Scholarly article on topic 'A Study on the Efficiency of Container Terminals in Korea and China'

A Study on the Efficiency of Container Terminals in Korea and China Academic research paper on "Earth and related environmental sciences"

CC BY-NC-ND
0
0
Share paper
Keywords
{"Container Terminal" / "Efficiency Evaluation" / DEA-CCR / DEA-BCC}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — Xue Bin Zheng, Nam Kyu Park

ABSTRACT The objective of this study is to derive implications required for efficiency improvement and management level enhancement by selecting container terminals within major large ports of Korea and China as comparison units, evaluating their relative efficiencies and analyzing the trend of changes in their efficiencies. Since the scope of comparison subjects has been narrowed down to container terminals unlike previous studies, it is expected that the study results would have significant meaning due to the fact that it would be possible to compare and analyze in more detail. To achieve the objective, 30 major container terminals in both countries are selected, input and output variables are defined for each terminal and the DEA (data envelopment analysis) model is used to conduct an analysis. The results show that the efficiency of major terminals in Korea (CCR: 0.815, BCC: 0.886) showed similar efficiency with China's terminals (CCR: 0.817, BCC: 0.887). While previous studies conclude that the efficiency of ports in Korea is far lower than that of ports in China.

Academic research paper on topic "A Study on the Efficiency of Container Terminals in Korea and China"

HOSTED BY

ELSEVIER

Contents lists available at ScienceDirect

The Asian Journal of Shipping and Logistics

Journal homepage: www.elsevier.com/locate/ajsl

Slit Abúiu Journal at O gapping aiift CiniitiItíü

A Study on the Efficiency of Container Terminals in Korea and China *

CrossMark

Xue Bin ZHENGa, Nam Kyu PARKb

a Doctor Student, Korea Maritime and Ocean University, Korea, E-mail:jhbl025@kmou.ac.kr (First .Author) b Professor, Tongmyong University, Korea, E-mail:nkpark@tu.ac.hr (Corresponding Author)

ARTICLE INFO

ABSTRACT

Article history: Received 2 June 2016

Received in revised form 25 November 2016 Accepted 30 November 2016

Keywords: Container Terminal Efficiency Evaluation

DEA-CCR DEA-BCC

The objective of this study is to derive implications required for efficiency improvement and management level enhancement by selecting container terminals within major large ports of Korea and China as comparison units, evaluating their relative efficiencies and analyzing the trend of changes in their efficiencies. Since the scope of comparison subjects has been narrowed down to container terminals unlike previous studies, it is expected that the study results would have significant meaning due to the fact that it would be possible to compare and analyze in more detail. To achieve the objective, 30 major container terminals in both countries are selected, input and output variables are defined for each terminal and the DEA (data envelopment analysis) model is used to conduct an analysis. The results show that the efficiency of major terminals in Korea (CCR: 0.815, BCC: 0.886) showed similar efficiency with China's terminals (CCR: 0.817, BCC: 0.887). While previous studies conclude that the efficiency of ports in Korea is far lower than that of ports in China.

Copyright © 2016 The Korean Association of Shipping and Logistics, Inc. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

1. Introduction

In recent times, the North East Asian region has globally large container ports where traffic volume is concentrated and generated the most in the world. The most effective means to enhance the competitiveness of a container terminal is to improve its service level, which can be realized by fully utilizing invested resources such as docks, berths, yards and equipment in efficiency improvement of the terminal. Multiple container terminals in many large container ports in Korea and

China are independently operated under the control of a single port authority. For example, many container terminals in Shanghai Port are operated by themselves respectively under the control of Shanghai International Port Group (SIPG). Similarly, there are several container terminals in Busan Port operated by themselves separately under the control of Busan Port Authority (BPA). However, previous studies on port efficiency evaluation of different countries are merely focusing on their

r This research was supported by the Tongmyong University Research Grant 2016.

http://dx.doi.org/10.1016/j.ajsl.2016.12.004

2092-5212/© 2016 The Korean Association of Shipping and Logistics, Inc. Production and hosting by Elsevier B.V. Peer review under responsibility ofthe Korean Association of Shipping and Logistics, Inc.

ports. They have not touched comparison on the container terminal operation between two or more countries. In this article, after selecting container terminals within major ports in Korea and China, we try to compare and analyse their relative efficiency. It is intended to find problems presented in the terminal operation and to derive implications required for efficiency improvement and operational level enhancement.

