Description
The report about the analytic hierarchy process (AHP) technique to determine the importance of various criteria in the transhipment port selection decision-making process.
International Journal of Logistics: Research and Applications Vol. 6, No. 4, 2003
Transhipment Port Selection and Decision-making Behaviour: Analysing the Taiwanese Case
TAIH-CHERNG LIRN,* HELEN A. THANOPOULOU & ANTHONY K. C. BERESFORD
Logistics and Operations Management Section, Cardiff Business School, Cardiff University, UK
ABSTRACT
The study uses the analytic hierarchy process (AHP) technique to determine the importance of various criteria in the transhipment port selection decision-making process. The authors propose a set of transhipment port selection criteria from a container carrier’s perspective. Sourcing the data from an AHP survey in Taiwan, transhipment port selection is found, tenuously at this ?rst stage of research, to depend mainly on port competitiveness as represented by the cost that carriers are faced with for loading and discharging of containers and on port ef?ciency as represented by the container loading and discharging rates. Fuzzy multiple criteria decision-making methodology (FMCDM) was applied to obtain evaluations of port alternatives and the relevant responses defuzzi?ed to derive crisp values of port performance.
Introduction The paper addresses transhipment port selection from a carrier’s perspective. Decision-making behaviour is becoming critical in order to compete in the liner container market as the market itself is under increasing competitive pressures. In
*Correspondence: Taih-Cherng Lirn, Logistics and Operations Management Section, Cardiff Business School, Cardiff University, Aberconway Building, Colum Drive, Cardiff CF10 3EU, UK, E-mail: [email protected] International Journal of Logistics ISSN 1367-5567 Print=ISSN 1469-848X online # 2003 Taylor & Francis Ltd http:==www.tandf.co.uk=journals DOI: 10.1080=13675560310001626990
230 T.-C. Lirn et al.
the 21st century major container carriers are competing globally for cargoes belonging to global shippers; this requires the provision of a global network by seacontainer carriers. The requirements of shippers who combine resources and distribute products globally are global coverage (i.e. coverage of the main trade routes), high frequency and speed, while maintaining the lowest possible cost and highest possible quality of service. This already ambitious target for today’s liner carriers is also combined with the requirements of shippers for the provision of more than just the shipping link of the transport and distribution chain. In this context, global carriers are not only required to provide the organisation of intermodal transport, but also the provision of total logistics often including added value services as discussed in Thanopoulou et al. (1999) and in Ryoo & Thanopoulou (1999). Amidst these pressures, port selection becomes a key element not only for achieving a cost advantage, but also for achieving product differentiation to the extent that differences in the quality of service are recorded as signi?cant. This is likely to be due to the demand of quality-sensitive clients such as shippers of valuable consumer and intermediate goods. This is unlike bulk shipping where research has indicated that, despite some level of nominal differences, there has been only marginal preference for a higher quality of service as highlighted by Tamvakis & Thanopoulou (2000). The liner container industry is different not just in this respect; it has also undergone major changes in terms of the world distribution of container capacity. Asia has risen to become a major power in liner shipping as well as in the world container trades. Therefore, the authors also aim to highlight more general trends in modern liner shipping. This research focuses on the transhipment decision-making behaviour of large Taiwanese international container carriers, which include some of the largest and most prominent international liner operators world-wide. Decision-making behaviour is particularly important not only for international container operators, but also for container port authorities and terminal operating companies. The increasing concentration within the industry has increased the potential impact of a move by a major port user on the individual port’s overall traf?c. The research approaches transhipment port selection decision-making by focusing on data sourced from a ?eld survey conducted in Taiwan by the ?rst-named author and using analytic hierarchy process (AHP) methodology. The main purpose of the research has been to identify the importance of factors that in?uence transhipment port choice by Taiwanese ocean container carriers. Identifying the criteria was an exercise based partly on a review of existing literature and on the results of the Delphi technique, which was applied to validate general port selection criteria in a transhipment context. The relative importance of the major criteria and sub-criteria revealed by the research indicates the directions to be explored by port operators in their effort not only to increase their own competitiveness vis-a-vis other ports, but also to facilitate transhipment activity by achieving lower costs, becoming more user-orientated and optimising resource allocation, especially as liner shipping networks are bound to expand and become more dense as globalisation intensi?es.
Background to Transhipment Port Selection Many container ports are expanding berth and loading=discharging facilities in order to meet the anticipated container throughput increase in the new century.
Transhipment Port Selection
231
FIGURE 1. Top World Container Ports by Throughput. Source: Compiled from Data in Different Issues of Containerisation International Yearbook 1999–2003.
Container traf?c has increased dramatically in the last 20 years. In the early 1980s, sea-container traf?c was just 1.56 million TEUs for Hong Kong and a little over 1 million TEUs for Singapore and Kaohsiung (Containerisation International, 1982, p. 77). By 1986, container throughput had increased to 2.78 million TEUs for Kaohsiung, 2.77 million TEUs for Hong Kong and 2.20 million TEUs for Singapore (Containerisation International Yearbook, 1987). Over the following 15 years container traf?c through Singapore and Hong Kong rose to 17.90 and 15.57 million TEUs, respectively (Containerisation International Yearbook, 2002). According to data for the year 2001, container throughput increases have been spectacular for the whole of the East and South Asian regions but the growth has not been evenly spread. Container throughput for Kaohsiung in 2001, at only 7.54 million TEUs, was well below Singapore’s and Hong Kong’s volumes. Although both the absolute ?gures and the respective rates of increase from the mid-1980s are impressive, they have to be seen in the context of port competition in that area. Kaohsiung was still ranked fourth among top container ports world-wide at the turn of the century, but the margin separating it from the top three ports has increased as its container traf?c is now less than half the volume of its major competitors (Figure 1). With competition between the different ports increasing, the risk of overinvestment in port facilities might become a serious problem in some of the major Asian ports. Knowing the preferences of carriers for transhipment ports and knowing how these preferences are formed becomes critical. The recent moves of major carriers between Asian ports and the increasing importance of individual global carriers in the traf?c of individual ports emphasise the importance of transhipment port selection.
Literature Review and Selection of Methodology The research adopted AHP to reveal the importance of transhipment port selection criteria in transhipment decision-making. The AHP methodology is a multiple criteria decision-making (MCDM) technique that provides the relative importance of factors involved in decisions through pairwise comparisons of the same level independent criteria. Pairwise comparisons are obtained through surveying either the whole population or a sample of decision-makers. The hierarchical structure of
232 T.-C. Lirn et al.
levels of criteria involved in the survey is critical for the successful application of AHP (Lai et al., 2002) as is the number of criteria adopted at each level. It has been observed that when ?ve or more items are involved in the survey questionnaire, the decision-maker (DM) statements regarding pairwise comparisons tend to become increasingly inconsistent in terms of transitivity (Bodin & Gass, 2003; Gass, 1998). Clearly structuring and de?ning the number of criteria is a critical preliminary stage of an AHP survey. In the literature, it has been suggested that de?ning independent criteria, and at the same time limiting their numbers at each level of an AHP hierarchy, can be achieved through using factor analysis (Park & Han, 2002; Sohn et al., 2001), or the Delphi technique (Azani & Khorramshahgol, 1990; Suh & Han, 2003) and=or brainstorming (Al-Harbi & Al-Subhi, 2001; Dey, 1999). The research for this paper adopted the Delphi technique to validate general port selection criteria in a transhipment context and then brainstorming to limit the number of criteria to a level acceptable for use in the ?nal AHP survey as most average values remained ‘‘high’’ to ‘‘very high’’ through both Delphi rounds. While factor analysis is only appropriate if a large enough number of questionnaire replies can be obtained, usually ?ve times larger than the number of factors included (Hair, 1995), the Delphi technique remains suitable even where only a small number of experts are involved in the process of inclusion=elimination of criteria through repeated surveying rounds.
