Consistent method of moments estimation of the true fixed effects model

Abstract

Abstract. A couple a years ago William H. Greene introduced the so called 'True fixed e↵ects estimator' (TFE), which is intended to separate between heterogeneity and e ciency in stochastic frontier analysis. We would say that it has had huge impact on applied stochastic frontier analysis. One problem with the original TFE estimator, is that it is biased in cases with finite time observations. For the normal-half-normal model this problem was solved by Chen et al. (2014) based on maximum likelihood estimation of the within-transformed model. In this study we show the possibilities with method of moment estimation. This approach is more straightforward computational and is more flexible than maximum likelihood estimation since the estimators are not as dependent on the distributional assumptions and do not hinges on an explicit distribution of the random error. We only assume symmetry and for more complicated models also a fixed fourth order cumulant.

Estimating efficiency effects in a panel data stochastic frontier model

“…In another related research, Wikstrom (2015) suggests a class of consistent method of moment estimators that goes beyond the normal half-normal TFE model proposed by Greene (2005). This is demonstrated by deriving a consistent normal-gamma TFE estimator.…”

Modelling Efficiency Effects in a True Fixed Effects Stochastic Frontier

“…They also conduct a simulation exercise and did not encounter any bias in the estimation of variance that Greene (2005) In another related research, Wikstrom (2015) suggests a class of consistent method of moment estimators that goes beyond the normal half-normal TFE model proposed by Greene (2005). This is demonstrated by deriving a consistent normal-gamma TFE estimator.…”

Efficient estimation of true fixed-effects stochastic frontier models

“…Another contribution is Wikström (2015), who defines a method-of-moments estimator and provides a simulation study for the same model considered by Chen et al (2014). The proposal has some interesting properties, such as its logical simplicity.…”

“…On the other hand, it is likely to be less efficient than the MMLE since the latter is based on a marginal likelihood (see Pace and Salvan 1997, Chap. 4), and the method of moments is at times prone to numerical instabilities, as testified by the "problematic replications" reported in Wikström (2015). Belotti and Ilardi (2018) provide some notable estimation strategies.…”