INTERNATIONAL JOURNAL OF PROJECT MANAGEMENT International Journal of Project Management 23 (2005) 483–491 www.elsevier.com/locate/ijproman

Evaluating project teaming strategies for construction of 101 using resource-based theory

Ren-Jye Dzeng *, Kuo-Sheng Wen 1

Department of Civil Engineering, National Chiao-Tung University, 1001 Ta-Hsieu Road, Hsinchu 30050, Taiwan

Received 18 May 2004; received in revised form 5 November 2004; accepted 25 February 2005

Abstract

Taipei 101, the tallest building in the world in 2004, is a super-large project that required many resources, exceeding a single con- tractorÕs capacity, and the project owner had to seek additional specialist contractors for the resource gaps. This paper presents an analytical model for the ownerÕs teaming strategies on the inclusion of additional contractors. The proposed model, based on the resource-based theory and the fuzzy Delphi method, allows the owner to identify critical resources required by the project, evaluate the capacities of the contractors, and identify the resource gaps. The paper also compares the analytical result with the ownerÕs actual decisions, and explains their differences. 2005 Elsevier Ltd and IPMA. All rights reserved.

Keywords: General management (strategy); Contractual (international projects); General (managing project); Project teaming; Procurement

1. Introduction which specialist contractors will be employed, in addi- tion to the prime contractor, because the success of , developed by Taipei Financial Center the project requires a sufficient provision of critical re- Corp. (TFCC), was designed by local architect C.Y. sources, such as trade know-how and special equipment Lee & Partners and was managed by Turner Interna- and materials. The prime contractor may provide gen- tional Ltd. It is a 101-story building that is officially eral resources (e.g., tower cranes and hoists) common the tallest building in the world today, being 508 m in to specialist contractors and maintain safety on the site. height and having a total floor area of 412,500 m2 Sometimes, as in the case of Taipei 101, the prime con- [21,26]. The US$700-million Taipei 101 project, initiated tractor may form a joint venture, encompassing several by Taipei city government, used a BOT (Build-Operate- firms, in order to ensure there are sufficient resources Transfer) approach with a 70-year concession period of available. property use. TFCC adopted a fast-track approach and As the construction project of Taipei 101 comes to an started foundation construction in January of 1998, be- end, this study was intended to review and rationalize fore the architectural design had been finalized [15,26]. the ownerÕs decisions in construction project teaming, In the construction of such a super-large project, the from the resource perspective, using resource-based the- owner often plays an important role in determining ory. A fuzzy Delphi survey approach was adopted to identify the critical resources required by the project. The capability of KTRT, the prime contractor, was as- * Corresponding author. Tel.: +886 35731982, fax: +886 35716257. E-mail addresses: [email protected] (R.-J. Dzeng), sessed for each critical resource, in order to identify [email protected] (K.-S. Wen). the resource gaps. The capabilities of the additional con- 1 Tel.: +886 35720327. tractors were also assessed to determine if they had filled

