of Nonparametric Stochastic Frontier Models

Abstract

The growth of semi- and nonparametric methods to estimate the stochastic frontier model has expanded rapidly in the preceding years. This chapter provides a critical eye toward this burgeoning and important literature, highlighting the various approaches to achieving near-nonparametric identification. From here, the importance of the relaxation of various modeling assumptions, issues of implementation and interpretation are offered to ease access to these approaches. Finally, several insights into what to date has seen limited focus, inference, are provided along with avenues for future research. The chapter curates the large literature using a consistent notation and describes the pros and cons of the available estimators for various features of the stochastic frontier model.