For subjects of this study, major terminals are selected within the global top 10 container ports as in Korea and China.

The paper consists of 5 sections. Section 2 deals with literature review, section 3 is about the DEA model review, which allows calculating the relativity efficiency of the container terminals in section 4. Finally, conclusions are drawn in section 5.

1. Literature Review

1.1. Port Efficiency Evaluation Using DEA

There are a large number of researches concerning port efficiency evaluation beginning in the 1970s, using various methodologies. Recently, the DEA model is commonly used to examine the efficiency of ports and has been the subject of many research studies estimating port efficiency. Table 1 summarizes previous studies on DEA and port efficiency measurements.

Table 1

Previous Study Summary

Researcher DMUs Input Factors Output Factors Model

Roll and Hayuth(1993) Israeli ports (20) -labor cost -capital cost -cargo characteristics -total cargo volume -service level -customer satisfaction -Number of vessels DEA-CCR

Martinez- Budria et al (1999) Spanish ports (26) -depreciation cost -labor cost -other costs -port facilities -lease income -cargo throughput DEA-BCC

Tongzon (2001) Australian container ports (4) & other countries' container ports (12) -number of cranes -number of berths -number of tugs -terminal area -vessel waiting time -number of port employees -container throughput -Number of disposed containers per work hour DEA-CCR DEA-Additive

Itoh (2002) Japanese ports (8) -terminal area -number of berths -number of cranes -number of employees -container throughput DEA-CCR DEA-BCC

Barrors (2003b) Portuguese ports (10) -number of employees -asset book value -Number of vessels -cargo throughput -general cargo throughput -container throughput -bulk cargo throughput -liquid cargo throughput DEA-Malmquist Tobit

Turner et al (2004) North American ports (26) -berth length -terminal area -number of cranes container throughput DEA-CCR Tobit

Cullinane et al (2004) global top 25 container ports -berth length -terminal area -Number of QCs -number of transfer cranes -number of straddle carriers container throughput DEA-CCR DEA-BCC

Lee et al (2007) ports in Korea, China and Japan (16) -number of berths -berth length -the depth of water -number of cranes -terminal area container throughput DEA-CCR DEA-BCC

Bang et al (2011) major ports in the world (76) -berth length -the average depth of water -number of QCs -terminal area container throughput DEA-CCR DEA-BCC

Kim et al (2012) Korean and Chinese ports(10) -number of berths -berth length -terminal area -number of QCs container throughput DEA-CCR DEA-BCC

Li et al (2015) 16 ports (Northeast Asia) -number of Berths -berth Length -depth -total area -QC Number container throughput DEA-CCR DEA-BCC

Source: Author

1.2. Differentiation to previous Studies

Mostly, previous studies have the following limitations:

First, most of the previous studies evaluated efficiency based on the port entity or as not strictly differentiating ports and terminals. In fact, relatively small number of studies compared container terminals among countries or regions.

Second, the number of selected DMUs: like researchers of Tongzon (2001), Itoh (2002), and Kim et al (2012) are too small in some studies and some studies lacked correctness or objectivity in the number of input and output factors and their correlations.

Third, previous studies compared container ports in Korea and China specified input factors with facilities and equipment of the entire port (total facilities and equipment of each terminal) and conducted the comparison. However, some of the reference data such as Containerization International Yearbook, are far different from the actual status of facilities and equipment within terminals and ports in China.