The Analytic Hierarchy Process The AHP methodology, ?rst developed by Thomas Saaty in 1971, has been widely used since its ?rst appearance in the Journal of Mathematical Psychology in 1977 (Saaty, 1977). The methodology falls into the category of MCDM approaches. As pointed out by Vreeker et al. (2002), the basic rules for solving multi-level hierarchical problems involve essentially four steps: (1) speci?cation of choice problem; (2) information analysis; (3) choice of appropriate method; and (4) evaluation of alternatives. As discussed in Saaty (1977, 1980) and more recently in Tam & Tummala (2001), the AHP modelling process involves four phases: (1) Structuring: formulating the AHP hierarchy in terms of objective, major criteria and sub-criteria, and rating scale used for the evaluation and for decisionalternatives to be evaluated. (2) Data collection: for this research a sample of Taiwanese container carriers was surveyed in 2001. (3) Obtaining the weights in different hierarchies: as in most of the literature known to the authors, the present paper has followed the geometric mean approach instead of the arithmetic one to improve results (Forman, 2000). The general form of the matrix of ratio comparisons of the AHP is described in Saaty (1977) and in Saaty & Vargas (1994). (4) Synthesis: in the case of the research presented in this paper, the ?nal step has been to propose a solution to transhipment port selection on the basis of AHP sub-criteria weights and the performance scores of transhipment port alternatives. The scope for successful applications of AHP has proved to be exceptionally wideranging, from the selection of fuel source for transportation in the future in
Transhipment Port Selection
233
Winebrake & Creswick (2003) to the evaluation of the top 20 decision-making support system journals (Forgionne et al., 2002). The limitations and drawbacks of AHP have been debated in several journals. One of the major criticisms of AHP is that it allows ‘‘rank reversals’’, where the order of the alternatives changes when any of them is added or deleted. This thus violates the invariance principle of utility theory. Dyer (1990a, b), Harker & Luis (1990) and Vargas (1997) have debated the validity of AHP, while some major criticisms of AHP have been successfully defended by Saaty (1990a, b, 2000). One of the problems AHP had to solve concerned the reliability of values attributed to the pairwise comparisons by participants in the surveys. Beynon (2002) found that under certain conditions alternative scales were following linguistic scales more effectively than Saaty’s original one- to nine-point scale, especially in relation to the symmetry of argument. Hurley (2001) reported that decision-makers are frequently certain about the rank order of the objects for a particular pairwise comparison matrix, but uncertain about the precise numerical weights that the AHP produces for that matrix. However, the applicability of AHP to a given data set can be enhanced through the application of fuzzy techniques. Triantaphyllou & Lin (1995) concluded that fuzzy AHP approaches are capable of capturing a human’s appraisal of ambiguity where complex decision-making problems are concerned. The scope for applying fuzzy techniques within AHP can have quite concrete applications. Tsaur et al. (1997) used fuzzy AHP to evaluate tourist risks for ?ve destinations, Kuo et al. (1999) proposed a decision support system for locating convenience stores through fuzzy AHP, while Ravi & Reddy (1999) adopted fuzzy AHP in order to rank the suitability of Indian coking and non-coking coals for industrial use.
Applying AHP to Transhipment Port Selection There are at least 20 known applications of AHP on transport problems, such as, for example Vreeker et al. (2002), Chang & Yeh (2001), Poh & Ang (1999), Tzeng & Wang (1994) and Frankel (1992). Within the wider transport sector, AHP continues to be used as a valid approach and this applies to many recent articles focusing on, or emanating from, Taiwan. Tsai & Su (2002) utilised AHP to assess the political risk attached to the development of an air logistics hub in Taiwan. Chou & Liang (2001) used AHP and proposed the use of linguistic values in order to create a model capable of evaluating the performance of shipping companies; the fuzzy approach they adopted has been considered as capable of addressing the problem of subjective=imprecise assessments related to the performance evaluation problem. The port performance evaluation problem is not very different in this respect as it involves subjective evaluations of the different attributes of ports that play a role in the port selection decision-making process. Fuzzy set theory was developed from the premise that the key elements in human thinking are not numbers, but linguistic terms or labels of fuzzy information sets (Zimmermann, 1991). The adoption of a fuzzy approach has been deemed an appropriate methodology for a ?rst approach to the transhipment port decisionmaking process in order to test the real extent of problems that assigning values could create in the context of wider research. Therefore, semantic words have been adopted to evaluate the performance of three major Taiwanese international ports
234 T.-C. Lirn et al.
and evaluators were requested to provide a triangular fuzzy number (TFN) within the scale range of 0–100 for each different semantic word used. Defuzzi?cation was then applied to convert the fuzzy set obtained into a real number. The centre of area (COA) method is the most widely used technique for computing defuzzi?ed values (Patel & Mohan, 2002), the mean of maximum method and the median rule being the two main alternative techniques used to defuzzify the initial fuzzy sets obtained. The COA method determines the centre area of membership functions. Equation (1) (Tsaur et al., 1997, 2002) was used in order to convert the fuzzy set of numbers denoted by Ri to a crisp value denoted as BNPi, the best non-fuzzy performance value of alternative i. BNP¼ i ½(URÀ LRiþ (MRÀ LRi Þ )] i i þ LRi : 3 (1 )
In equation (1), BNPi represents the best non-fuzzy performance value, and URi, MRi, and LRi are the components of a fuzzy number Ri that is the result of a fuzzy multiplication. MRi is a product of the representative value of the highest grade of membership in a triangular fuzzy set multiplied by the weight obtained from AHP. The terms LRi and URi denote the products of the representative values of the lower boundary and upper boundary of a triangular fuzzy set, where the membership is lowest, multiplied by the weight obtained from AHP.
Deriving AHP Criteria The starting point of this preparatory phase of the AHP survey used in this research was the identi?cation of relevant criteria through a literature review. While there is limited literature on transhipment port choice, factors affecting port choice in general and factors affecting the general competitiveness of container ports are well documented in the literature. Combining criteria declared in recent transhipment port selection cases (Porcari, 1999; Villalon, 1998) and port choice criteria used in academic literature gave an overall total of 47 port selection criteria (see Table 1). The large number of registered criteria made the use of factor analysis at this preliminary stage prohibitive as the number of experts to be surveyed would have to be at least two to three times larger than the 47 criteria found. An added limitation would be that experts involved in the hierarchical structuring task through an eventual factor analysis and those being surveyed in the AHP survey must not overlap. As there are no minimum limits on expert numbers involved in the Delphi technique, the non-overlap rule was easy to follow. The Delphi technique using ?ve-point Likert scales was used at a ?rst stage as the basis for validating criteria and at a second stage as the basis for fusing the large number of criteria identi?ed from the literature review through brainstorming. The criteria were presented to a total of 10 experts, senior industry executives and scholars in the area of shipping, during a number of interviews and evaluation rounds conducted in Taiwan. The initial large number of criteria was reduced by being further classi?ed by the surveying author into six major criteria: (1) Water depth of the port. (2) Local infrastructure and port transportation network. (3) Area of the container yard and of marshalling yard.
TABLE 1. Factors Affecting the Carriers’ Port Choice Behaviour
Writer
Criteria Villalon (1998) Hayuth (1995) Slack (1985) Baird (2000) Porcari (1999) Burnson (2001) Jansson & Shneers (1987) Thomas (1998) 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 (continued) 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Brooks (2000) Frankel (2001) Fleming & Baird (1999) Murphy et al. (1989) Browne et al. (1989) Branch (1986)
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.
Available number of berths Back-up space on terminal Congestion Cargo volume Cargo-generating effect Other modes Competitiveness Container handling ef?ciency Containerised cargo proportion Degree of integration (EDI) Depth of the port Geographical advantage Flexible operation process Free time Frequency of feeder service Frequent port of call Infrastructure Inland freight rates Labour problems Loading=discharging rate Low cost Major container centre Numbers of sailing Operation Related business operation Port accessibility (land and sea) Port operation=working hours Port berthing time length Port charges Price conditions Port equipment Port security Port service coverage Port tradition and organisation Privilege contract to carrier Port productivity
Transhipment Port Selection 235
236 T.-C. Lirn et al.
TABLE 1. Continued
Writer
Criteria Villalon (1998) Hayuth (1995) Slack (1985) 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Baird (2000) Porcari (1999) Burnson (2001) Jansson & Shneers (1987) Thomas (1998) Brooks (2000) Frankel (2001) Fleming & Baird (1999) Murphy et al. (1989) Browne et al. (1989) Branch (1986)
35. Proximity to alternative loading centre 36. Quality of customs handling 37. Regulations 38. Service considerations 39. Size of hinterland 40. Size of port=terminal 41. State aides and in?uence on cost 42. Superstructure 43. Intermodal link=network 44. Transportation and port-user cost 45. Time on the route 46. Transit time 47. Trade inertia
Transhipment Port Selection
237
TABLE 2. Qualitative Categorising 47 Criteria into Four Major Criteria and 16 Sub-criteria
Notes: A1, A2, mode quantity; B1, B2, average value; C1, C2, standard deviation; D, mode value; E, mode quantity difference between round one and round two surveys ( ¼ A1 7A2); CAT, category; F, standard deviation reduction value ( ¼ C1 7C2); No., numbering of CAT. III sub-criteria; C.Y., container yard. The Delphi panel is composed of 10 experts, ?ve persons are selected from academia and ?ve persons are from the shipping industry in Taiwan. Criteria in the literature for general port selection are all validated for transhipment ports with high average scores by the Delphi technique. CAT. IV are variables derived from literature review, CAT. III and CAT. II are allocated by corresponding author, CAT. I is categorised by consultation with two academics in Taiwan Ocean University.
238 T.-C. Lirn et al.
(4) Basic cargo volume (non-transhipment cargo). (5) Geographical advantage: location near the main service routes. (6) Port ef?ciency and cost of container-handling for carriers. This list was further shortened through brainstorming sessions with academics in Taiwan to derive the four major criteria that were used in the survey questionnaire along with 16 sub-criteria (see Table 2. The major criteria and their subdivision into relevant sets of sub-criteria are the two system components of the AHP (see Figure 2).
