Contents 1. Intro r duction 2. Research Method 3. Applications of DEA t A o Container Po rts 4. Efficiency Results 5. Conclusion

Transcription

1 Contents 1. Introduction 2. Research Method 3. Applications of DEA to Container Ports 4. Efficiency Results 5. Conclusion - 1 -

2 Purpose of Research As the competition among the world ports has become increasingly fierce, every port is striving to increase its investments constantly and lower its operational costs in order to maintain the competitive edge and provide satisfactory services to port users. The unreasoning behavior, however, has induced that substantial waste and inefficiency exists in container port production. It is important to note, however, that pu ure physical expansion is constrained by a limited supply of available land, especially for urban centre ports, and escalating environmental concerns. In addition, the excessive and inappropriate investment also can induce the phenomenon of inefficiency and wasting of resources. Therefore, it is of great importance for the port to improve the productive efficiency of container port appears to be the viable solution and know whether it has fully used its existing infrastructures and that output has been maximized given the input

3 Purpose of Research From this perspective, data envelopment analysis model provides a more appropriate benchmark for the container port. The aim of this study is assumed to be the minimization of the use of input(s) and maximization of the output(s), by applying with DEA-CCR, DEA-BCC, and DEA-Super Efficiency, three models, to acquire a variety of analytical results about the productivity efficiency for the seventeen Korean container ports. According to efficiency value analysis, this study first finds the reason of inefficiency. It is followed by identificationn of the potential areas of improvement for inefficient ports by applying slack variable method. Finally, return to scale approach is used to assess whether each port is in a state of increasing, decreasing, or constant return to scale

4 Data Envelopment Analysis (DEA) Conceptual Exposition The conceptual basis of data envelopment analysis (DEA) came from Farrell, but it was initially restricted to the performance evaluation of decision-making units (DMUs) with multiple inputs and only a single output. Charnes, Cooper and Rhodes later established their model (dubbed the CCR model ), which included multiple inputs and multiple outputs. This model confirmed the utility of DEA in performance evaluation. In analyzing inputs and outputs, the term production frontier is used to refer to the maximum output obtained under different combinations of input variables. In general, it can be distinguished into a parametric production frontier and a non-parametric production frontier; the former requires the production function to be known, whereas the latter does not. DEA utilizes the non-parametric type and applies the concept of Pareto optimization for performance evaluation. Using DEA, it is possible to identify a DMU as an efficient unit or an inefficient unit, and the DEA analysis provides useful guidance regarding how a given DMU s performance might be improved. -4-

5 Data Envelopment Analysis (DEA) Conceptual Exposition By using a simple example involving two inputs (x1, and x2) to produce a single output (y), under the assumption of constant return to scale. Knowledge of the unit isoquant of the fully efficient firm, represented by SS' in Figure, permits the measurement of technical efficiency. If a given firm uses quantities of inputs, defined by the point P, to produce a unit of output, the technical inefficiency of that firm could be represented by the distance QP, which is the amount by which all inputs could be proportionally reduced without a reduction in output. This is usually expressed in percentage terms by the ratio QP/0P. which represents the percentage by which all inputs could be reduced

6 Data Envelopment Analysis (DEA) Slack Variable The slack variable analysis addresses the utilization rate of input and output variables. It does this by assessing how to improve the operational performance of inefficient DMUs by indicating how many inputs to decrease, and/or how many outputs to increase, so as to render the inefficient DMUs efficient. In Figure, the input slack associated with the point A' is CA' of input x2. In cases when there are more inputs and outputs than considered in this simple example, the identification of the "nearest" efficient frontier point (such as C), and hence the subsequent calculation of slacks, is not a trivial task

7 Data Envelopment Analysis (DEA) Conceptual Exposition These efficiency measures assume the production function of the fully efficient firm is known. In practice this is not the case, and the efficient isoquant must be estimated from the sample data. Farrelll suggested the use of either a non- constructed such that no observed parametric piecewise-linear convex isoquant point should lie to the left or below it

8 Data Envelopment Analysis (DEA) DEA-BCC Model Specification DEA has become a popular method for general business management because it has a number of advantages: it can evaluate a DMU s performance with multiple inputs and multiple outputs; it allows the units of input and output variables to be different; and it is not necessary to know the type of production function in advance. However, DEA also has several limitations: the DMUs must be homogeneous; to obtain the best results, the number of DMUs must be at least twice the total number of input and output variables; and isotonicity must exist, that is, the output must not decrease while the input increases

9 Data Envelopment Analysis (DEA) DEA-CCR Model Specification As noted above, Charnes, Cooper and Rhodes established the so-called CCR model to estimate the aggregate efficiency (AE) under constant return to scale(crs). Banker, Charnes and Cooper later established the so-called BCC model to calculate the technical efficiency (TE) under variable return to scale (VRS). By dividing AE by TE, the scale efficiency (SE) is obtained. This allows the identification of an inefficient DMU as being technically inefficient (TIE) or scale inefficient (SIE). The assumptions of the BCC model are essentially the same as those of the CCR model. However, an intercept -u0 is added to identify the kind of return to scale provided by the DMU in question. When the value of -u0 is equal to zero, this indicates that the DMU is under constant return to scale (CRS); less than zero means a decreasing return to scale (DRS); and greater than zero indicates an increasing return to scale (IRS). -9-