Accordingly, aspects, as follows, are improved in this study to derive more objective study results.

First, the appropriate number of homogeneous DMUs are selected and factors are selected that reflect the actual relationship between input and output factors to enhance the accuracy of the efficiency analysis results.

Second, the scope of comparison subjects is narrowed down to container terminals in Korea and China to measure the terminal operation efficiency in detail so as to derive implications more significant to operators.

Third, data is collected from statistics published by related organizations, official websites of ports, data published by ports, official websites of terminal operators for the ports, actual site visits, and it is comprehensively reviewed to ensure objectivity and accuracy of input and output factors to the extent possible.

2. Methodology

2.1. Summary of DEA

The DEA model can be divided into several types depending on the nature of the applied problem and characteristics of given data. Typical basic models widely used for the DEA model are four of the constant return to scale (CRS) based input and output oriented CCR models and variable return to scale (VRS) input and output oriented BCC models. The output oriented CCR and BCC models are selected for the analysis in this study.

2.2. DEA-CCR Model

DEA-CCR model is the first DEA based model proposed by Charnes, Cooper and Rhoders (1978), and it is characterized in input or output criteria based on the constant return to scale (CRS). The CCR model derives the efficiency by estimating the ratio of input and output variables of each DMU. Its limiting condition is that the input and output ratios should be equal to or less than a single constant value. For 'n' DMUs, if 'm' inputs xig (i=l,2,3,...,m)>0) are used to calculate outputs y^.(r=l,2,3,...,s)>0 from DMUj(j=l,2,3...n), the linear programming form of the CCR model is Equation 1 as follows:

s. t. ^ Uj-yj-j — ^ Vi%f <0, j= 1,2,..., n

r=1 i=l

V:Xin = 1

Ur, Vi > s, Vr,i

Where, Xf is the amount of the 'i' types of input factors for the j-th DMU, yrj is the amount of the 'r' types of output factors for the 'j'-th DMU, ur is the weight of the 'r'-th output factor, vt is the weight of its input factor, e is the non-Archimedean constant which is very small (10~6 in general), and h0 is the relativity efficiency of the 'o'-th DMU.

Then, when slack variables and s+ are introduced, it can be presented in the linear programming model as shown in Equation 2 as follows:

min h„ = 9

i = 1,2, ...,m

, J A,

^ AiVrj - Sr = yr0. r = 1,2,... ,s

s. t. ^ Aixs + s; = ®xi i=i

sr,sr+,Aj > 0, Vi,r,j

If the optimal solution of the above linear programming equation is 1, the DMUy0 is technically effective, and if the optimal solution is 1 and the optimal solution is 9* = 1, sf* = 0, s^* = 0 at the same time, it is thought the DMUy0 is effective.

Equation 1 and Equation 2 are the different form of the input oriented CCR model. The output oriented CCR linear programming model after duality transformation is Equation 3 as follows:

max h„ = 9

s.t. - ^ Âj yrJ + ^ 9yr0 +

j=1 r=1

^ Aj Xf + s[ = xiB, i = 1,2, ...,m

sJ" = 0, r= 1,2,..., s

sr,sr+,Aj> 0, Vi,r,j

2.3. DEA-BCC model

Since the CCR model assumes returns to scale are constant, it has a drawback that the scale efficiency and the pure technical efficiency are not differentiated. If the production technology is based on variable returns to scale, the estimate derived from the CCR model may appear inefficient despite actually efficient DMU. To improve on such a problem, Banker, Charnes, and Cooper (1984) eased the constant returns to scale to apply the assumption called as the variable returns to scale and added block requisites. The BCC linear programming model becomes Equation 4 as follows:

max ha = uryr0 + ua

s.t. ^ uryrj — ^ ViXj + u0 < 0, j = 1,2, ...,n

r=1 [=1

V,X,a = 1

max h0 = Y,r=iUryro

ur,xi0 > e, Vr, i

Where, since u0 is not subject to condition limitation, it reflects the return to a scale characteristic of the 'j'-th DMU. Ifu0 = 0 , the DMU will be under the most optimal production scale, if under the constant return to scale condition and u0 > 0 , the DMU will be under the decreasing returns to scale since it is under the production scale greater than optimal, and if u0 < 0, the DMU will be under the increasing returns to scale since it is under the production scale smaller than optimal.