FIGURE 2. Determining Optimum Transhipment Port for Major Taiwan Container Carriers using AHP.
Transhipment Port Selection
239
The AHP Survey and Results A ?eld survey was conducted among four major international container carriers in Taiwan. A questionnaire was presented to the users in face-to-face meetings. In the ?rst part, the respondents conducted pairwise comparisons of the importance of various transhipping port selection criteria. In the second part, the evaluators adopted ?ve linguistic variables, namely excellent, good, average, fair and poor (Tsaur et al., 1997), to evaluate port performance on the basis of the 16 sub-criteria. Non-fuzzy performance measures were then derived using equation (1) to arrive at a crisp value of performance for each of the three major international ports in Taiwan based on the carrier evaluations. Carriers are identi?ed in this paper with the sequential letters A, B, C and D to avoid any breach of anonymity. The results show that the different participants in the survey attribute varying importance to the major criteria for transhipment port selection. Both A and B carriers are mostly concerned with ‘‘port management’’. B carrier also attributed signi?cant importance to ‘‘port geographical location’’. The C and D carriers were found to be concerned mostly with ‘‘port basic physical characteristics’’ and ‘‘carriers cost perspective’’, respectively (see Table 3). In terms of the 16 sub-criteria used (see Figure 2), there were also differences among the respondents. When the global weights of the criteria are considered as an average for all respondents, the results indicate that all four major criteria have almost the same weight in in?uencing transhipment port selection behaviour. In terms of subcriteria, the companies surveyed seem to attribute most weight to ‘‘proximity of the feeder ports’’ (global G.M. weight 9.06%), ‘‘berthing delay and loading=discharging rate’’ (global G.M. weight 10.84%) and ‘‘carriers loading=discharging cost’’ (global G.M. weight 11.59%), as shown in Table 4. Performance of Taiwanese Ports By using semantic words, the carriers evaluated the performance of the three major international ports according to the sub-criteria for transhipment port selection. Equation (1) was then applied to calculate the BNP values. The survey responses also reveal the limits of the methodology applied for performance evaluation purposes as one of the carriers did not proceed to evaluate the overall performance of the three ports involved, possibly in order to avoid offending port terminal operators. The performance evaluation by the remaining respondents resulted in port X being consistently ranked ?rst (Table 5). The overall results can also be deemed as
TABLE 3. Importance of Each Major Criterion Revealed Through the AHP Survey Carrier A (%)
Port basic physical characteristics Port geographical location Port management Carriers’ cost perspective 8.48 5.99 42.77 42.77
Carrier B (%)
5.89 45.58 42.64 5.89
Carrier C (%)
69.16 13.97 9.88 6.99
Carrier D (%)
8.04 39.81 5.62 46.53
240 T.-C. Lirn et al.
TABLE 4. Global Weights of Transhipment Port Selection Criteria Major Criteria (First-tier Factors)
Port basic physical characteristics
Major Criteria: Global Weight (%)
22.89
Sub-criteria (Second-tier Factors)
Infrastructure improvement Port facilities and equipment Convenience of inter-modal link Size of marshalling yard and container yard The closeness to the import= export consumption areas The closeness to the main navigation route Proximity of the feeder ports Proximity of competing ports and modes Port administration and custom regulation Berthing delay and loading= discharging rate Port safety Terminal security Carrier’s loading=discharging cost Ownership of the port and terminal Privileged terms to the carriers Government levy and duty
Sub-criteria: Global Weight (%)
1.55 7.48 8.72 5.14 8.07 5.43 9.06 3.78 3.85 10.84 9.01 1.53 11.59 8.15 3.65 2.15
Port geographical location
26.34
Port management
25.23
Carriers’ cost perspective*
25.54
Note: Italic characters indicate the most important criterion and sub-criterion as revealed by AHP.
objectively consistent. The prominence of the carrier’s cost perspective and of the loading and discharging cost among the sub-criteria was con?rmed by the comparison of port rankings through the survey and the cost ranking of ports involved. Following a similar approach in the case of airlines (Chang & Yeh, 2001), actual port loading and discharging cost data were used to substitute for the scores obtained through the AHP performance evaluation; this resulted in the same ranking of the three major international ports in Taiwan as the one derived through the AHP survey. However, the comparison was made on the basis of actual data relevant only to one of the sub-criteria and should be taken into account only as such.
TABLE 5. BNP Values Calculated from the Evaluation of Three Major Ports in Taiwan Port
Carriers Carrier A Carrier B Carrier C Carrier D Total ( ¼ B þ C þ D) Port X NA 90.50 77.40 88.82 256.72 Port Y NA 81.72 75.21 73.84 230.77 Port Z NA 77.38 74.70 70.77 222.85
Notes: BNP represents best non-fuzzy performance; NA, not available.
Transhipment Port Selection
241
Conclusions, Limitations and Further Research The set of data used for this research was limited as carriers surveyed were from one country and this is acknowledged as a signi?cant limitation. However, the small sample surveyed included all global carriers and regional ones. All ports and carriers surveyed were in and from Taiwan, respectively. This allowed the research to remain free from considerations regarding regional pressures, which could eventually affect port choice. If a broader regional study were to be carried out, the political risk factor, which has not been included to date in port selection literature, could eventually become a more in?uential attribute, although that would depend on the de?nition of the geographical range of the region considered. In any case, the ?ndings of the research are consistent with the feeling of the market on the prevalence of cost as a major criterion from a carrier’s perspective. The initial ?ndings are also consistent with informally expressed views from the industry (personal interview with a senior executive of a global carrier in autumn 2002), while the initial results of an ongoing research on global carriers seem to indicate the prevalent signi?cance of the handling cost among the criteria used in the survey (Lirn et al., 2003). There is also scope for improving the methodology adopted in the future. Factor analysis can be used to provide an alternative approach for narrowing down the number of proposed criteria and a more objective method for distinguishing between criteria and sub-criteria. Furthermore, the fuzzy performance evaluation by semantic language was found to be dif?cult by most interviewees taking part in this research and it required face-to-face explanations before the interviewees felt able to provide a score range for their related semantic evaluations. This problem has also been reported by Shieh (2000) and seems to indicate that broader research through questionnaires can only be undertaken through a non-fuzzy AHP in order to increase the response rate. Another important conclusion emanating from the survey itself was that of the familiarity of interviewees with the basics of the AHP methodology. The survey indicated that if the basics of AHP had not been explained to the interviewees, it was very easy for them to answer all questions but their answers could be seriously inconsistent, with eventual repercussions for the validity of the results (Saaty, 2001). However, if the AHP methodology is explained to the interviewees, the value of the consistency index reduces signi?cantly and results are much more reliable. The major drawback of providing information on AHP in person is that of the potential introduction of a bias, and the solution seems to lie in providing summary information on AHP through the questionnaire itself. Research which two of the authors are currently conducting on transhipment port selection by major global carriers has adopted this approach, taking account of other limitations that have arisen in this research related to the uniformity of language expressions and the large number of sub-criteria. Finally, AHP can also provide the basis for required updates on the transhipment port selection problem. As no carrier operates in isolation, decisionmaking is bound to change following shifts in the environment in which carriers operate. A systematic update of the research would reveal evolving trends.
Acknowledgements The authors would like to thank two anonymous referees for helpful and constructive comments as well as the editors of the issue. Thanks are expressed to
242 T.-C. Lirn et al.
Dr Malcolm Beynon for comments in the early stages of the writing of the paper. Professor S. Yahalom and Mr P. Chou from the National Taiwan Ocean University provided valuable comments to the ?rst author in the initial stages of this research.
