Stochastic Frontier Models with Bounded Inefficiency

摘要: This paper introduces a new model of the stochastic production frontier that incorporates an unobservable bound for inefficiency, which is naturally instituted by market competition forces. We consider doubly truncated normal, half-normal, and exponential distributions to inefficiency component error term. derive analytical form density function term each specification, expressions conditional mean levels, provide proofs local identifiability these models under differing assumptions about deep parameters distributions. examine skewness properties our estimators explanation finding positive (“wrong”) in many applied studies using traditional model. extend panel data setting specify time-varying as well efficiencies. A Monte Carlo study conducted finite sample performance maximum likelihood cross-sectional settings. Lastly, we illustrate use analysis efficiencies US banking industry from 1984 2009 recently developed over 4,000 banks also compare findings those based on set competing specifications find substantial increases efficiency after regulatory reforms 1980s but backsliding during 2005–2009 period presaging financial meltdown experienced worldwide.