Abstract
The productive efficiency of a firm can be seen as composed of two parts, one persistent and one transient. The received empirical literature on the measurement of productive efficiency has paid relatively little attention to the difference between these two components. Ahn and Sickles (Econ Rev 19(4):461–492, 2000) suggested some approaches that pointed in this direction. The possibility was also raised in Greene (Health Econ 13(10):959–980, 2004. doi:10.1002/hec.938), who expressed some pessimism over the possibility of distinguishing the two empirically. Recently, Colombi (A skew normal stochastic frontier model for panel data, 2010) and Kumbhakar and Tsionas (J Appl Econ 29(1):110–132, 2012), in a milestone extension of the stochastic frontier methodology have proposed a tractable model based on panel data that promises to provide separate estimates of the two components of efficiency. The approach developed in the original presentation proved very cumbersome actually to implement in practice. Colombi (2010) notes that FIML estimation of the model is ‘complex and time consuming.’ In the sequence of papers, Colombi (2010), Colombi et al. (A stochastic frontier model with short-run and long-run inefficiency random effects, 2011, J Prod Anal, 2014), Kumbhakar et al. (J Prod Anal 41(2):321–337, 2012) and Kumbhakar and Tsionas (2012) have suggested other strategies, including a four step least squares method. The main point of this paper is that full maximum likelihood estimation of the model is neither complex nor time consuming. The extreme complexity of the log likelihood noted in Colombi (2010), Colombi et al. (2011, 2014) is reduced by using simulation and exploiting the Butler and Moffitt (Econometrica 50:761–764, 1982) formulation. In this paper, we develop a practical full information maximum simulated likelihood estimator for the model. The approach is very effective and strikingly simple to apply, and uses all of the sample distributional information to obtain the estimates. We also implement the panel data counterpart of the Jondrow et al. (J Econ 19(2–3):233–238, 1982) estimator for technical or cost inefficiency. The technique is applied in a study of the cost efficiency of Swiss railways.
Similar content being viewed by others
Notes
Books and surveys on the measurement of the level of productive efficiency do make the distinction between models that estimate time-varying inefficiency indicators and models that produce time invariant indicators. See, for example, Kumbhakar and Lovell (2003) or Greene (2008). Measurement of the distinct parts has proved challenging.
For a discussion of these models see Greene (2008).
Other extensions of the basic frontier model have also considered exponential and truncated normal distributions for the inefficiency term. See for instance Battese and Coelli (1992).
The received literature provides many alternative stochastic frontier models that allow the estimation of time-varying inefficiency. In addition to those noted above, we would include the model proposed by Lee and Schmidt (1993) that extends the Kumbhakar (1990) and Battese and Coelli (1992) models, and the model proposed by Kumbhakar and Wang (2005). There are others as well. See Greene (2008) or Parmeter and Kumbhakar (2015) for surveys.
The Mundlak auxiliary equation has been proposed for a random effects linear regression model. This approach, based on normality and the linear model, might not strictly apply to stochastic frontier models estimated by ML, as these models possess an asymmetric composite error term ε i . As the model captures the correlation between the individual effects and the explanatory variables at least partly, the resulting heterogeneity bias is expected to be minimal. The general approach has been used elsewhere in a variety of settings under the heading of ‘correlated random effects models.’ See, e.g., Wooldridge (2010).
There is no requirement that the number of observations, T i , be the same for each firm. T i is assumed here to be constant only for convenience to avoid another subscript in the presentation.
The estimation strategy in (11) could, in principle, be applied to estimation of the stochastic frontier model in (6) by integrating u it out of the conditionally normal linear regression model in (5). Maximization of (6) directly is extremely straightforward, however, and the MSLE would provide no improvement over direct MLE.
The authors allowed \(\sigma_{u}^{2}\) to vary by period. This aspect can be added to the estimating equations by changing σ 2 u I to Ψ = diag(σ 2 u,1 ,…,σ 2 u,T ) in the definition of V. In the formulations below, the parameterization in terms of σ and λ would have to be replaced with the original parameterization in terms of σ v and σ u,t . Their results on the presence of this type of heteroscedasticity are mixed.
In the original work, Colombi (2010) notes use of a self-developed R routine named SNF-maxlik.
The computation was done using a modified version of LIMDEP/NLOGIT that will be generally available in 2015. Adaptation of the method to other platforms, such as the mixed estimation packages in R or Stata should be straightforward given the availability of the TRE model to begin with. Running time for estimation of the cost model (20) was approximately 2 min and we did not experience convergence problems. The model is essentially a true random effects model, which has proved to be straightforward to estimate in a number of received applications since 2004. The computation of technical efficiency firm by firm based on (17), which does require a one time, post estimation application of GHK simulation, takes roughly the same amount of time as estimation of the model, itself.
The data set is available at http://people.stern.nyu.edu/wgreene/Text/Edition7/TableF19-1.txt.
The values of the coefficients of the MGTRE are very close to the coefficients obtained using a classical fixed effects model.