REFERENCES
AL-HARBI, K. & AL-SUBHI, M. (2001) Application of the AHP in project management, International Journal of Project Management, 19, (1), pp. 19–27. AZANI, H. & KHORRAMSHAHGOL, R. (1990) Analytic Delphi Method (ADM): A strategic decisionmaking model applied to location planning, Engineering Costs and Production Economics, 20, (1), pp. 23–28. BAIRD, A.J. (2000) Port privatisation: Objectives, extent, process and the UK experience, International Journal of Maritime Economics, 2, (3), pp. 177–194. BEYNON, M. (2002) An analysis of distributions of priority values from alternative comparison scales within AHP, European Journal of Operational Research, 140, (1), pp. 104–117. BODIN, L. & GASS, S.I. (2003) On teaching the analytic hierarchy process, Computers and Operations Research, 30, (10), pp. 1487–1497. BRANCH, A.E. (1986) Elements of Port Operation And Management (London, Chapman and Hall). BROOKS, M. (2000) Sea Change in Liner Shipping—Regulation and Managerial Decision-making in a Global Industry (Oxford, Elsevier Science). BROWNE, M., DOGANIS, R. & BERGSTRAND, S. (1989) Transhipment of UK Trade (London, British Ports Federation). BURNSON, P. (2001) Sea change, Global Logistics, 1 July, LexisNexis (online). CHANG, Y.H. & YEH, C.H. (2001) Evaluating airline competitiveness using multiattribute decisionmaking, Omega, 29, (5), pp. 405–415. CHOU, T.Y. & LIANG, G.S. (2001) Application of a fuzzy multi-criteria decision-making model for shipping company performance evaluation, Maritime Policy and Management, 28, (4), pp. 375–392. Containerisation International (1982) Informa Maritime and Transport, Essex, June. Containerisation International Yearbook (1987, 1999–2003) Informa Maritime and Transport, Essex. DEY, U.K. (1999) Process re-engineering for effective implementation of projects, International Journal of Project Management, 17, (3), pp. 147–159. DYER, S.J. (1990a) Remarks on the analytic hierarchy process, Management Science, 36, pp. 249–258. DYER, S.J. (1990b) A clari?cation of remarks on the analytic hierarchy process, Management Science, 36, pp. 274–275. FLEMING, D.K. & BAIRD, A.J. (1999) Some re?ections on port competition in the United States and Western Europe, Maritime Policy and Management, 26, (4), pp. 383–394. FORGIONNE, G.A., KOHLIB, R. & JENNINGS, D. (2002) An AHP analysis of quality in AI and DSS journals, Omega, 30, (3), pp. 171–183. FORMAN, E.H. (2000) Decision by Objectives (Washington, George Washington University), available at http:==www.expertchoice.com=dbo= FRANKEL, E.G. (1992) Hierarchical logic in shipping policy and decision-making, Maritime Policy and Management, 19, (3), pp. 211–221. FRANKEL, E.G. (2001) Economics of transhipment in container shipping logistics, Inaugural International Conference on Port and Maritime R and D and Technology, Singapore, pp. 29–31. GASS, S.I. (1998) Tournaments, transitivity and pairwise comparison matrices, Journal of the Operational Research Society, 49, pp. 616–624. HAIR, J.F. (1995) Multivariate Data Analysis (Englewood Cliffs, NJ, Prentice Hall). HARKER, T.P. & LUIS, G.V. (1990) Reply to remarks on the analytic hierarchy process, Management Science, 36, pp. 269–273. HAYUTH, Y. (1995) Container traf?c in ocean shipping policy, Proceedings of International Conference ‘‘Ports for Europe’’ Conference, Europacollege, Zeehaven Brugge, 23–24 November. HURLEY, W.J. (2001) The analytic hierarchy process: A note on an approach to sensitivity which preserves rank order, Computers and Operations Research, 28, pp. 185–188. JANSSON, J.O. & SHNEERS, D. (1987) Liner Shipping Economics (New York, Chapman and Hall). KUO, R.J., CHI, S.C. & KAO, S.S. (1999) A decision support system for locating convenience store through fuzzy AHP, Computers and Industrial Engineering, 37, (1–2), pp. 323–326. LAI, V.S., WONG, B.K. & CHEUNG, W. (2002) Group decision making in a multiple criteria environment: A case using the AHP in software selection, European Journal of Operational Research, 137, (1), pp. 134–144. LIRN, T.C., THANOPOULOU, H.A. & BEYNON, M.J. (2003) Applying AHP on Transhipment Port Selection: A Global Perspective, (Mimeo, Cardiff Business School).
Transhipment Port Selection
243
MURPHY, P.R., DALENBURG, D.R. & DALEY, J.M. (1989) Assessing international port operations, International Journal of Physical Distribution and Logistics Management, 19, (9), pp. 3–10. PARK, C.-S. & HAN, I. (2002) A case-based reasoning with the feature weights derived by analytic hierarchy process for bankruptcy prediction, Expert Systems with Applications, 23, (3), pp. 255–264. PATEL, A.V. & MOHAN, B.M. (2002) Some numerical aspects of centre of area defuzzi?cation method, Fuzzy Sets and Systems, 132, (3), pp. 401–409. POH, K.L. & ANG, B.W. (1999) Transportation fuels and policy for Singapore: An AHP planning approach, Computers and Industrial Engineering, 37, pp. 507–525. PORCARI, J.D. (1999) The logical choice for a Northeast hub, Journal of Commerce, 30 April 30, Lexis Nexis (online). RAVI, V. & REDDY, P.J. (1999) Ranking of Indian coals via fuzzy multi attribute decision-making, Fuzzy Sets and Systems, 103, pp. 369–377. RYOO, D.K. & THANOPOULOU, H.A. (1999), Liner alliances in the globalization era: A strategic tool for Asian container carriers, Maritime Policy and Management, 26, (4), pp. 349–367. SAATY, T.L. (1977) A scaling method for priorities in hierarchical structures, Journal of Mathmatical Psychology, 15, pp. 234–281. SAATY, T.L. (1980) The Analytic Hierarchy Process (New York, McGraw-Hill). SAATY, T.L. (1990a) An exposition of the AHP in reply to the paper remarks on the analytic hierarchy process, Management Science, 36, pp. 259–268. SAATY, T.L. (1990b) How to make a decision: The analytic hierarchy process, Interfaces, 24, (6), pp. 19–43. S AATY, T.L. (2000) On the relativity of relative measure—accommodating both rank preservation and rank reversals in the AHP, European Journal of Operational Research, 121, pp. 205–212. SAATY, T.L. (2001) Decision Making for Leaders, new edn (Pittsburgh, PA, RWS Publications). SAATY, T.L. & VARGAS, L.G. (1994) Decision Making in Economic, Political, Social, and Technological Environments with the Analytic Process (Pittsburgh, PA, RWS Publications). SHIEH, Y.W. (2000) The interactive supply-demand model for supply chain in electronic commerce, Doctoral Dissertation, National Chengchi University, Taipei. SLACK, B. (1985) Containerization, inter-port competition and port selection, Maritime Policy and Management, 12, (4), pp. 293–303. SOHN, K.Y., YANG, J.W. & KANG, C.S. (2001) Assimilation of public opinions in nuclear decisionmaking using risk perception, Annals of Nuclear Energy, 28, (6), pp. 553–563. SUH, B. & HAN, I. (2003) The IS risk analysis based on a business model, Information and Management, 41 (2), pp. 149–158. TAM, M.C.Y. & TUMMALA, V.M.R. (2001) An application of the AHP in vendor selection of a telecommunications system, Omega-The International Journal of Management Science, 29, (2), pp. 171–182. TAMVAKIS, M. & THANOPOULOU, H.A. (2000) Does quality pay? The case of the dry bulk market, Transportation Research Part E: Logistics and Transportation Review, 36, pp. 297–307. THANOPOULOU, H.A., RYOO, D.K. & LEE, T.W. (1999) Korean liner shipping in the era of global alliances, Maritime Policy and Management, 26, (3), pp. 209–229. THOMAS, B.J. (1998) Structure changes in the maritime industry’s impact on the inter-port competition in container trade, Proceedings of the International Conference on Shipping Development and Port Management, Kaohsiung, 26–29 March. TRIANTAPHYLLOU, E. & LIN, C.T. (1995) Development and evaluation of ?ve fuzzy multiattribute decision-making methods, International Journal of Approximate Reasoning, 14, pp. 281–310. TSAI, M.C. & SU, Y.S. (2002) Political risk assessment on air logistics hub developments in Taiwan, Journal of Air Transport Management, 8, (6), pp. 373–380. TSAUR, S.H., CHANG, T.Y. & YEN, C.H. (2002) The evaluation of airline service quality by fuzzy MCDM, Tourism Management, 23, (2), pp. 107–115. TSAUR, S.H., TZENG, G.H. & WANG, K.C. (1997) Evaluating tourist risks from fuzzy perspective, Annals of Tourism Research, 24, (4), pp. 796–812. TZENG, G.H. & WANG, R.T. (1994) Application of AHP and fuzzy MADM to evaluation of bus system performance in Taipei City, Third International Symposium on the Analytic Hierarchy Process, George Washington University, Washington, DC, 11–13 July. VARGAS, L.G. (1997) Why the multiplicative AHP is invalid: A practical example, Journal of Multicriteria Decision Analysis, 6, (3), pp. 169–170. VILLALON, W. (1998) Smarter beats bigger, World Economic Development Congress 1998, Transportation Infrastructure Summit, excerpted and edited by the Journal of Commerce, available at http:==www.containtheport.com=contain=info=jocv.htm VREEKER, R., NIJKAMP, P. & WELLE, C.T. (2002) A multicriteria decision support methodology for evaluating airport expansion plans, Transportation Research Part D, 7, pp. 27–47.
244 T.-C. Lirn et al.
WIND, Y. & SAATY, T.L. (1980) Marketing applications of the analytic hierarchy process, Management Science, 26, (7), pp. 641–658. WINEBRAKE, J.J. & CRESWICK, B.P. (2003) The future of hydrogen fuelling systems for transportation: An application of perspective-based scenario analysis using the analytic hierarchy process, Technological Forecasting and Social Change, 70, (4), pp. 359–384. ZIMMERMANN, H.J. (1991) Fuzzy Set Theory and its Application (Boston, USA: Kluwer Academic).
doc_619218264.pdf
The report about the analytic hierarchy process (AHP) technique to determine the importance of various criteria in the transhipment port selection decision-making process.