Then, when slack variables and s+ are introduced, it can be presented in the linear programming model as shown in Equation 5 as follows:

min h0 = 8 (5)

s.t. ^^AjXj + s[ = 0xiB, i = l,2,...,m i=i

^¿jyrj - Sr = yr0. r = l,2, ...,s

sf.s+.Aj > 0, V i,r,j

Equation 4 and Equation 5 are the different form of input oriented BCC model. The output oriented BCC linear programming model after duality transformation is Equation 6 as follows:

max h0 = 6 + €K™! sr + Zr=i sr+] (6)

S.t. — X Aj yrj + X flyro + 5r = r=l,2,...,s

~ xio>

sf.s;,^. > o, V i,r,j

2.4. Scale Efficiency

The constant return to scale refers to the case that output factors increase as the same as levels of input factors, the increasing return to scale refers to the case that output factors increase beyond levels of input factors, and the decreasing return to scale represent that output factors increase under levels of input factors. The scale efficiency (SE) is indicated in Equation 7 as follows:

Scale Effciency = Technical Effciency ^ Pure Technical Effciency (7)

Since the CCR efficiency index does not consider the scale effect, it is also called as the technical efficiency (TE). Meanwhile, the BCC efficiency index represents the local pure technical efficiency (PTE) under the assumption of the variable returns to scale, and it is indicated in Equation 8 as follows:

Technical Effciency = Pure Technical Effciency x Scale Effciency (8)

Such decompositions show whether the inefficiency source is due to inefficient management activities or caused by inefficiency. If the scale efficiency of a DMU is 1, it means it is operated in the very productive scale.

4. Empirical Analysis and Evaluation

4.1 Selection of Analysis DMUs and Variables

For evaluation accuracy, the number of DMUs should be sufficiently greater in comparison to the number of input and output variables. According to Banker et al. (1984), the proper number of DMUs should be at least three times greater than the total number of input and output variables, and Boussofiane et al. (1991) state that the number of DMUs should be at least greater than the number of output variables multiplied by the number of input variables.

DMUs selected in this study are terminals of nine ports: Busan Port in Korea and Continental Chinese ports in Shanghai, Shenzhen, Ningbo, Qingdao, Guangzhou, Tianjin, Dalian, and Xiamen. The factors considered on selection are as follows: First, the terminal should have been operated for at least 5 years and entered a stable operation stage. Second, significant data should be available from the official website of the terminal and/or that of its port. Third, the total traffic volume of the terminal should account for most of the traffic volume of its port.

Data of analysis subjects are mostly collected from sources as follows:

® Data from the official website of the port and published data from the port.

® Data from the official website of the terminal and published data from the terminal.

® Monthly statistics by container terminals published by China Ports and Harbours Association.

@ Annual statistics published by Korea Port Logistics Association.

© Data collected by direct contact with terminal operators as well as site visits.

