Abstract
This paper introduces new and flexible classes of inefficiency distributions for stochastic frontier models. We consider both generalized gamma distributions and mixtures of generalized gamma distributions. These classes cover many interesting cases and accommodate both positively and negatively skewed composed error distributions. Bayesian methods allow for useful inference with carefully chosen prior distributions. We recommend a two-component mixture model where a sensible amount of structure is imposed through the prior to distinguish the components, which are given an economic interpretation. This setting allows for efficiencies to depend on firm characteristics, through the probability of belonging to either component. Issues of label-switching and separate identification of both the measurement and inefficiency errors are also examined. Inference methods through MCMC with partial centring are outlined and used to analyse both simulated and real data. An illustration using hospital cost data is discussed in some detail.
Similar content being viewed by others
Notes
Predictive results are obtained by integrating out the parameters (of the sampling model) using the posterior. In the case of Fig. 1, this leads to the distribution of efficiencies for an unobserved firm in the industry, given all the information in the data.
Actually, from the point of view of statistical identification, they could appear in both places, although this would perhaps be hard to justify from an economic point of view (with the possible exception of time variables).
Actually, the data do not contain an explicit price for materials, so we assume that material prices are constant across hospitals but not over time. The latter dynamics is then approximated by a quadratic time trend.
This is the ratio of clinical personnel to average daily census, transformed to a dummy variable.
These Bayes factors are, again, computed through the Savage-Dickey density ratio, as explained in Sect. 6.
References
Aigner D, Lovell CAK, Schmidt P (1977) Formulation and estimation of stochastic frontier production function models. J Econometrics 6:21–37
Albert JH, Chib S (1993) Bayesian analysis of binary and polychotomous response data. J Am Stat Assoc 88:669–679
Battese GE, Coelli TJ (1992) Frontier production functions, technical efficiency and panel data: with application to paddy farmers in India. J Prod Anal 3:153–169
Bernardo JM, Smith AFM (1994) Bayesian theory. Wiley, Chichester
Box GEP, Tiao GC (1973) Bayesian inference in statistical analysis. Addison-Wesley, Reading
Carree MA (2002) Technological inefficiency and the skewness of the error component in stochastic frontier analysis. Econ Lett 77:101–107
Celeux G, Hurn M, Robert CP (2000) Computational and inferential difficulties with mixture posterior distributions. J Am Stat Assoc 95:957–970
Cornwell C, Schmidt P, Sickles R (1990) Production frontiers with cross-sectional and time-series variation in efficiency levels. J Econometrics 46:185–200
Fernández C, Koop G, Steel MFJ (2002) Multiple-output production with undesirable outputs: an application to nitrogen surplus in agriculture. J Am Stat Assoc 97:432–442
Fernández C, Osiewalski J, Steel MFJ (1997) On the use of panel data in stochastic frontier models with improper priors. J Econometrics 79:169–193
Gelfand AE, Sahu SK, Carlin BP (1995) Efficient parameterizations for normal linear mixed models. Biometrika 82:479–488
Gilks WR, Wild P (1992) Adaptive rejection sampling for Gibbs sampling. J R Stat Soc C 34:198–200
Greene WH (1990) A gamma-distributed stochastic frontier model. J Econometrics 46:141–163
Greene WH (2003) Simulated likelihood estimation of the normal-gamma stochastic frontier function. J Prod Anal 19:179–190
Griffin JE, Steel MFJ (2004) Semiparametric Bayesian inference for stochastic frontier models. J Econometrics 123:121–152
Hall P, Simar L (2002) Estimating a changepoint, boundary or frontier in the presence of observation error. J Am Stat Assoc 97:523–534
Johnson NL, Kotz S, Balakrishnan N (1994) Continuous univariate distributions. vol 1, 2nd edn. Wiley, New York
Koop G, Osiewalski J, Steel MFJ (1997) Bayesian efficiency analysis through individual effects: Hospital cost frontiers. J Econometrics 76:77–105
Koop G, Steel MFJ, Osiewalski J (1995) Posterior analysis of stochastic frontiers models using Gibbs sampling. Computational Stat 10:353–373
Kozumi H, Zhang XY (2005) Bayesian and non-Bayesian analysis of gamma stochastic frontier models by Markov chain Monte Carlo methods. Comput Stat 20:575–593
Kumbhakar SC (1990) Production frontiers, panel data and time-varying technical inefficiency. J Econometrics 46:201–212
Kumbhakar SC, Ghosh S, McGuckin JT (1991) A generalized production frontier approach for estimating determinants of inefficiency in US dairy farms. J Bus Econ Stat 9:279–286
Meeusen W, van den Broeck J (1977) Efficiency estimation from Cobb-Douglas production functions with composed errors. Int Econ Rev 8:435–444
Papaspiliopoulos O, Roberts GO, Sköld M (2003) Non-centered parameterisations for hierarchical models and data augmentation. Bernardo JM, Bayarri MJ, Berger JO, Dawid AP, Heckerman D, Smith AFM, West M (eds) Bayesian statistics, vol 7. Clarendon Press, Oxford, pp 307–326
Reinhard S, Lovell CAK, Thijssen G (1999) Econometric application of technical and environmental efficiency: an application to dutch dairy farms. Am J Agr Econ 81:44–60
Richardson S, Green PJ (1997) On Bayesian analysis of mixtures with an unknown number of components (with discussion). J Roy Stat Soc B 59:731–792
Ritter C, Simar L (1997) Pitfalls of Normal-Gamma stochastic frontier models. J Prod Anal 8:167–182
Stacy EW (1962) A generalization of the gamma distribution. Ann Math Stat 33:1187–1192
Stephens M (2000) Dealing with label switching in mixture models. J R Stat Soc B 62:795–809
Stevenson RE (1980) Likelihood functions for generalized stochastic frontier estimation. J Econometrics 13:57–66
Spiegelhalter D, Grigg O, Kinsman R, Threasure T (2003) Risk-adjusted sequential probability ratio tests: applications to Bristol, Shipman and adult cardiac surgery. Int J Qual Health Care 15:7–13
Tsionas EG (2000) Full likelihood inference in Normal-gamma stochastic frontier models. J Productivity Anal 13:183–205
van den Broeck J, Koop G, Osiewalski J, Steel MFJ (1994) Stochastic frontier models: a Bayesian perspective. J Econometrics 61:273–303
Verdinelli I, Wasserman L (1995) Computing Bayes factors using a generalization of the Savage-Dickey density ratio. J Am Stat Assoc 90:614–618
Acknowledgements
Jim Griffin acknowledges research support from The Nuffield Foundation grant NUF-NAL/00728. We thank three Referees for constructive comments.