International Journal of Logistics: Research and Applications Vol. 6, No. 4, 2003
Transhipment Port Selection and Decision-making Behaviour: Analysing the Taiwanese Case
TAIH-CHERNG LIRN,* HELEN A. THANOPOULOU & ANTHONY K. C. BERESFORD
Logistics and Operations Management Section, Cardiff Business School, Cardiff University, UK
ABSTRACT
The study uses the analytic hierarchy process (AHP) technique to determine the importance of various criteria in the transhipment port selection decision-making process. The authors propose a set of transhipment port selection criteria from a container carrier’s perspective. Sourcing the data from an AHP survey in Taiwan, transhipment port selection is found, tenuously at this ?rst stage of research, to depend mainly on port competitiveness as represented by the cost that carriers are faced with for loading and discharging of containers and on port ef?ciency as represented by the container loading and discharging rates. Fuzzy multiple criteria decision-making methodology (FMCDM) was applied to obtain evaluations of port alternatives and the relevant responses defuzzi?ed to derive crisp values of port performance.
Introduction The paper addresses transhipment port selection from a carrier’s perspective. Decision-making behaviour is becoming critical in order to compete in the liner container market as the market itself is under increasing competitive pressures. In
*Correspondence: Taih-Cherng Lirn, Logistics and Operations Management Section, Cardiff Business School, Cardiff University, Aberconway Building, Colum Drive, Cardiff CF10 3EU, UK, E-mail: [email protected] International Journal of Logistics ISSN 1367-5567 Print=ISSN 1469-848X online # 2003 Taylor & Francis Ltd http:==www.tandf.co.uk=journals DOI: 10.1080=13675560310001626990
230 T.-C. Lirn et al.
the 21st century major container carriers are competing globally for cargoes belonging to global shippers; this requires the provision of a global network by seacontainer carriers. The requirements of shippers who combine resources and distribute products globally are global coverage (i.e. coverage of the main trade routes), high frequency and speed, while maintaining the lowest possible cost and highest possible quality of service. This already ambitious target for today’s liner carriers is also combined with the requirements of shippers for the provision of more than just the shipping link of the transport and distribution chain. In this context, global carriers are not only required to provide the organisation of intermodal transport, but also the provision of total logistics often including added value services as discussed in Thanopoulou et al. (1999) and in Ryoo & Thanopoulou (1999). Amidst these pressures, port selection becomes a key element not only for achieving a cost advantage, but also for achieving product differentiation to the extent that differences in the quality of service are recorded as signi?cant. This is likely to be due to the demand of quality-sensitive clients such as shippers of valuable consumer and intermediate goods. This is unlike bulk shipping where research has indicated that, despite some level of nominal differences, there has been only marginal preference for a higher quality of service as highlighted by Tamvakis & Thanopoulou (2000). The liner container industry is different not just in this respect; it has also undergone major changes in terms of the world distribution of container capacity. Asia has risen to become a major power in liner shipping as well as in the world container trades. Therefore, the authors also aim to highlight more general trends in modern liner shipping. This research focuses on the transhipment decision-making behaviour of large Taiwanese international container carriers, which include some of the largest and most prominent international liner operators world-wide. Decision-making behaviour is particularly important not only for international container operators, but also for container port authorities and terminal operating companies. The increasing concentration within the industry has increased the potential impact of a move by a major port user on the individual port’s overall traf?c. The research approaches transhipment port selection decision-making by focusing on data sourced from a ?eld survey conducted in Taiwan by the ?rst-named author and using analytic hierarchy process (AHP) methodology. The main purpose of the research has been to identify the importance of factors that in?uence transhipment port choice by Taiwanese ocean container carriers. Identifying the criteria was an exercise based partly on a review of existing literature and on the results of the Delphi technique, which was applied to validate general port selection criteria in a transhipment context. The relative importance of the major criteria and sub-criteria revealed by the research indicates the directions to be explored by port operators in their effort not only to increase their own competitiveness vis-a-vis other ports, but also to facilitate transhipment activity by achieving lower costs, becoming more user-orientated and optimising resource allocation, especially as liner shipping networks are bound to expand and become more dense as globalisation intensi?es.
Background to Transhipment Port Selection Many container ports are expanding berth and loading=discharging facilities in order to meet the anticipated container throughput increase in the new century.
Transhipment Port Selection
231
FIGURE 1. Top World Container Ports by Throughput. Source: Compiled from Data in Different Issues of Containerisation International Yearbook 1999–2003.
Container traf?c has increased dramatically in the last 20 years. In the early 1980s, sea-container traf?c was just 1.56 million TEUs for Hong Kong and a little over 1 million TEUs for Singapore and Kaohsiung (Containerisation International, 1982, p. 77). By 1986, container throughput had increased to 2.78 million TEUs for Kaohsiung, 2.77 million TEUs for Hong Kong and 2.20 million TEUs for Singapore (Containerisation International Yearbook, 1987). Over the following 15 years container traf?c through Singapore and Hong Kong rose to 17.90 and 15.57 million TEUs, respectively (Containerisation International Yearbook, 2002). According to data for the year 2001, container throughput increases have been spectacular for the whole of the East and South Asian regions but the growth has not been evenly spread. Container throughput for Kaohsiung in 2001, at only 7.54 million TEUs, was well below Singapore’s and Hong Kong’s volumes. Although both the absolute ?gures and the respective rates of increase from the mid-1980s are impressive, they have to be seen in the context of port competition in that area. Kaohsiung was still ranked fourth among top container ports world-wide at the turn of the century, but the margin separating it from the top three ports has increased as its container traf?c is now less than half the volume of its major competitors (Figure 1). With competition between the different ports increasing, the risk of overinvestment in port facilities might become a serious problem in some of the major Asian ports. Knowing the preferences of carriers for transhipment ports and knowing how these preferences are formed becomes critical. The recent moves of major carriers between Asian ports and the increasing importance of individual global carriers in the traf?c of individual ports emphasise the importance of transhipment port selection.
Literature Review and Selection of Methodology The research adopted AHP to reveal the importance of transhipment port selection criteria in transhipment decision-making. The AHP methodology is a multiple criteria decision-making (MCDM) technique that provides the relative importance of factors involved in decisions through pairwise comparisons of the same level independent criteria. Pairwise comparisons are obtained through surveying either the whole population or a sample of decision-makers. The hierarchical structure of
232 T.-C. Lirn et al.
levels of criteria involved in the survey is critical for the successful application of AHP (Lai et al., 2002) as is the number of criteria adopted at each level. It has been observed that when ?ve or more items are involved in the survey questionnaire, the decision-maker (DM) statements regarding pairwise comparisons tend to become increasingly inconsistent in terms of transitivity (Bodin & Gass, 2003; Gass, 1998). Clearly structuring and de?ning the number of criteria is a critical preliminary stage of an AHP survey. In the literature, it has been suggested that de?ning independent criteria, and at the same time limiting their numbers at each level of an AHP hierarchy, can be achieved through using factor analysis (Park & Han, 2002; Sohn et al., 2001), or the Delphi technique (Azani & Khorramshahgol, 1990; Suh & Han, 2003) and=or brainstorming (Al-Harbi & Al-Subhi, 2001; Dey, 1999). The research for this paper adopted the Delphi technique to validate general port selection criteria in a transhipment context and then brainstorming to limit the number of criteria to a level acceptable for use in the ?nal AHP survey as most average values remained ‘‘high’’ to ‘‘very high’’ through both Delphi rounds. While factor analysis is only appropriate if a large enough number of questionnaire replies can be obtained, usually ?ve times larger than the number of factors included (Hair, 1995), the Delphi technique remains suitable even where only a small number of experts are involved in the process of inclusion=elimination of criteria through repeated surveying rounds.
