1 / 12

Ratio-Based Efficiency Analysis

Ratio-Based Efficiency Analysis. Antti Punkka and Ahti Salo Systems Analysis Laboratory Aalto University School of Science P.O. Box 11100, 00076 Aalto Finland antti.punkka@tkk.fi, ahti.salo@tkk.fi. Efficiency Ratio and preference statements. Efficiency Ratio of DMU k , k = 1 ,...,K

fionan
Download Presentation

Ratio-Based Efficiency Analysis

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Ratio-Based Efficiency Analysis Antti Punkka and Ahti SaloSystems Analysis LaboratoryAalto University School of ScienceP.O. Box 11100, 00076 AaltoFinlandantti.punkka@tkk.fi, ahti.salo@tkk.fi

  2. Efficiency Ratio and preference statements • Efficiency Ratio of DMUk, k = 1,...,K • Possible preference statements constrain the relative values of outputs and inputs • Linear constraints on output and input weights, cf. assurance regions of type I • ”A doctoral thesis is at least as valuable as 2 master’s theses, but not more valuable than 7 master’s theses”: udoctoral≥ 2umaster’s,udoctoral≤ 7umaster’s, • Feasible weights (u,v) fulfill these linear constraints • Without preference statements, all non-negative u≠0, v≠0 are feasible

  3. Efficiency Ratio in CCR-DEA • Efficient DMUs maximize Efficiency Ratio with some (u,v) • For any (u,v), let E*(u,v) = max {E1(u,v),...,EK(u,v)} • Efficiency score of DMUk is maxu,v [Ek(u,v)/E*(u,v)] • Based on comparisons with one weights, with one DMU • Order of two DMUs’ efficiency scores can depend on what other DMUs are considered • Does not show how ’bad’ a DMU can be • Efficiency score of an efficient DMU is 1 • DMU1 and DMU3are efficient • If DMU5 is included, then DMU2 becomes more efficient than DMU3 in terms of efficiency score E E5 E3 / E*=1 E1 / E*=1 E3 / E*=0.98 E1 E* E2 E3 E4 / E*=0.82 E4 2 outputs,1 input u1

  4. New results for Ratio-Based Efficiency Analysis (REA) • All results are based on comparing DMUs’ Efficiency Ratios 1. Given a pair of DMUs, is the first DMU more efficient than the second for all feasible weights? • Dominance relations 2. What are the best and worst possible efficiency rankings of a DMU overall feasible weights? • Ranking intervals 3. Considering all feasible weights, how efficientis a DMU compared to the most (or the least) efficient DMU of a benchmark group? • Efficiency bounds • Can be computed in presence of preference statements about the relative values of outputs and inputs

  5. Dominance relation (1/2) • DMUk dominates DMUl iff its Efficiency Ratio is (i) at least as high as that of DMUlfor all feasible weights (ii) is higher than that of DMUlfor some feasible weights • Example: 2 outputs, 1 input • Feasible weights such that 2u1 ≥u2≥u1 • DMU3 and DMU2 dominate DMU4 • CCR-DEA-inefficient DMU2 is non-dominated,too • Computation: LP models E E1 E* E2 E3 E4 u1=1/3 u2=2/3 u1=1/2 u2=1/2

  6. Dominance relation (2/2) 1 2 3 3 • A graph shows dominance relations • Transitive: • Asymmetric: no DMU dominates itself and • Additional preference statements can lead to new relations • Relation ”A dominates B” still holds, unless EA = EB throughout the revised weight set • Statement 5u1≥ 4u2 leads to new relations • Dominance vs. CCR-DEA-efficiency • Efficient DMUs are non-dominated • A dominates B⇔B is inefficient among {A,B} • Dominance between two DMUs does not depend on other DMUs 2 4 1 4 E E1 E* E2 E3 E4 5u1= 4u2 u1=1/3 u2=2/3 u1=1/2 u2=1/2