The number of terminals finally selected are 30 as listed in Table 2:

Table 2

Summary on Comparison Subject Terminals

Country (Quantity) Port Terminal Subtotal

Korea(5) Busan KBCT, PNIT, PNC, HJNC, HPNT 5

Shanghai SSICT, SGICT, SPICT, SIPGZCT, SECT, SMCT, Yidong 7

Shenzhen SCT, CCT, YICT 3

Ningbo NBSCT, NBCT, Gangji, YDCT, CMICT 5

China(25) Qingdao QQCT 1

Guangzhou GOCT, GNICT 2

Tianjin TPCT, FICT, TCT, TACT, TOCT 5

Dalian DCT 1

Xiamen Guojigangwu 1

Source: Author

In section 2, we have observed that many previous studies used the same natures of variables such as number of berth and berth length. This reduces the accuracy of the results. Thus, in this study, only different natures of variables that mostly associated with efficiency are selected. They are berth length, the yard area, the number of quay cranes and the number of yard cranes and the output variable is the container traffic volume. The evaluation indicators are listed in Table 3:

Table 3

Summary of input and output variables

Input Variable Output Variable

Berth length(m)

Yard Area (m2) Container Throughput(TEU)

Number of Quay Cranes

Number of Yard Cranes

Source: Author

Table 4 lists 2014 data for input and output variables used in this

analysis.

4.2 Empirical Analysis

The output-oriented DEA-CCR, DEA-BCC, and Scale Efficiency model are applied for the evaluation of 30 container terminals as of 2014, and DEAP 2.1 developed by Tim Coelli (1996) is used in this analysis. The analysis results are listed in Table 5. The efficiency of major terminals in Korea (OCR: 0.815, BCC: 0.886) show similar efficiency with China's terminals (CCR: 0.817, BCC: 0.887).

Table 4

Summary of input and output variables

Port Terminal Berth Length (m) Yard Area (m!) Number of QC Number of TC Container Throughput (TEU)

Shanghai SSICT 3,000 1,490,000 34 105 8,100,018

SGICT 2,600 1,418,000 30 72 7,101,700

SPICT 900 278,000 11 42 2,373,619

SIPGZCT 1,985 1,080,200 26 73 6,173,682

SECT 1,437 920,000 14 48 3,451,595

SMCT 2,258 1,192,000 26 87 5,164,888

Yidong 1,641 608,000 14 35 2,705,579

Shenzhen SCT 4,090 1,000,000 33 105 5,065,753

CCT 3,138 1,020,000 37 108 4,769,889

YICT 6,390 3,730,000 74 200 11,672,798

Busan KBCT 1,500 804,000 15 34 2,190,665

PNIT 1,200 384,000 11 30 1,712,728

PNC 2,000 525,000 19 61 3,895,202

HJNC 1,100 283,000 12 42 2,467,741

HPNT 1,150 246,000 12 38 2,552,383

Ningbo NBSCT 1,258 700,000 18 58 3,656,287

NBCT 900 400,000 11 39 2,020,014

Gangji 1,700 911,700 20 60 4,700,419

YDCT 1,710 615,000 20 50 3,214,879

CMICT 1,500 800,000 16 48 2,500,916

Qingdao QQCT 6,563 4,620,000 74 193 16,624,400

Guangzhou GOCT 2,100 2,030,000 22 64 4,650,736

GNICT 1,820 1,148,000 25 52 6,036,557

Tianjin TPCT 2,300 1,800,000 23 58 3,009,771

FICT 1,202 350,000 12 31 2,569,599

TCT 1,223 530,000 14 38 2,410,381

TACT 1,100 562,000 11 33 2,610,926

TOCT 1,138 450,000 10 21 1,451,073

Dalian DCT 1,846 585,000 14 41 3,787,813

Xiamen Guojigangwu 3,036 1,348,000 26 60 3,746,954

Source: Official websites of ports and terminals; the equipment status in 2014 summarized from the materials published by Korea Port Logistics Association; throughput statistics of major port container terminals at the end of December in China from the magazine of China Ports Association (2015); and Busan Port Authority.

Note: For data accuracy, we've visited Ningbo, Tianjin, Shanghai and Shenzhen in 2013 and 2014. The berth length of YICT was officially announced as 7,885m (2014), but it was incorrect and it was excluded. 30,000m2 of ODCY was added to HPNT. For Qingdao Port, data for QQCT and QQCTU terminals was combined due to the throughput collection base.