The Analytic Hierarchy Process The AHP methodology, ?rst developed by Thomas Saaty in 1971, has been widely used since its ?rst appearance in the Journal of Mathematical Psychology in 1977 (Saaty, 1977). The methodology falls into the category of MCDM approaches. As pointed out by Vreeker et al. (2002), the basic rules for solving multi-level hierarchical problems involve essentially four steps: (1) speci?cation of choice problem; (2) information analysis; (3) choice of appropriate method; and (4) evaluation of alternatives. As discussed in Saaty (1977, 1980) and more recently in Tam & Tummala (2001), the AHP modelling process involves four phases: (1) Structuring: formulating the AHP hierarchy in terms of objective, major criteria and sub-criteria, and rating scale used for the evaluation and for decisionalternatives to be evaluated. (2) Data collection: for this research a sample of Taiwanese container carriers was surveyed in 2001. (3) Obtaining the weights in different hierarchies: as in most of the literature known to the authors, the present paper has followed the geometric mean approach instead of the arithmetic one to improve results (Forman, 2000). The general form of the matrix of ratio comparisons of the AHP is described in Saaty (1977) and in Saaty & Vargas (1994). (4) Synthesis: in the case of the research presented in this paper, the ?nal step has been to propose a solution to transhipment port selection on the basis of AHP sub-criteria weights and the performance scores of transhipment port alternatives. The scope for successful applications of AHP has proved to be exceptionally wideranging, from the selection of fuel source for transportation in the future in
Transhipment Port Selection
233
Winebrake & Creswick (2003) to the evaluation of the top 20 decision-making support system journals (Forgionne et al., 2002). The limitations and drawbacks of AHP have been debated in several journals. One of the major criticisms of AHP is that it allows ‘‘rank reversals’’, where the order of the alternatives changes when any of them is added or deleted. This thus violates the invariance principle of utility theory. Dyer (1990a, b), Harker & Luis (1990) and Vargas (1997) have debated the validity of AHP, while some major criticisms of AHP have been successfully defended by Saaty (1990a, b, 2000). One of the problems AHP had to solve concerned the reliability of values attributed to the pairwise comparisons by participants in the surveys. Beynon (2002) found that under certain conditions alternative scales were following linguistic scales more effectively than Saaty’s original one- to nine-point scale, especially in relation to the symmetry of argument. Hurley (2001) reported that decision-makers are frequently certain about the rank order of the objects for a particular pairwise comparison matrix, but uncertain about the precise numerical weights that the AHP produces for that matrix. However, the applicability of AHP to a given data set can be enhanced through the application of fuzzy techniques. Triantaphyllou & Lin (1995) concluded that fuzzy AHP approaches are capable of capturing a human’s appraisal of ambiguity where complex decision-making problems are concerned. The scope for applying fuzzy techniques within AHP can have quite concrete applications. Tsaur et al. (1997) used fuzzy AHP to evaluate tourist risks for ?ve destinations, Kuo et al. (1999) proposed a decision support system for locating convenience stores through fuzzy AHP, while Ravi & Reddy (1999) adopted fuzzy AHP in order to rank the suitability of Indian coking and non-coking coals for industrial use.
Applying AHP to Transhipment Port Selection There are at least 20 known applications of AHP on transport problems, such as, for example Vreeker et al. (2002), Chang & Yeh (2001), Poh & Ang (1999), Tzeng & Wang (1994) and Frankel (1992). Within the wider transport sector, AHP continues to be used as a valid approach and this applies to many recent articles focusing on, or emanating from, Taiwan. Tsai & Su (2002) utilised AHP to assess the political risk attached to the development of an air logistics hub in Taiwan. Chou & Liang (2001) used AHP and proposed the use of linguistic values in order to create a model capable of evaluating the performance of shipping companies; the fuzzy approach they adopted has been considered as capable of addressing the problem of subjective=imprecise assessments related to the performance evaluation problem. The port performance evaluation problem is not very different in this respect as it involves subjective evaluations of the different attributes of ports that play a role in the port selection decision-making process. Fuzzy set theory was developed from the premise that the key elements in human thinking are not numbers, but linguistic terms or labels of fuzzy information sets (Zimmermann, 1991). The adoption of a fuzzy approach has been deemed an appropriate methodology for a ?rst approach to the transhipment port decisionmaking process in order to test the real extent of problems that assigning values could create in the context of wider research. Therefore, semantic words have been adopted to evaluate the performance of three major Taiwanese international ports
234 T.-C. Lirn et al.
and evaluators were requested to provide a triangular fuzzy number (TFN) within the scale range of 0–100 for each different semantic word used. Defuzzi?cation was then applied to convert the fuzzy set obtained into a real number. The centre of area (COA) method is the most widely used technique for computing defuzzi?ed values (Patel & Mohan, 2002), the mean of maximum method and the median rule being the two main alternative techniques used to defuzzify the initial fuzzy sets obtained. The COA method determines the centre area of membership functions. Equation (1) (Tsaur et al., 1997, 2002) was used in order to convert the fuzzy set of numbers denoted by Ri to a crisp value denoted as BNPi, the best non-fuzzy performance value of alternative i. BNP¼ i ½(URÀ LRiþ (MRÀ LRi Þ )] i i þ LRi : 3 (1 )
In equation (1), BNPi represents the best non-fuzzy performance value, and URi, MRi, and LRi are the components of a fuzzy number Ri that is the result of a fuzzy multiplication. MRi is a product of the representative value of the highest grade of membership in a triangular fuzzy set multiplied by the weight obtained from AHP. The terms LRi and URi denote the products of the representative values of the lower boundary and upper boundary of a triangular fuzzy set, where the membership is lowest, multiplied by the weight obtained from AHP.
Deriving AHP Criteria The starting point of this preparatory phase of the AHP survey used in this research was the identi?cation of relevant criteria through a literature review. While there is limited literature on transhipment port choice, factors affecting port choice in general and factors affecting the general competitiveness of container ports are well documented in the literature. Combining criteria declared in recent transhipment port selection cases (Porcari, 1999; Villalon, 1998) and port choice criteria used in academic literature gave an overall total of 47 port selection criteria (see Table 1). The large number of registered criteria made the use of factor analysis at this preliminary stage prohibitive as the number of experts to be surveyed would have to be at least two to three times larger than the 47 criteria found. An added limitation would be that experts involved in the hierarchical structuring task through an eventual factor analysis and those being surveyed in the AHP survey must not overlap. As there are no minimum limits on expert numbers involved in the Delphi technique, the non-overlap rule was easy to follow. The Delphi technique using ?ve-point Likert scales was used at a ?rst stage as the basis for validating criteria and at a second stage as the basis for fusing the large number of criteria identi?ed from the literature review through brainstorming. The criteria were presented to a total of 10 experts, senior industry executives and scholars in the area of shipping, during a number of interviews and evaluation rounds conducted in Taiwan. The initial large number of criteria was reduced by being further classi?ed by the surveying author into six major criteria: (1) Water depth of the port. (2) Local infrastructure and port transportation network. (3) Area of the container yard and of marshalling yard.
TABLE 1. Factors Affecting the Carriers’ Port Choice Behaviour
Writer
Criteria Villalon (1998) Hayuth (1995) Slack (1985) Baird (2000) Porcari (1999) Burnson (2001) Jansson & Shneers (1987) Thomas (1998) 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 (continued) 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Brooks (2000) Frankel (2001) Fleming & Baird (1999) Murphy et al. (1989) Browne et al. (1989) Branch (1986)
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.
Available number of berths Back-up space on terminal Congestion Cargo volume Cargo-generating effect Other modes Competitiveness Container handling ef?ciency Containerised cargo proportion Degree of integration (EDI) Depth of the port Geographical advantage Flexible operation process Free time Frequency of feeder service Frequent port of call Infrastructure Inland freight rates Labour problems Loading=discharging rate Low cost Major container centre Numbers of sailing Operation Related business operation Port accessibility (land and sea) Port operation=working hours Port berthing time length Port charges Price conditions Port equipment Port security Port service coverage Port tradition and organisation Privilege contract to carrier Port productivity
Transhipment Port Selection 235
236 T.-C. Lirn et al.
TABLE 1. Continued
Writer
Criteria Villalon (1998) Hayuth (1995) Slack (1985) 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Baird (2000) Porcari (1999) Burnson (2001) Jansson & Shneers (1987) Thomas (1998) Brooks (2000) Frankel (2001) Fleming & Baird (1999) Murphy et al. (1989) Browne et al. (1989) Branch (1986)
35. Proximity to alternative loading centre 36. Quality of customs handling 37. Regulations 38. Service considerations 39. Size of hinterland 40. Size of port=terminal 41. State aides and in?uence on cost 42. Superstructure 43. Intermodal link=network 44. Transportation and port-user cost 45. Time on the route 46. Transit time 47. Trade inertia
Transhipment Port Selection
237
TABLE 2. Qualitative Categorising 47 Criteria into Four Major Criteria and 16 Sub-criteria
Notes: A1, A2, mode quantity; B1, B2, average value; C1, C2, standard deviation; D, mode value; E, mode quantity difference between round one and round two surveys ( ¼ A1 7A2); CAT, category; F, standard deviation reduction value ( ¼ C1 7C2); No., numbering of CAT. III sub-criteria; C.Y., container yard. The Delphi panel is composed of 10 experts, ?ve persons are selected from academia and ?ve persons are from the shipping industry in Taiwan. Criteria in the literature for general port selection are all validated for transhipment ports with high average scores by the Delphi technique. CAT. IV are variables derived from literature review, CAT. III and CAT. II are allocated by corresponding author, CAT. I is categorised by consultation with two academics in Taiwan Ocean University.
238 T.-C. Lirn et al.
(4) Basic cargo volume (non-transhipment cargo). (5) Geographical advantage: location near the main service routes. (6) Port ef?ciency and cost of container-handling for carriers. This list was further shortened through brainstorming sessions with academics in Taiwan to derive the four major criteria that were used in the survey questionnaire along with 16 sub-criteria (see Table 2. The major criteria and their subdivision into relevant sets of sub-criteria are the two system components of the AHP (see Figure 2).