  7. E Ranking intervals E1 • For any (u,v), the DMUs can be ranked based on Efficiency Ratios • DMUs’ minimum and maximum rankings • Properties • Addition / removal of a DMU changes the rankings by at most 1 • Show how ’good’ and ’bad’ DMUs can be • Minimum ranking of a CCR-DEA-efficient DMU is 1 • Computation: MILP models • K-1 binary variables • Additional preference statements do not widen the intervals E* E2 E3 E4 3rd DMU4 ranked 4th u1=1/3 u2=2/3 u1=1/2 u2=1/2 ranking 1 ranking 2 ranking 3 ranking 4 DMU1 DMU2 DMU3 DMU4

  8. E E3 / E* = 1 E1 / E*=1 E1 E2 / E*=0.98 E* E2 E2 / E*=0.85 E3 E E3 / E* = 0.7 E3 / E0 = 1.33 E1 / E*=0.75 E4 / E*=0.82 E4 E4 / E*=0.6 E1 / E0=1.67 E1 E* E2 E2 / E0=1.1 E2 / E0=1.42 E3 E3 / E0 = 1.17 E0 u1=1/3 u2=2/3 u1=1/2 u2=1/2 E1 / E0=1 E4 E4 / E0=1 u1=1/3 u2=2/3 u1=1/2 u2=1/2 Efficiency bounds • Select a benchmark group and compare against its most or least efficient DMU with all feasible weights • ”How efficient is DMU1 compared to the most efficient of other DMUs?” [0.75,1.18] • ”How efficient are the DMUs compared to • ... the most efficient of all DMUs, DMU*? • ... the least efficient of all DMUs, DMU0?” • Computation: LP models • Additional preference statements do not widen the intervals E4 / E0 =1.07 Compared to DMU* E1[0.75,1.00]E* E2[0.85,0.98]E* E3[0.70,1.00]E* E4[0.60,0.82]E* Compared to DMU0 E1[1.00,1.67]E0 E2[1.10,1.42]E0 E3[1.17,1.33]E0 E4[1.00,1.07]E0

  9. Example: Efficiency analysis of TKK’s departments • 2 inputs and 44 outputs describe the 12 university dept’s • Data from TKK’s reporting system • 7 Resources Committee members responded to preference elicitation questions which yielded crisp weightings • E.g. ”How many master’s theses are as valuable as a dissertation?” • Feasible weights modeled as all convex combinations of these 7 weightings y1 (Master’s Theses) Department x1 (Budget funding) y2 (Dissertations) x2 (Project funding) y3 (Int’l publications) TKK = Helsinki University of Technology. As of 1.1.2010, TKK is part of the Aalto University

  10. A J L K I C, E G B D, F, H Ranking intervals Efficiency bounds compared to DMU* • Dept’s A, J and L are CCR-DEA-efficient • But A can attain ranking 7 > 4, the worst ranking of K • For some feasible weights, EA/E* is only 57 % • For K, the smallest such ratio is 71% • Intervals set by Efficiency bounds of D, F and H overlap with those of B and G • Yet, B and G are more efficient for all feasible weights Dominance relations

  11. A J L K I C, E G B D, F, H Specification of performance targets: examples • How big a radial increase in its outputs must Department D make to be among the 6 most efficient departments • ... for some feasible weights? • 25,97 % • ... for all feasible weights? • 54,40 % • Computation: MILP models • How big an increase to be non-dominated? • 88,18% • Computation: LP models

  12. Conclusion • REA results use all feasible weights to compare DMUs • Dominance relations compare DMUs pairwise and provide a dominance structure for the DMUs • Ranking intervals show which efficiency rankings the DMUs can attain • Efficiency bounds extend CCR-DEA-Efficiency scores by allowing comparisons to the most or least efficient unit of any benchmark group • Computation with (MI)LP models allows comparing dozens of DMUs • Consistent with CCR-DEA results; they are obtained as special cases • Admits preference statements • Helps exclude use of extreme weights • More information  narrower intervals, more dominance relations • A. Salo, A. Punkka (2011): Ranking Intervals and Dominance Relations for Ratio-Based Efficiency Analysis, Management Science 57(1), pp. 200-214

More Related