Table 5

CCR, BCC, scale efficiency analysis results as of 2014

Port Terminal CCR BCC Scale Efficiency

Shanghai SSICT 0.962 1 drs 0.962

SGICT 0.953 1 drs 0.953

SPICT 1 1 crs 1

SIPGZCT 1 1 crs 1

SECT 0.962 0.991 irs 0.971

SMCT 0.8 0.817 drs 0.979

Yidong 0.759 0.841 irs 0.903

Shenzhen SCT 0.641 0.849 drs 0.755

CCT 0.598 0.788 drs 0.759

YICT 0.636 0.822 drs 0.773

Busan KBCT 0.589 0.678 irs 0.869

PNIT 0.652 0.804 irs 0.811

PNC 0.888 1 drs 0.888

HJNC 0.948 0.95 drs 0.998

HPNT 1 1 crs 1

Ningbo NBSCT 0.92 0.962 irs 0.956

NBCT 0.789 1 irs 0.789

Gangji 0.95 0.956 irs 0.993

YDCT 0.751 0.786 drs 0.956

CMICT 0.621 0.635 irs 0.978

Qingdao QQCT 0.901 1 drs 0.901

Guangzhou GOCT 0.837 0.85 drs 0.985

GNICT 1 1 crs 1

Tianjin TPCT 0.513 0.53 drs 0.969

FICT 1 1 crs 1

TCT 0.725 0.758 irs 0.957

TACT 0.931 1 irs 0.931

TOCT 0.605 1 irs 0.605

Dalian DCT 1 1 crs 1

Xiamen Guojigangwu 0.578 0.596 drs 0.97

China Average 0.817 0.887 - 0.922

Korea Average 0.815 0.886 - 0.913

Source: Author

In the CCR analysis, the number of terminals evaluated as efficient are 6, which account for 20% of the total terminals. Major productivity indicators are analysed for 6 terminals of which efficiency is evaluated as 1 as of 2014. The analysis results are listed in Table 6. SPICT Terminal shows high yard throughput since a relatively large number of yard cranes are used in the relatively small yard area. SIPGZT shows high berth throughput (3,110TEU/m) with a lot of QCs and smoothly handles the traffic generated from the berth with the relatively large yard area secured (190,463m2 per berth). Consequently, it maintains overall high efficiency. The reason for the high efficiency of HPNT Terminal is that it handles a relatively large traffic volume compared to the significantly smaller yard area (74,870m2 per berth) than any other terminals. Like SIPGZCT, GNICT also smoothly handled the traffic volume from the berth with a significantly amount of cranes and a large yard area. The reason why FICT and DCT show high efficiency is that the large traffic volume is handled with a relatively small number of resources. Particularly, the number of QCs and that of transfer cranes of DCT

Terminal are significantly lower than any other terminal. Unlike the DEA-CCR model, the DEC-BCC model assumes that the returns to scale vary because given DMUs are composed of block groups. That is, the model assumes variable returns to scale (VRS). In the BCC analysis, the number of terminals evaluated as efficient are increased compared to that of the DEA-CCR model, which implies that the reason why the total technical efficiency (CCR) is lower is due to scale inefficiency. Terminals analysed as efficient as of 2014 total 14, which are SSICT (Yangshan Port Phases 1 and 2), SGICT (Yangshan Port Phase 3), SPICT (Waigaoqiao Phase 1), and SIPGZCT (Waigaoqiao Phase 2), of Shanghai Port, PNC and HPNT of Busan Port, NBCT of Ningbo Port, QQCT of Qingdao Port, GNICT of Guangzhou Port, FICT, TACT and TOCT of Tianjin Port, and DCT of Dalian Port. Average efficiency of the total terminals is 0.887. In total, the number of terminals evaluated as efficient is higher than that of the DEA-CCR model, which implies that the reason why the total technical efficiency (CCR) is lower is due to scale inefficiency. Terminals that show 1 for both the BCC efficiency and the CCR efficiency are the 6 terminals of SPICT, SIPGZCT, HPNT, GNICT, FICT, and DCT, which

are analysed as the most efficient terminals.