References
Abdulai A, Tietje H (2007) Estimating technical efficiency under unobserved heterogeneity with stochastic frontier models: application to northern German dairy farms. Eur Rev Agric Econ 34(3):393–416
Ahn SC, Sickles RC (2000) Estimation of long-run inefficiency levels: a dynamic frontier approach. Econometric Reviews 19(4):461–492
Aigner D, Lovell CAK, Schmidt P (1977) Formulation and estimation of stochastic frontier production function models. J Econ 6(1):21–37
Azzalini A (1985) A class of distributions which includes the normal ones. Scand J Stat 12:171–178
Baranzini A, Faust AK (2014) Water supply: the cost structure of water utilities in Switzerland. J Prod Anal 41:383–397
Battese GE, Coelli TJ (1992) Frontier production functions, technical efficiency and panel data: with application to paddy farmers in India. In: Gulledge TRG Jr, Lovell CAK (eds) International applications of productivity and efficiency analysis. Springer, Netherlands, pp 149–165
Battese GE, Coelli TJ (1995) A model for technical inefficiency effects in a stochastic frontier production function for panel data. Empir Econ 20(2):325–332
Butler J, Moffitt R (1982) A computationally efficient quadrature procedure for the one factor multinomial probit model. Econometrica 50:761–764
Chen Y, Schmidt P, Wang P (2014) Consistent estimates of the fixed effects stochastic frontier model. J Econ 181:65–76
Colombi R (2010) A skew normal stochastic frontier model for panel data. In: Proceedings of the 45-th scientific meeting of the Italian statistical society
Colombi R, Martini G, Vittadini G (2011) A stochastic frontier model with short-run and long-run inefficiency random effects. In: Department of Economics and Technology Management, Universita Di Bergamo, Italy
Colombi R, Kumbhakar S, Martini G, Vittadini G (2014) Closed-skew normality in stochastic frontiers with individual effects and long/short-run efficiency. J Prod Anal (Published online)
Cornes R (1992) Duality and modern economics. CUP Archive
Cornwell C, Schmidt P, Sickles RC (1990) Production frontiers with cross-sectional and time-series variation in efficiency levels. J Econ 46(1–2):185–200
Cuesta RA (2000) A production model with firm-specific temporal variation in technical inefficiency: with application to spanish dairy farms. J Prod Anal 13(2):139–158
Farsi M, Filippini M, Greene W (2005a) Efficiency measurement in network industries: application to the swiss railway companies. J Regul Econ 28(1):69–90
Farsi M, Filippini M, Kuenzle M (2005b) Unobserved heterogeneity in stochastic cost frontier models: an application to Swiss nursing homes. Appl Econ 37(18):2127–2141
Greene W (2004) Distinguishing between heterogeneity and inefficiency: stochastic frontier analysis of the World Health Organization’s panel data on national health care systems. Health Econ 13(10):959–980. doi:10.1002/hec.938
Greene W (2005a) Reconsidering heterogeneity in panel data estimators of the stochastic frontier model. J Econ 126(2):269–303. doi:10.1016/j.jeconom.2004.05.003
Greene W (2005b) Fixed and random effects in stochastic frontier models. J Prod Anal 23(1):7–32
Greene W (2008) The econometric approach to efficiency analysis. In: Fried HO, Lovell CAK, Shelton SS (eds) The measurement of productivity efficiency and productivity growth. Oxford University Press, Oxford, pp 92–250
Jondrow J, Lovell CAK, Materov IS, Schmidt P (1982) On the estimation of technical inefficiency in the stochastic frontier production function model. J Econ 19(2–3):233–238
Kumbhakar SC (1990) Production frontiers, panel data, and time-varying technical inefficiency. J Econ 46(1–2):201–211
Kumbhakar SC, Lovell CAK (2003) Stochastic frontier analysis. Cambridge University Press, Cambridge
Kumbhakar EG, Tsionas SC (2012) Firm heterogeneity, persistent and transient technical inefficiency: a generalized true random effects model. J Appl Econ 29(1):110–132
Kumbhakar SC, Wang H-J (2005) Production frontiers, panel data, and time-varying technical inefficiency. J Econ 46(1):201–211
Kumbhakar SC, Lien G, Hardaker JB (2012) Technical efficiency in competing panel data models: a study of Norwegian grain farming. J Prod Anal 41(2):321–337
Lee Y, Schmidt P (1993) A production frontier model with flexible temporal variation in technical efficiency. In: Fried KLH, Schmidt S (eds) The measurement of productive efficiency. Oxford University Press, Oxford
Mundlak Y (1978) On the pooling of time series and cross section data. Econometrica 46(1):69–85
Parmeter C, Kumbhakar S (2015) Efficiency analysis: a primer on recent advances. Found Trends Econ 7(3–4):1–115
Pitt MM, Lee L-F (1981) The measurement and sources of technical inefficiency in the Indonesian weaving industry. J Dev Econ 9(1):43–64
Schmidt P, Sickles RC (1984) Production frontiers and panel data. J Bus Econ Stat 2(4):367–374
Train K (2003) Discrete choice methods with simulation. Cambridge University Press, Cambridge
Wooldridge JM (2010) Correlated random effects models with unbalanced data. MIMEO, University of Michigan, Ann Arbor
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Filippini, M., Greene, W. Persistent and transient productive inefficiency: a maximum simulated likelihood approach. J Prod Anal 45, 187–196 (2016). https://doi.org/10.1007/s11123-015-0446-y
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11123-015-0446-y