FIGURE 2. Determining Optimum Transhipment Port for Major Taiwan Container Carriers using AHP.
Transhipment Port Selection
239
The AHP Survey and Results A ?eld survey was conducted among four major international container carriers in Taiwan. A questionnaire was presented to the users in face-to-face meetings. In the ?rst part, the respondents conducted pairwise comparisons of the importance of various transhipping port selection criteria. In the second part, the evaluators adopted ?ve linguistic variables, namely excellent, good, average, fair and poor (Tsaur et al., 1997), to evaluate port performance on the basis of the 16 sub-criteria. Non-fuzzy performance measures were then derived using equation (1) to arrive at a crisp value of performance for each of the three major international ports in Taiwan based on the carrier evaluations. Carriers are identi?ed in this paper with the sequential letters A, B, C and D to avoid any breach of anonymity. The results show that the different participants in the survey attribute varying importance to the major criteria for transhipment port selection. Both A and B carriers are mostly concerned with ‘‘port management’’. B carrier also attributed signi?cant importance to ‘‘port geographical location’’. The C and D carriers were found to be concerned mostly with ‘‘port basic physical characteristics’’ and ‘‘carriers cost perspective’’, respectively (see Table 3). In terms of the 16 sub-criteria used (see Figure 2), there were also differences among the respondents. When the global weights of the criteria are considered as an average for all respondents, the results indicate that all four major criteria have almost the same weight in in?uencing transhipment port selection behaviour. In terms of subcriteria, the companies surveyed seem to attribute most weight to ‘‘proximity of the feeder ports’’ (global G.M. weight 9.06%), ‘‘berthing delay and loading=discharging rate’’ (global G.M. weight 10.84%) and ‘‘carriers loading=discharging cost’’ (global G.M. weight 11.59%), as shown in Table 4. Performance of Taiwanese Ports By using semantic words, the carriers evaluated the performance of the three major international ports according to the sub-criteria for transhipment port selection. Equation (1) was then applied to calculate the BNP values. The survey responses also reveal the limits of the methodology applied for performance evaluation purposes as one of the carriers did not proceed to evaluate the overall performance of the three ports involved, possibly in order to avoid offending port terminal operators. The performance evaluation by the remaining respondents resulted in port X being consistently ranked ?rst (Table 5). The overall results can also be deemed as
TABLE 3. Importance of Each Major Criterion Revealed Through the AHP Survey Carrier A (%)
Port basic physical characteristics Port geographical location Port management Carriers’ cost perspective 8.48 5.99 42.77 42.77
Carrier B (%)
5.89 45.58 42.64 5.89
Carrier C (%)
69.16 13.97 9.88 6.99
Carrier D (%)
8.04 39.81 5.62 46.53
240 T.-C. Lirn et al.
TABLE 4. Global Weights of Transhipment Port Selection Criteria Major Criteria (First-tier Factors)
Port basic physical characteristics
Major Criteria: Global Weight (%)
22.89
Sub-criteria (Second-tier Factors)
Infrastructure improvement Port facilities and equipment Convenience of inter-modal link Size of marshalling yard and container yard The closeness to the import= export consumption areas The closeness to the main navigation route Proximity of the feeder ports Proximity of competing ports and modes Port administration and custom regulation Berthing delay and loading= discharging rate Port safety Terminal security Carrier’s loading=discharging cost Ownership of the port and terminal Privileged terms to the carriers Government levy and duty
Sub-criteria: Global Weight (%)
1.55 7.48 8.72 5.14 8.07 5.43 9.06 3.78 3.85 10.84 9.01 1.53 11.59 8.15 3.65 2.15
Port geographical location
26.34
Port management
25.23
Carriers’ cost perspective*
25.54
Note: Italic characters indicate the most important criterion and sub-criterion as revealed by AHP.
objectively consistent. The prominence of the carrier’s cost perspective and of the loading and discharging cost among the sub-criteria was con?rmed by the comparison of port rankings through the survey and the cost ranking of ports involved. Following a similar approach in the case of airlines (Chang & Yeh, 2001), actual port loading and discharging cost data were used to substitute for the scores obtained through the AHP performance evaluation; this resulted in the same ranking of the three major international ports in Taiwan as the one derived through the AHP survey. However, the comparison was made on the basis of actual data relevant only to one of the sub-criteria and should be taken into account only as such.
TABLE 5. BNP Values Calculated from the Evaluation of Three Major Ports in Taiwan Port
Carriers Carrier A Carrier B Carrier C Carrier D Total ( ¼ B þ C þ D) Port X NA 90.50 77.40 88.82 256.72 Port Y NA 81.72 75.21 73.84 230.77 Port Z NA 77.38 74.70 70.77 222.85
Notes: BNP represents best non-fuzzy performance; NA, not available.
Transhipment Port Selection
241
Conclusions, Limitations and Further Research The set of data used for this research was limited as carriers surveyed were from one country and this is acknowledged as a signi?cant limitation. However, the small sample surveyed included all global carriers and regional ones. All ports and carriers surveyed were in and from Taiwan, respectively. This allowed the research to remain free from considerations regarding regional pressures, which could eventually affect port choice. If a broader regional study were to be carried out, the political risk factor, which has not been included to date in port selection literature, could eventually become a more in?uential attribute, although that would depend on the de?nition of the geographical range of the region considered. In any case, the ?ndings of the research are consistent with the feeling of the market on the prevalence of cost as a major criterion from a carrier’s perspective. The initial ?ndings are also consistent with informally expressed views from the industry (personal interview with a senior executive of a global carrier in autumn 2002), while the initial results of an ongoing research on global carriers seem to indicate the prevalent signi?cance of the handling cost among the criteria used in the survey (Lirn et al., 2003). There is also scope for improving the methodology adopted in the future. Factor analysis can be used to provide an alternative approach for narrowing down the number of proposed criteria and a more objective method for distinguishing between criteria and sub-criteria. Furthermore, the fuzzy performance evaluation by semantic language was found to be dif?cult by most interviewees taking part in this research and it required face-to-face explanations before the interviewees felt able to provide a score range for their related semantic evaluations. This problem has also been reported by Shieh (2000) and seems to indicate that broader research through questionnaires can only be undertaken through a non-fuzzy AHP in order to increase the response rate. Another important conclusion emanating from the survey itself was that of the familiarity of interviewees with the basics of the AHP methodology. The survey indicated that if the basics of AHP had not been explained to the interviewees, it was very easy for them to answer all questions but their answers could be seriously inconsistent, with eventual repercussions for the validity of the results (Saaty, 2001). However, if the AHP methodology is explained to the interviewees, the value of the consistency index reduces signi?cantly and results are much more reliable. The major drawback of providing information on AHP in person is that of the potential introduction of a bias, and the solution seems to lie in providing summary information on AHP through the questionnaire itself. Research which two of the authors are currently conducting on transhipment port selection by major global carriers has adopted this approach, taking account of other limitations that have arisen in this research related to the uniformity of language expressions and the large number of sub-criteria. Finally, AHP can also provide the basis for required updates on the transhipment port selection problem. As no carrier operates in isolation, decisionmaking is bound to change following shifts in the environment in which carriers operate. A systematic update of the research would reveal evolving trends.
Acknowledgements The authors would like to thank two anonymous referees for helpful and constructive comments as well as the editors of the issue. Thanks are expressed to
242 T.-C. Lirn et al.
Dr Malcolm Beynon for comments in the early stages of the writing of the paper. Professor S. Yahalom and Mr P. Chou from the National Taiwan Ocean University provided valuable comments to the ?rst author in the initial stages of this research.
