Abstract
Standard axioms of free disposability, convexity and constant returns to scale employed in Data Envelopment Analysis (DEA) implicitly assume continuous, real-valued inputs and outputs. However, the implicit assumption of continuous data will never hold with exact precision in real world data. To address the discrete nature of data explicitly, various formulations of Integer DEA (IDEA) have been suggested. Unfortunately, the axiomatic foundations and the correct mathematical formulation of IDEA technology has caused considerable confusion in the literature. This chapter has three objectives. First, we re-examine the axiomatic foundations of IDEA, demonstrating that some IDEA formulations proposed in the literature fail to satisfy the axioms of free disposability of continuous inputs and outputs, and natural disposability of discrete inputs and outputs. Second, we critically examine alternative efficiency metrics available for IDEA. We complement the IDEA formulations for the radial input measure with the radial output measure and the directional distance function. We then critically discuss the additive efficiency metrics, demonstrating that the optimal slacks are not necessarily unique. Third, we consider estimation of the IDEA technology under stochastic noise, modeling inefficiency and noise as Poisson distributed random variables.
Abbreviations of key concepts referred to in this chapter: DEA = Data Envelopment Analysis, DMU = Decision Making Unit, CNLS = Convex Nonparametric Least Squares, IDEA = Integer DEA, MILP = Mixed Integer Linear Programming, RTS = Returns To Scale, SFA = Stochastic Frontier Analysis, StoNED = Stochastic Nonparametric Envelopment of Data.
Abbreviations of articles frequently cited in this chapter: KJS = Kuosmanen, Johnson and Saastamoinen (in this volume), KKM = Kuosmanen and Kazemi Matin (2009), KMK = Kazemi Matin and Kuosmanen (2009), KSM = Khezrimotlagh, Salleh, and Mohsenpour (2012, (2013a, (2013b), LV = Lozano and Villa (2006, 2007).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
We will henceforth use the term “DMU” to refer to any entity that transforms inputs to output, including both non-profit firms and for-profit companies. DMU can refer to a production plant, facility, or sub-division of a company, or to an aggregate entity such as an industry, a region, or a country.
- 2.
Previous studies such as Banker and Morey (1986), Kamakura (1988), and Rousseau and Semple (1993) (among others) consider inputs and outputs measured on the categorical or ordinal scale, which are obviously integer valued. However, input-output variables defined on the interval or ratio scales can be integer valued as well.
- 3.
- 4.
For clarity, we denote vectors by bold lower case letters (e.g., x) and matrices by bold capital letters (e.g., X).
- 5.
- 6.
This is another issue that has caused confusion: see Cherchye et al. (2001).
- 7.
KSM (2012) state: ”Now, if it has been supposed that only the integer numbers set is considered, then it should not have been used the real number variable in the integer axioms! In fact, a new axiom must not have any doubts or parallel affects with those previous axioms. In other words, an axiom is an evident premise as to be accepted as true without controversy.” This discussion reveals that KSM do not understand the economic meaning of axioms in DEA. In fact, none of the standard DEA axioms can meet the requirements of KSM.
- 8.
DEA is a nonparametric estimator subject to the curse of dimensionality. This implies that the precision of DEA estimator deteriorates rapidly as the number of input and output variables increases. Also the discriminating power of DEA is affected: when the dimensionality is large, almost all DMUs appear as inefficient.
- 9.
In the single output case, Afriat (1972) proves the similar minimum extrapolation result for the smallest production function satisfies axioms (A1), (A1) + (A2), or (A1) + (A2) + (A5).
- 10.
In their original manuscript, KKM stated their MILP formulation using inequality constraints. They later introduced slacks by request of a reviewer.
- 11.
In most applications we can think of, it would seem more natural to treat integer-valued inputs as quasi-fixed factors, and project DMUs to the frontier in the direction of continuous inputs.
- 12.
An interested reader can easily verify the empirical results reported by KKM and KMK. For transparency, the data and the computational codes for GAMS and LINGO are freely available on the website:
- 13.
The Poisson distribution is the most widely used discrete probability distribution in statistics. It can be derived from the probability of a given number of events occurring in a fixed interval of time and/or space when the events occur with a known average rate and independently of the time since the last event. Note that the Poisson distribution can be derived as a limiting case to the binomial distribution as the number of trials approaches to infinity and the expected number of successes is fixed.
- 14.
- 15.
See, e.g., Simar and Wilson (2010) for a more detailed discussion about the wrong skewness problem in stochastic frontier estimation.
- 16.