Author information
Authors and Affiliations
Corresponding author
Appendix: Some Details on the MCMC Sampler
Appendix: Some Details on the MCMC Sampler
1.1 A.1 Drawing the inefficiencies
The full conditional for the inefficiencies is, of course, different from the one in the existing literature. The firm-specific inefficiencies u i will be independent given all the other parameters and the observations, with density function:
where \(\mu_i=T_i\alpha+\beta^{\prime} X_i^{\prime} \iota-y_i^{\prime} \iota\) for a cost frontier and its negative for a production frontier, if we define \(y_i=(y_{i1},{\ldots},y_{i{T_i}})^\prime,\;X_i=(x_{i1},{\ldots},x_{i{T_i}})^\prime\) and \(\iota\) is a T i -dimensional vector of ones. We can easily show that the conditional in (10) is log-concave if
which is always satisfied if cϕ > 1 and c > 1. For this parameter combination, we use adaptive rejection sampling (see Gilks and Wild 1992) to sample directly from (10). In the other cases, we use random walk Metropolis-Hastings with a lognormal candidate generator. Random walk Metropolis-Hastings is also used to generate drawings for the parameters ϕ and c while λ can simply be drawn from a Gamma conditional. We use the reparameterization from (ϕ,c) to (ψ,c), where ψ = ϕc, since it leads to better mixing properties of the sampler.
1.2 A.2 Centring
As indicated in Sect. 6, we use a sampler which randomly mixes updates from the centred and the uncentred parameterizations. This basic idea is called “partial centring” in Papaspiliopoulos et al. (2003), who show that this generally leads to more robust sampling algorithms. We choose a centred update with probability 1/4, as we only need to “recentre” α once in a while. In addition, we found that centring works best if we integrate out λ while updating α (i.e. we effectively draw α and λ jointly). Thus, in the centred parameterization we use a random walk Metropolis-Hastings step to sample from the following conditional for α < min i {z i }:
which can be shown to be log-concave whenever cϕ > 1 and c > 1. We use adaptive rejection sampling in the latter case and a random walk Metropolis-Hastings step otherwise.
1.3 A.3 The mixture model
In the case of the mixture inefficiency distribution, as described in Sect. 3.1, we extend the MCMC sampler of the basic case by augmenting with an indicator variable s i ,i = 1,…,n which can take the values 0 or 1 and assigns firm i to one of the two efficiency groups (i.e. one of the two inefficiency components). For the mixture model the sampler will, thus, generate a chain on (α,β,σ2,w,θ,u,s), where s = (s 1,…,s n ). Inference about the parameters of each component (c j , ϕ j , λ j ), j = 1,2 now only depends on those firms for which s i = j − 1. The full conditional distribution for w will be
and the values of s i are updated through
For the decomposition case, where we have included a prior probability of w = 0, the full conditional distribution of w has the same form as above except when \(\sum_{i=1}^n s_i=n\). In this case w is zero with probability \(q^{\star}\) and w∼Be(w 0, w 1 + n) otherwise where
Here we only considered switching to the one-component model when \(\sum_{i=1}^n s_i=n\), which worked well. However, in applications with very large n it might be more efficient to also consider jumps between the models when \(\sum_{i=1}^n s_i\) is relatively close to n.
In case we allow for the probability of being in the efficient group to depend on firm characteristics, we can use simple Gibbs sampling after data augmentation as in Albert and Chib (1993) to update the probit regression coefficients.
Rights and permissions
About this article
Cite this article
Griffin, J.E., Steel, M.F.J. Flexible mixture modelling of stochastic frontiers. J Prod Anal 29, 33–50 (2008). https://doi.org/10.1007/s11123-007-0064-4
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11123-007-0064-4