Table 6

Key productivity indicators of terminals with 1 for total efficiency (CCR) as of 2014

Terminal Berth Throughput/ m(TEU) Yard Throughput (TEU/l,000m2) Yard Area/350m ofBerth (m2) Number ofQC/350m ofBerth Number of TC/350m of Berth

SPICT, CN 2,637 8,538 108,111 4.3 16

SIPGZCT.CN 3,110 5,715 190,463 4.6 13

HPNT, KR 2,219 10,376 74,870 3.7 12

GNICT, CN 3,317 5,258 220,769 4.8 10

FICT, CN 2,138 7,342 101,913 3.5 9

DCT.KR 2,052 6,475 110,915 2.7 8

Source: Author

The scale efficiency is estimated from the technical efficiency (CCR) are divided by the pure technical efficiency (BCC). Terminals that are analysed inefficient are SSICT, SGICT, PNC, NBCT, QQCT, TACT, and TOCT terminals by the scale efficiency analysis results even though the DEA-BCC model demonstrates their efficiency is 1 as of 2014. It means that overall inefficiency occurs due to the scale inefficiency and implies that inefficiency is the input scale should be reviewed if appropriate. 6 terminals show the constant return to scale (CRS) as of 2014. Among inefficient terminals, 13 terminals show the decreasing return to scale (DRS) and 11 terminals show the increasing return to scale (IRS). That is, 11 terminals can further improve their efficiency by expanding the scale, and 13 terminals should prepare a plan to enhance their efficiency by reducing their scale or enhancing the efficiency of their facilities.

Table 7 summarizes analysis results on CCR, BCC, and scale efficiency of comparison subject terminals. The average efficiency is maintained at approximately 0.8 as a whole by the CCR model, and the average efficiency under the BCC model is higher than that of the CCR model, which means relatively high efficiency is shown. The average scale efficiency of terminals is not less than 0.9, which is relatively high.

Port, particularly terminals within its new port, show high efficiency close to those of terminals in China by the result compared in the terminal unit.

It is necessary to carefully review the operational overload of terminals in Busan New port and vacant facilities in Busan North Port. Even though KBCT Terminal selected for evaluation handle the largest container throughput among the terminals in Busan North Port, its efficiency is relatively low. PNC, HJNC, and HPNT terminals within Busan New Port are operated very efficiently. It is necessary to watch SIPGZCT Terminal of Shanghai Port and GNICT Terminal of Guangzhou Port in China since both terminals are very efficient with high container throughput handled compared to facilities and equipment through the evaluation period. Particularly GNICT Terminal of Guangzhou Port handles the highest container throughput per berth length in Continental China, which is 3,300TEU per meter of a berth. However, some terminals in Busan New Port show efficiency close to it. Their high efficiency deserves lull appreciation, but to maintain such efficiency, it is necessary to review if service quality is impacted.

The current research will be extended in the future. More method such as Super-efficient DEA, Stochastic Frontier Analysis (SFA) and extensive evaluation objects should be applied to further research.

Table 7

Summary on Analysis Result by Each Model as of 2014

Division Analysis Model

CCR BCC Scale Efficiency

DMU Efficient 6(20%) 13(43%) CRS 6

Inefficient 24(80%) 17(57%) DRS 13

Total 30(100%) 30(100%) IRS 11

Average Efficiency 0.817 0.887 0.920

Source: Author

References

A. BOUSSOFIANE, R.G. DYSON and E. THANASSOULIS. (1991), "Applied data envelopment analysis", European Journal of Operational Research, Vol. 52, No. 1, pp. 1-15.

