【正文】
Data Envelopment Analysis (DEA) 數(shù)據(jù)包絡(luò)分析 Over 4000 published articles from 1978 to 2023 DEA ? What it is ? Concepts ? Farrell measures of Efficiency – technical – allocative – scale ? Running DEA ? Dangers of DEA ?Data Envelopment Analysis (DEA) ?This method is a special but important application of linear programming. ?DEA is widely applied to evaluate the performance of Decision Making Units (DMUs) such as private firms, public agencies, schools, and hospitals. Important Features (1) Efficiency (Productivity) Management (2) Relative Measure vs. Absolute Measure (Social Science) (Natural Science) (3) MultiDimensional Analysis vs. Cost / Benefit Analysis ? (1) Relative Comparison ? (2) Multiple Inputs and Outputs ? (3) Efficiency Measurement (0%100%) ? (4) Avoid the Specification Error between Inputs and Outputs ? (5) Production/Cost Analysis Important Features Data Envelopment Analysis (DEA) ? A mathematical optimisation technique where the unknown variables are the weights ? Work through some examples to explain ? Start with a simple example of 8 similar anisations (decision making units DMUs) producing two outputs using a single input S t o r e A B C D E F G HE mp l o y e e 2 3 3 4 5 5 6 8S a l e 1 3 2 3 4 2 3 5S a l e / E mp l o y e e 0 . 5 1 0 . 6 6 7 0 . 7 5 0 . 8 0 . 4 0 . 5 0 . 6 2 5Table 1. Single Input and Single Output Case Concepts Employee 0 1 2 3 4 5 6 7 8 9 6 5 4 3 2 1 0 Figure 1. Comparisons of Branch Stores Efficiency Frontier A B C D F G E H Employee 0 1 2 3 4 5 6 7 8 9 Figure 2. Regression Line vs. Frontier Line Efficiency Frontier A B C D F G E H Regression Line B: the best store The efficiency of the other stores are determined by 1 of em ployeeper Sale s others of em ployeeper Sale s0 ?? B1 = B E D C H A = G F = 6 5 4 3 2 1 0 S t o r e A B C D E F G HE f f i c i e n c y 0 . 5 1 0 . 6 6 7 0 . 7 5 0 . 8 0 . 4 0 . 5 0 . 6 2 5Table 2. Efficiency Employee 0 1 2 3 4 4 3 2 1 0 Figure 3. Improvement of Store A A1 A B A2 S t o r e A B C D E F G H IE mp l o y e e1x 4 7 8 4 2 5 6 5 . 5 6F l o o r A r e a2x 3 3 1 2 4 2 4 2 . 5 2 . 5S a l e y 1 1 1 1 1 1 1 1 1Table 3. Two Inputs and One Output Case Figure 4. Two Inputs and One Output Case Production Possibility Set Efficient Frontiers A E D F C H I G B 0123450 1 2 3 4 5 6 7 8 9Em plo y ee / Sa lesArea/S