REFERENCES
AL-HARBI, K. & AL-SUBHI, M. (2001) Application of the AHP in project management, International Journal of Project Management, 19, (1), pp. 19–27. AZANI, H. & KHORRAMSHAHGOL, R. (1990) Analytic Delphi Method (ADM): A strategic decisionmaking model applied to location planning, Engineering Costs and Production Economics, 20, (1), pp. 23–28. BAIRD, A.J. (2000) Port privatisation: Objectives, extent, process and the UK experience, International Journal of Maritime Economics, 2, (3), pp. 177–194. BEYNON, M. (2002) An analysis of distributions of priority values from alternative comparison scales within AHP, European Journal of Operational Research, 140, (1), pp. 104–117. BODIN, L. & GASS, S.I. (2003) On teaching the analytic hierarchy process, Computers and Operations Research, 30, (10), pp. 1487–1497. BRANCH, A.E. (1986) Elements of Port Operation And Management (London, Chapman and Hall). BROOKS, M. (2000) Sea Change in Liner Shipping—Regulation and Managerial Decision-making in a Global Industry (Oxford, Elsevier Science). BROWNE, M., DOGANIS, R. & BERGSTRAND, S. (1989) Transhipment of UK Trade (London, British Ports Federation). BURNSON, P. (2001) Sea change, Global Logistics, 1 July, LexisNexis (online). CHANG, Y.H. & YEH, C.H. (2001) Evaluating airline competitiveness using multiattribute decisionmaking, Omega, 29, (5), pp. 405–415. CHOU, T.Y. & LIANG, G.S. (2001) Application of a fuzzy multi-criteria decision-making model for shipping company performance evaluation, Maritime Policy and Management, 28, (4), pp. 375–392. Containerisation International (1982) Informa Maritime and Transport, Essex, June. Containerisation International Yearbook (1987, 1999–2003) Informa Maritime and Transport, Essex. DEY, U.K. (1999) Process re-engineering for effective implementation of projects, International Journal of Project Management, 17, (3), pp. 147–159. DYER, S.J. (1990a) Remarks on the analytic hierarchy process, Management Science, 36, pp. 249–258. DYER, S.J. (1990b) A clari?cation of remarks on the analytic hierarchy process, Management Science, 36, pp. 274–275. FLEMING, D.K. & BAIRD, A.J. (1999) Some re?ections on port competition in the United States and Western Europe, Maritime Policy and Management, 26, (4), pp. 383–394. FORGIONNE, G.A., KOHLIB, R. & JENNINGS, D. (2002) An AHP analysis of quality in AI and DSS journals, Omega, 30, (3), pp. 171–183. FORMAN, E.H. (2000) Decision by Objectives (Washington, George Washington University), available at http:==www.expertchoice.com=dbo= FRANKEL, E.G. (1992) Hierarchical logic in shipping policy and decision-making, Maritime Policy and Management, 19, (3), pp. 211–221. FRANKEL, E.G. (2001) Economics of transhipment in container shipping logistics, Inaugural International Conference on Port and Maritime R and D and Technology, Singapore, pp. 29–31. GASS, S.I. (1998) Tournaments, transitivity and pairwise comparison matrices, Journal of the Operational Research Society, 49, pp. 616–624. HAIR, J.F. (1995) Multivariate Data Analysis (Englewood Cliffs, NJ, Prentice Hall). HARKER, T.P. & LUIS, G.V. (1990) Reply to remarks on the analytic hierarchy process, Management Science, 36, pp. 269–273. HAYUTH, Y. (1995) Container traf?c in ocean shipping policy, Proceedings of International Conference ‘‘Ports for Europe’’ Conference, Europacollege, Zeehaven Brugge, 23–24 November. HURLEY, W.J. (2001) The analytic hierarchy process: A note on an approach to sensitivity which preserves rank order, Computers and Operations Research, 28, pp. 185–188. JANSSON, J.O. & SHNEERS, D. (1987) Liner Shipping Economics (New York, Chapman and Hall). KUO, R.J., CHI, S.C. & KAO, S.S. (1999) A decision support system for locating convenience store through fuzzy AHP, Computers and Industrial Engineering, 37, (1–2), pp. 323–326. LAI, V.S., WONG, B.K. & CHEUNG, W. (2002) Group decision making in a multiple criteria environment: A case using the AHP in software selection, European Journal of Operational Research, 137, (1), pp. 134–144. LIRN, T.C., THANOPOULOU, H.A. & BEYNON, M.J. (2003) Applying AHP on Transhipment Port Selection: A Global Perspective, (Mimeo, Cardiff Business School).
Transhipment Port Selection
243
MURPHY, P.R., DALENBURG, D.R. & DALEY, J.M. (1989) Assessing international port operations, International Journal of Physical Distribution and Logistics Management, 19, (9), pp. 3–10. PARK, C.-S. & HAN, I. (2002) A case-based reasoning with the feature weights derived by analytic hierarchy process for bankruptcy prediction, Expert Systems with Applications, 23, (3), pp. 255–264. PATEL, A.V. & MOHAN, B.M. (2002) Some numerical aspects of centre of area defuzzi?cation method, Fuzzy Sets and Systems, 132, (3), pp. 401–409. POH, K.L. & ANG, B.W. (1999) Transportation fuels and policy for Singapore: An AHP planning approach, Computers and Industrial Engineering, 37, pp. 507–525. PORCARI, J.D. (1999) The logical choice for a Northeast hub, Journal of Commerce, 30 April 30, Lexis Nexis (online). RAVI, V. & REDDY, P.J. (1999) Ranking of Indian coals via fuzzy multi attribute decision-making, Fuzzy Sets and Systems, 103, pp. 369–377. RYOO, D.K. & THANOPOULOU, H.A. (1999), Liner alliances in the globalization era: A strategic tool for Asian container carriers, Maritime Policy and Management, 26, (4), pp. 349–367. SAATY, T.L. (1977) A scaling method for priorities in hierarchical structures, Journal of Mathmatical Psychology, 15, pp. 234–281. SAATY, T.L. (1980) The Analytic Hierarchy Process (New York, McGraw-Hill). SAATY, T.L. (1990a) An exposition of the AHP in reply to the paper remarks on the analytic hierarchy process, Management Science, 36, pp. 259–268. SAATY, T.L. (1990b) How to make a decision: The analytic hierarchy process, Interfaces, 24, (6), pp. 19–43. S AATY, T.L. (2000) On the relativity of relative measure—accommodating both rank preservation and rank reversals in the AHP, European Journal of Operational Research, 121, pp. 205–212. SAATY, T.L. (2001) Decision Making for Leaders, new edn (Pittsburgh, PA, RWS Publications). SAATY, T.L. & VARGAS, L.G. (1994) Decision Making in Economic, Political, Social, and Technological Environments with the Analytic Process (Pittsburgh, PA, RWS Publications). SHIEH, Y.W. (2000) The interactive supply-demand model for supply chain in electronic commerce, Doctoral Dissertation, National Chengchi University, Taipei. SLACK, B. (1985) Containerization, inter-port competition and port selection, Maritime Policy and Management, 12, (4), pp. 293–303. SOHN, K.Y., YANG, J.W. & KANG, C.S. (2001) Assimilation of public opinions in nuclear decisionmaking using risk perception, Annals of Nuclear Energy, 28, (6), pp. 553–563. SUH, B. & HAN, I. (2003) The IS risk analysis based on a business model, Information and Management, 41 (2), pp. 149–158. TAM, M.C.Y. & TUMMALA, V.M.R. (2001) An application of the AHP in vendor selection of a telecommunications system, Omega-The International Journal of Management Science, 29, (2), pp. 171–182. TAMVAKIS, M. & THANOPOULOU, H.A. (2000) Does quality pay? The case of the dry bulk market, Transportation Research Part E: Logistics and Transportation Review, 36, pp. 297–307. THANOPOULOU, H.A., RYOO, D.K. & LEE, T.W. (1999) Korean liner shipping in the era of global alliances, Maritime Policy and Management, 26, (3), pp. 209–229. THOMAS, B.J. (1998) Structure changes in the maritime industry’s impact on the inter-port competition in container trade, Proceedings of the International Conference on Shipping Development and Port Management, Kaohsiung, 26–29 March. TRIANTAPHYLLOU, E. & LIN, C.T. (1995) Development and evaluation of ?ve fuzzy multiattribute decision-making methods, International Journal of Approximate Reasoning, 14, pp. 281–310. TSAI, M.C. & SU, Y.S. (2002) Political risk assessment on air logistics hub developments in Taiwan, Journal of Air Transport Management, 8, (6), pp. 373–380. TSAUR, S.H., CHANG, T.Y. & YEN, C.H. (2002) The evaluation of airline service quality by fuzzy MCDM, Tourism Management, 23, (2), pp. 107–115. TSAUR, S.H., TZENG, G.H. & WANG, K.C. (1997) Evaluating tourist risks from fuzzy perspective, Annals of Tourism Research, 24, (4), pp. 796–812. TZENG, G.H. & WANG, R.T. (1994) Application of AHP and fuzzy MADM to evaluation of bus system performance in Taipei City, Third International Symposium on the Analytic Hierarchy Process, George Washington University, Washington, DC, 11–13 July. VARGAS, L.G. (1997) Why the multiplicative AHP is invalid: A practical example, Journal of Multicriteria Decision Analysis, 6, (3), pp. 169–170. VILLALON, W. (1998) Smarter beats bigger, World Economic Development Congress 1998, Transportation Infrastructure Summit, excerpted and edited by the Journal of Commerce, available at http:==www.containtheport.com=contain=info=jocv.htm VREEKER, R., NIJKAMP, P. & WELLE, C.T. (2002) A multicriteria decision support methodology for evaluating airport expansion plans, Transportation Research Part D, 7, pp. 27–47.
244 T.-C. Lirn et al.
WIND, Y. & SAATY, T.L. (1980) Marketing applications of the analytic hierarchy process, Management Science, 26, (7), pp. 641–658. WINEBRAKE, J.J. & CRESWICK, B.P. (2003) The future of hydrogen fuelling systems for transportation: An application of perspective-based scenario analysis using the analytic hierarchy process, Technological Forecasting and Social Change, 70, (4), pp. 359–384. ZIMMERMANN, H.J. (1991) Fuzzy Set Theory and its Application (Boston, USA: Kluwer Academic).
doc_619218264.pdf