KJS, Sect. 7.4.8, discusses some of this literature in the context of StoNED estimation.
References
Afriat SN (1972) Efficiency estimation of production functions. Int Econ Rev 13(3):568–598
Alirezaee MR, Sani MRR (2011) An enumeration algorithm for integer-valued data envelopment analysis. Int Trans Oper Res 18(6):729–740
Banker RD, Morey RC (1986) The use of categorical variables in data envelopment analysis. Manage Sci 32(12):1613–1627
Banker RD, Charnes A, Cooper WW, (1984) Some models for estimating technical and scale inefficiencies in data envelopment analysis. Manage Sci 30(9):1078–1092
Bogetoft P (1996) DEA on relaxed convexity assumptions. Manage Sci 42(3):457–465
Bogetoft P, Tama JM, Tind J (2000) Convex input and output projections of nonconvex production possibility sets. Manage Sci 46(6):858–869
Chambers RG, Chung YH, Färe R (1996) Benefit and distance functions. J Econ Theory 70(2):407–419
Chambers RG, Chung YH, Färe R (1998) Profit, directional distance function, and Nerlovian efficiency. J Optim Theory Appl 98(2):351–364
Charnes A, Cooper WW, Rhodes E (1978) Measuring the efficiency of decision making units. Eur J Oper Res 2(6):429–444
Charnes A, Cooper WW, Golany B, Seiford LM, Stutz J (1985) Foundations of data envelopment analysis for Pareto-Koopmans efficient empirical production functions. J Econom 30(1):91–107
Cherchye L, Kuosmanen T, Post T (2001). Alternative treatments of congestion in DEA. Eur J Oper Res 132:75–80
Chen CM., Du J, Huo J, Zhu J (2012) Undesirable factors in integer-valued DEA: evaluating the operational efficiencies of city bus systems considering safety records. Decis Support Syst 54(1):330–335
Chen Y, Djamasbi S, Du J, Lim S (2013) Integer-valued DEA super-efficiency based on directional distance function with an application of evaluating mood and its impact on performance. Int J Prod Econ 146(2):550–556
Cooper WW, Park KS, Pastor JT (1999) RAM: a range adjusted measure of inefficiency for use with additive models, and relations to other models and measures in DEA. J Prod Anal 11:5–42
Deprins D, Simar L, Tulkens H, (1984) Measuring labor-efficiency in post offices. In: Marchand M, Pestieau P, Tulkens H, (eds) The performance of public enterprises, concepts and measurement. Elsevier Science Ltd., North Holland
Du J, Chen CM, Chen Y, Cook WD., Zhu J (2012) Additive super-efficiency in integer-valued data envelopment analysis. Eur J Oper Res 218(1):186–192
Farrell M (1957) The measurement of productive efficiency. J Royal Stat Soc Ser A 120(3):253–290
Färe R, Grosskopf S (2009) A Comment on weak disposability in nonparametric production analysis. Am J Agric Econ 91(2):535–538
Jondrow J, Lovell CAK, Materov IS, Schmidt P (1982) On estimation of technical inefficiency in the stochastic frontier production function model. J Econom 19:233–238
Kamakura WA (1988) A note on the use of categorical variables in data envelopment analysis. Manage sci 34(10):1273–1276
Kazemi Matin, R., Emrouznejad, A. 2011. An integer-valued data envelopment analysis model with bounded outputs. Int Trans Oper Res 18(6):741–749
Kazemi Matin R, Kuosmanen T (2009) Theory of integer-valued data envelopment analysis under alternative returns to scale axioms. Omega 37(5):988–995
Keshvari A Kuosmanen T (2013) Stochastic non-convex envelopment of data: applying isotonic regression to frontier estimation. Eur J Oper Res 231(2):481–491
Khezrimotlagh D, Salleh S, Mohsenpour Z (2012) A comment on theory of integer-valued data envelopment analysis. Appl Math Sci 6(116):5769–5774
Khezrimotlagh D, Salleh S, Mohsenpour Z (2013a) A new robust mixed integer-valued model in DEA. Appl Math Model 37(24):9885–9897
Khezrimotlagh D, Salleh S, Mohsenpour Z (2013b) A note on integer-valued radial model in DEA. Comput Ind Eng 66(1):199–200
Koopmans TC (1951a) Analysis of production as an efficient combination of activities. In: Koopmans TC (ed) Activity analysis of production and allocation. Wiley, New York, pp 33–97
Koopmans TC (ed) (1951b) Activity analysis of production and allocation, cowles commission for research in economics. Wiley, New York