BANG HEE SEOK, KANG DONG JOON and PARK JAE HEON. (2011), "A Study on the Efficiency Analysis of Major Container Ports", Korea Trade Review, Vol 36, No. 2, pp. 1-23.

5. Conclusion and further research

In this article, the relative efficiency of container terminals in Korean and China is researched.

When comparing the efficiency of container ports, if the comparison of the objects involves different countries, we need to consider the business model of the container port in question. In most cases, they are operated by terminal units. Therefore, if we want to draw a correct and objective conclusion, we need to reduce the comparison scopes to the terminal unit.

Terminals in Korea show a similar level with the overall average. While previous studies (Kim et al., 2012; Li et al., 2015) conclude that the efficiency of ports in Korea is far lower than that of ports in China, Busan

BANKER R. D., CHARNES. A. and COOPER. W. W. (1984), "Some Models for Estimating Technical and Scale Inefficiencies in Data Envelopment Analysis", Management Science, Vol. 30, pp. 1078-1092.

BARROS C. P. (2003b), "The Measurement of Efficiency of Portuguese Seaport Authorities with DEA", International Journal of Transport Economics, Vol. 30, No. 3, pp. 335-354.

CHARNES, A., COOPER, W. W. and RHODES, E. (1978), "Measuring the Efficiency of Decision Making Units", European Journal of Operational Research, Vol. 2, No. 6, pp. 429-444.

CHINA PORT ASSOCIATION. (2015), "Throughput statistics of major port container terminals at the end of December in China", China Ports, Vol. 271, Issue. l,pp. 47-48.

CULLINANE KEVIN, SONG DONG WOOK, JIPING and WANG TENG FEI. (2004), "An application of DEA Window Analysis to Container Port Production Efficiency", Review of Network Economics, Vol. 3, No. 2, pp. 1-23.

ITOH. (2002), "Efficiency Changes at Major Container Ports in Japan: A Window Application of Data Envelopment Analysis", Review of Urban and Regional Development Studies, Vol. 14, No. 2, pp. 133-152.

KIM MIN SOO and HUANG TIAN CI. (2012), "A Comparative Study on the Competitiveness of Major Shipping Ports in Korea and China after the Global Financial Crisis", China and Sinology, No. 16, pp. 1-34.

KOREA PORT LOGISTICS ASSOCIATION. (2015), "2014 Equipment Status", Korea Port Logistics Association.

LEE JANG WON, KIM HYEONG KI and KIM SUNG HO. (2007), "A Comparative Study of the Competitiveness among Korean, Chinese, and Japanese Major Ports", Journal of International Area Studies, Vol. 11, No. 4, pp. 333-360.

LI. D, KWAK. K. S, NAM. K. C, and AN. Y. M. (2015), "A Comparative Analysis of Terminal Efficiency in Northeast Asia Container Ports", Journal of Korean navigation andPort Research, Vol. 39, No. 1, pp. 55-60.

MARTINEZ BUDRIA, DIAZ-ARMAS, NAVARRO IBANEZ, and RAVELO MESA. (1999), "A Study of the Efficiency of Spanish Port Authorities-using Data Envelopment Analysis", International Journal of TransportEconomics, XXVI, pp. 237-253.

ROLL Y & HAYUTH Y. (1993), "Port Performance Comparison applying data envelopment analysis", Maritime Police and Management, Vol. 20, No. 2, pp. 153-161.

TIM COELLI. (1996), "A guide to DEAP version 2.1: a data envelopment analysis (computer) program", Centre for Efficiency and Productivity Analysis.

TONGZON, J. (2001), "Efficiency Measurement of selected Australian and Other International Ports Using Data Envelopment Analysis", Transportation Research Part ,4, Vol. 35, No. 2, pp. 113-128.

TURNER HUGH, WINDLE ROBERT and DRESNER MARTIN. (2004), "North American container port productivity: 1984-1997", Transportation Research Part E, Vol. 40 No. 4, pp. 339-356.