NONPARAMETRIC IDENTIFICATION AND ESTIMATION OF STOCHASTIC FRONTIER MODELS

<p>This dissertation studies nonparametric identication and estimation of stochastic frontiermodels. It is composed of three chapters. The rst chapter investigates the identication and estimation of a cross sectional stochastic frontier model with Laplacian errors and unknown variance, which is built on a nonparametric density deconvolution strategy. Chapter two studies a zero-ineciency stochastic frontier model utilizing a penalized sieve estimator, which allows flexible function forms and arbitrary distributions of ineciency. The third chapter explores identication and estimation of a nonparametric panel stochastic frontier model based on Kotlarski's Lemma and moments derived from conditional characteristic functions.</p>

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Economics

Year

2020

Author dc:creator

• Cai, Jun

Contributors dc:contributor

• William W. Horrace

• Deconvolution
• Kernel Estimation
• Panel Data Model
• Productivity
• Sieve Estimation
• Stochastic Frontier Model
• Social and Behavioral Sciences

Repository record dc:identifier

https://surface.syr.edu/etd/1196

OAI identifier oai:identifier

oai:surface.syr.edu:etd-2197