10 Data Envelopment Analysis (DEA) DEA-Super-Efficiency Model Specification Because the CCR model gives a value of 1 for all efficient DMUs, it is unable to establish any further distinctions among the efficient DMUs. Andersen and Petersen therefore presented a new DEA model, the Super-efficiency. This model removes an efficient DMU, and then estimates the production frontier again; this provides a new efficiency value for the effic cient DMU that had previously been removed. The new efficiency value can thus be greater than 1. However, if an inefficient DMU is removed, the original production frontier does not change. Therefore, the efficiency values of inefficient DMUs do not change in the Super-efficiency model

11 Data Envelopment Analysis (DEA) Overall, the two basic models of DEA (i.e. CCR and BCC models) are used to provide the efficiency values for port operational performance. The super-efficiency model is used to make further distinctions among the efficient DMUs since they all have efficiency values of 1 in the CCR and BCC models. Thus, the varied and complementary information can be extracted from these three models to provide a more complete and comprehensive performance evaluation. DEA efficiency values of each DMU Scale efficiency > Technical efficiency Need improve ment Inefficient < 1 Scale efficiency < Technical efficiency Efficient = 1 Increasing Scale inefficient Pure technical in efficient Scale efficient and pure technical efficient Efficiency value analysis Decreasing No improvement Slack variable analysis Flow Process of DEA Analyses Constant Return to scale analysis

12 Research Procedure Selecting seventeen Korean ports Eliminating duplication factors Literature research Establishing output /dependent variable Deciding independent variables Survey Collecting data Interview CCR, BCC, Super-efficiency Efficiency value / Slack variable Return to scale Three DEA models Three DEA approaches Conclusions and suggestions Research Procedure

13 Scope of Research Container Terminals

14 Definition of Input Variables Terminal Crane Quay Crane Yard Tractor Water Depth Yard Area Berth Length Output Variable : Throughput per Berth

15 Data Collection of Korean Major Container Ports In 2008 Year Ports Variable Yard area/ Per berth ( m2 ) QC/ Per berth (Number) Busan-KBCT Busan-HBCT Busan-INTERGIS Busan-Hanjin Busan-HGCT Busan-DPCT Busan-UTC Busan-BICT Gwangyang-KIT Gwangyang-KX Gwangyang-HKTL Gwangyang-GICT Gwangyang-DBE Incheon-ICT Incheon-SGCT Ulsan-JUCT Pyeongtaek-PCTC Average TC/ Per berth (Number) Inputs YT/ Per berth (Number) Berth Length (m) Water Depth (m) Output Throughput/ per berth (TEU)

16 Efficiency Under Three DEA Models of Korea Model Efficiency Ports CCR Efficiency Super Efficiency Busan-KBCT Busan-HBCT Busan-INTERGIS Busan-Hanjin Busan-HGCT Busan-DPCT Busan-UTC Busan-BICT Gwangyang-KIT Gwangyang-KX Gwangyang-HKTL Gwangyang-GICT Gwangyang-DBE Incheon-ICT Incheon-SGCT Ulsan-JUCT Pyeongtaek-PCTC Average Reasons of inefficiency Score Rank BCC Scale Return to scale Efficiency Efficiency Decreasing Constant Constant Increasing Constant Increasing Increasing Constant Increasing Decreasing Increasing Increasing Increasing Increasing Increasing Increasing Constant

17 Slack Variable Analysis Results of Korea Variable Inputs Ports Yard area/ per berth QC/ per berth TC/ per berth YT/ per berth Berth Length Water depth Busan-KBCT 100, Busan-HBCT Busan-INTERGIS Busan-Hanjin Busan-HGCT Busan-DPCT Busan-UTC 6, Busan-BICT Gwangyang-KIT Gwangyang-KX Gwangyang-HKTL Gwangyang-GICT Gwangyang-DBE Incheon-ICT Incheon-SGCT Ulsan-JUCT Pyeongtaek-PCTC

18 Case Study for KBCT

19 2.6 기존컨테이너터미널적정하하역능력산정 Case Study for KBCT(Adjusting yard area as slack variable) In the case that yard area is decreased, the throughput per berth is same as previous condition Current Situation 시뮬레이션결과안벽 장치장 항만 터미널명 선석수 선석길 선석 연간 연간총처 이 규모 입항 TGS 리량 (m) (DWT) 척수 (TEU) 연간선석당처리량 (TEU) 점유 대기 율 율 (%) (%) 연간선석당적정량 (TEU) 점유율 (%) 장치장적정처리량 (TEU) 부산북항 신선대 KBCT 5 (300x5) 1,500 50,000 1,848 12,660 2,341, ,353 56% 1.3% 449,296 83% 469,132 19

20 Slack Variable Analysis Results of Korea simulation

21 Empirical results reveal that substantial waste exists in the production process of the container ports in the sample. For instance, the average efficiency of container terminals using the DEA-CCR model amounts to This indicates that, on average, the ports under study can dramatically increase the level of their outputs by 1.42 times as much as their current level while using the same inputs. Empirical results also reveal that the container terminals under the study were found to exhibit a mix of decreasing, increasing and constant returns to scale at current levels of output. Such information is particularly useful for port managers or policy makers to decide on the scale of production

22 Thank You