Kuosmanen T (2001) DEA with efficiency classification preserving conditional convexity. Eur J Oper Res 132(2):326–342
Kuosmanen T (2003) Duality theory of non-convex technologies. J Prod Anal 20:273–304
Kuosmanen T (2005) Weak disposability in nonparametric production analysis with undesirable outputs. Am J Agric Econ 87:1077–1082
Kuosmanen T (2008) Representation theorem for convex nonparametric least squares. Econom J 11:308–325
Kuosmanen T, Johnson AL (2010) Data envelopment analysis as nonparametric least squares regression. Oper Res 58(1):149–160
Kuosmanen T, Kazemi Matin R (2009) Theory of integer-valued data envelopment analysis. Eur J Oper Res 192(2):658–667
Kuosmanen T, Kortelainen M (2012) Stochastic non-smooth envelopment of data: semi-parametric frontier estimation subject to shape constraints. J Prod Anal 38(1):11–28
Kuosmanen T, Podinovski VV (2009) Weak disposability in nonparametric production analysis: reply to Färe and Grosskopf. Am J Agric Econ 91(2):539–545
Kuosmanen T, Post GT, Sipiläinen T (2004) Shadow price approach to total factor productivity measurement: with an application to Finnish grass-silage production. J Prod Anal 22(1):95–121
Kuosmanen T, Johnson AL, Saastamoinen A (2014) Stochastic nonparametric approach to efficiency analysis: a unified framework. In: Zhu J (ed) data envelopment analysis, Springer
Lozano S (2013) Using DEA to find the best partner for a horizontal cooperation. Comput Ind Eng 66(2):286–292
Lozano S, Villa G (2006) Data envelopment analysis of integer-valued inputs and outputs. Comput Oper Res 33(10):3004–3014
Lozano S, Villa G (2007) Integer DEA models. In Zhu J, Cook WD (eds) Modeling data irregularities and structural complexities in data envelopment analysis. Springer, New York, pp 271–290
Lozano S, Villa G, Canca D (2011) Application of centralised DEA approach to capital budgeting in Spanish ports. Comput Ind Eng 60(3):455–465
Nöhren M, Heinzl A (2012) Measuring the relative efficiency of global delivery models in IT outsourcing. Lect Notes Bus Inf Process 130:61–75
Petersen, N. C. (1990) Data Envelopment analysis on a relaxed set of assumptions, Management Science 20(3), 305–314
Podinovski VV (2004) Bridging the gap between the constant and variable returns-to-scale models: selective proportionality in data envelopment analysis. J Oper Res Soc 55(3):265–276
Post GT (2001) Transconcave data envelopment analysis. Eur J Oper Res 132(2):374–389
Rousseau JJ, Semple JH (1993) Notes: categorical outputs data envelopment analysis. Manage Sci 39(3):384–386
Shephard R, (1970) Theory of cost and production functions. Princeton University Press, Princeton
Simar L, Wilson PW (2010) Inferences from cross-sectional, stochastic frontier models. Econom Rev 29(1):62–98
Skellam JG (1946) The frequency distribution of the difference between two Poisson variates belonging to different populations. J Royal Stat Soc Ser A 109(3):296–296
Tone K (2001) A slacks-based measure of efficiency in data envelopment analysis. Eur J Oper Res 130:498–509
Tulkens H (1993) On FDH efficiency analysis: some methodological issues and applications to retail banking, courts, and urban transit. J Prod Anal 4:183–210
Von Neumann J (1945–1946) A model of general economic equilibrium. Rev Econ Stud 13(1):1–9
Wu J Zhou Z, Liang L (2009) Measuring the performance of nations at Beijing summer olympics using integer-valued DEA model. J Sports Econ 11(5):549–566
Wu J, Liang L, Song H (2010) Measuring hotel performance using the integer DEA model. Tour Econ 16(4):867–882
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer Science+Business Media New York
About this chapter
Cite this chapter
Kuosmanen, T., Keshvari, A., Matin, R. (2015). Discrete and Integer Valued Inputs and Outputs in Data Envelopment Analysis. In: Zhu, J. (eds) Data Envelopment Analysis. International Series in Operations Research & Management Science, vol 221. Springer, Boston, MA. https://doi.org/10.1007/978-1-4899-7553-9_4
Download citation
DOI: https://doi.org/10.1007/978-1-4899-7553-9_4
Published:
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4899-7552-2
Online ISBN: 978-1-4899-7553-9
eBook Packages: Business and EconomicsBusiness and Management (R0)