Container Terminal Efficiency: A ContextDependent Data Envelopment Analysis Approach

1
Container Terminal Efficiency A
Context-Dependent Data Envelopment Analysis
Approach
Dr. Teng-Fei WANG (speaker) University of
Plymouth, United Kingdom Professor Xinlian
XIE Professor Zan YANG Dalian Maritime
University, China
2
Introduction

• Data Envelopment Analysis (DEA) has widely been
  applied to the container port/terminal industry
• Roll and Hayuth 1993 Martinez-Budria et al,
  1999 Tongzon, 2001 Valentine and Gray, 2001
  Barros and Athanassiou, 2004 Turner, Windle and
  Dresner, 2004 Cullinane, Song and Wang, 2005
  Cullinane, Song, Ji and Wang, 2004 Cullinane, Ji
  and Wang, 2005 Cullinane, Wang, Song Ji, 2006

3
Some (potential) weaknesses of the existing
research

• Do these most frequently DEA approaches always
  make sense in practice?
• Can annual data, which have most frequently been
  used in the existing literature, be a problem?
• Does Chinas container terminal industry deserve
  further research given its rapid development?

4
Data Envelopment Analysis in Theory

• To measure the relative efficiency of a
  homogeneous set of production units termed
  Decision Making Units (DMUs) where the presence
  of multiple inputs and outputs makes comparisons
  difficult.
• DEA utilises the concept of the frontier of the
  production possibility set and constructs a
  piecewise linear frontier to benchmark each
  individual production unit. The efficiency of
  each production unit is no more than unity.

5
In practice what is homogeneous? Is each
container terminal unique?
Sample
Too small
6
In practice what is homogeneous? Is every
container terminal unique?
Sample
Too small
Too big
A risk to lose the homogeneity!!
7
Traditional DEA
a
Output 2
k
d
e
c
f
h
Output 1
8
Context-dependent DEA

• Context-dependent DEA refers to a DEA approach
  where a set of DMUs are evaluated against a
  particular evaluation context.
• Each evaluation context represents an efficient
  frontier composed by DMUs in a specific
  performance level.

9
Can annual data, which have most frequently been
used in the existing literature, be a problem?
10
Top container ports in the world (2005)
11
The locations of the main ports in China
12
Data collection

• There is little that can be measured on a whole
  port basis. Most comparable data must concentrate
  on a terminal basis (Alderton, 1999).
• Two groups of data relating to 32 container
  terminals in China are collected for analysis.
• The first group of data include the annual
  container throughput of these container
  terminals.
• The second group of data include the maximum
  monthly volume of containers handled by different
  terminals.

13
Average efficiency of container terminals in
different regions
north
Central
south
14
Data and efficiency
15
A selection of terminals with highest and lowest
efficiency
16
Different layer of frontier identified by
context-dependent DEA approach
17
Concluding remarks

• This paper demonstrates the potential usefulness
  of context-dependent DEA approach, further
  research should be directed towards the
  implementation of this approaches to practice.
• During this process, a close communication with
  terminal/port practitioners is absolutely
  necessary.