0263-7863/$30.00 2005 Elsevier Ltd and IPMA. All rights reserved. doi:10.1016/j.ijproman.2005.02.004 484 R.-J. Dzeng, K.-S. Wen / International Journal of Project Management 23 (2005) 483–491 the gaps and had better capability than the prime While this study focuses on the project ownerÕs deci- contractor. Finally, the ownerÕs decisions were com- sions of when to acquire specialist contractors, from the pared against the survey results and discrepancies perspective of resource-based theory, several researchers explained. (e.g., [23,3,20,25]) have identified criteria for pre-qualify- ing and mechanisms for selecting prospective contrac- tors or subcontractors from the perspective of the 2. Resource-based theory and construction teaming owner or the prime contractor; this information may be used subsequently, once the owner has determined According to Hoskisson et al. [11], Mahoney and whether or not to hire a specialist contractor. Pandian [16] and Yasemin and Mahoney [28], re- source-based theory is related to the resource-based view (RBV) of the firm in the field of strategic management 3. Classification of resources research, which focuses on the firmÕs internal strengths and weaknesses relative to their external opportunities Barney [2] divided a firmÕs resources into physical and threats. Traditional strategy models, such as Por- capital resources (e.g., equipment), human capital re- terÕs five forces model [22], relies heavily on the analysis sources (e.g., experience), and organizational resources of a companyÕs external competitive environment. In (e.g., a firmÕs formal structure). Hitt et al. [10] viewed contrast, the resource-based view is grounded in the per- afirmÕs resources as the input for its business process, spective that a firmÕs internal environment, in terms of and divided them into tangible resources (e.g., financial its resources and capabilities, is more critical to the resources) and intangible resources (e.g., patents). Grant determination of strategic action than is the external [8] viewed resources as the foundation for firm profits environment. Rather than being defined by the compet- and the major source of firm capabilities, and divided itive environment, Barney [2] and Grant [6] argue that them into financial resources, reputation, physical re- the parameters of a firmÕs competitive strategy are criti- sources, human resources, technological resources and cally influenced by its accumulated resources. In other organizational resources. Grant [9] re-organized re- words, what a firm possesses would determine what it sources into four categories, namely tangible resources, accomplishes. intangible resources, human resources and communica- Each organization is a collection of unique re- tive & interactive abilities. sources and capabilities that provides the basis for its This study considered the characteristics of building strategy and the primary source of its returns. A capa- construction projects, as well as the applicability of the bility is a capacity for a set of resources to integra- aforementioned resource categories, and divided con- tively perform a task. Readers interested in the tractor resources into assets and capabilities. The assets difference between RBV and other theories of firms include tangible assets (e.g., plant, construction equip- may refer to Conner [5], who compared RBV with five ment) and intangible assets (e.g., patented construction theories that have been significant in the evolution of methods, in-house geographic data) while the abilities industrial organization, including neoclassical theoryÕs include technological management abilities, operational perfect competition model [17], Bain-type industrial abilities, innovation abilities and experience with high- organization [1], the Schumpeterian competition [19], rise buildings. Chicago School responses [7] and transaction cost the- ory [27]. RBV seems particularly appropriate for examining an 4. Teaming of project Taipei 101 ownerÕs strategic teaming for a construction project be- cause such a project requires valuable and various re- The owner of Taipei 101, Taipei Financial Center sources from team companies. A super-large project Corporation (TFCC), first awarded the foundation normally requires resource capabilities exceeding what work to local contractor, Da-Hsin Construction Co. a single contractor can provide and needs several con- Later, TFCC selected KTRT, a consortium of - tractors to work as a team. Construction project team- based Kumagai Gumi Co., Taiwan Kumagai Construc- ing can be viewed as a strategic decision process which tion Co., and two other local contractors, RSEA will ensure that sufficient resource capabilities can be Engineering Corp. and Ta-Yo-Wei Construction Co., provided by the team as a whole. Thus, in a fast-track as the prime contractor, for an amount of US$590 mil- project, the process includes identifying the critical re- lion [26]. Kumagai Gumi Co., founded in 1898, has built sources required by the project, assessing the capabilities many skyscrapers, such as the 62-story AT&T Gateway provided by existing team members (whose contracts Tower in Seattle. Taiwan Kumagai, a branch of Kuma- have been tendered) for each critical resource, determin- gai Gumi Co. Ltd., was established in Taiwan in 1974. ing the resource gaps, and adding new team members to RSEA is the largest and the most capable heavy con- fill the resource gaps. struction enterprise in Taiwan, which also has great R.-J. Dzeng, K.-S. Wen / International Journal of Project Management 23 (2005) 483–491 485 experience in large building and infrastructure projects, quantitative values, the fuzzy Delphi method (FDM) such as the Jeddah Control Tower in Saudi Arabia. [4] was used in Step 2. Ta-Yo-Wei Construction Company, established in Since its inception by Dalkey and Helmer [6], the 1976, has been an important player in public works pro- Delphi method has been widely accepted as an effective jects (e.g., bridge and road constructions) in Taiwan [15]. forecasting tool, and has been used in a wide range of After determining the prime contractor, TFCC also applications, such as trend forecasting, knowledge cap- selected several specialist contractors, including the joint turing, public policy conglomeration and planning, etc. venture of Nippon Steel Corp. and China Steel Struc- The reported advantages of this method include: the ture Co. Ltd. for structural steel work; GFC Ltd. for ele- pooling of expert talents; the anonymity of experts, vator and escalator work; Gartner Ltd. for curtain wall thus avoiding domination of a group by individuals; work; Motioneering Inc. for the wind tuned mass dam- and iterations, with structured feedback from group re- per (TMD); and Samsung Corporation for interior sponses, to help reach a consensus [4]. One weakness of decoration. this method is that it requires the experts to be sur- The owner did not use any specific measures to deter- veyed repeatedly, usually more than twice, to allow mine which part of the work should be included within the forecast values to converge. Another drawback, the scope of the prime contractorÕs work or which found in many other survey techniques, is that the should be awarded to another specialist contractor. problems of ambiguity and uncertainty still exist, in These decisions were made based on unstructured dis- both survey questions and responses [24]; however, cussions during meetings. Our retrospective review of the fuzzy set theory can be adopted to deal with such these decisions, using resource-based theory, is described problems. below. Several versions of FDM (e.g., Murray et al. [18], Kaufmann and Gupta [13]) have been developed. They differ in the way participants are questioned, in fuzzy 5. Research objective and methodology membership functions, and in the number of survey iter- ations required. This study adopted the FDM proposed The objective of this study is to analyze the ownerÕs by Chang et al. [4], which is a modified version of the decisions for including additional specialist contractors, Max-Min FDM developed by Ishikawa and Amagasa from the resource capability perspective. More specifi- [12]. This was chosen because the respondents were re- cally, the authorsÕ intention was to discover: quired to answer a single-point estimate for each ques- tion, rather than a three-point estimate. 1. What were the critical resources required by the To determine the critical resources using FDM, the project? experts were requested to assess the degree of criticality 2. What were the resource gaps (i.e., what critical for each identified resource in the questionnaire, via an resources was KTRT lacking, in sufficient capacity)? interval value, on a scale of one to ten. A resource with 3. Were the ownerÕs selected additional contractors able one criticality implies that insufficient provision of a re- to fill the resource gaps, and provide better resource source only slightly affects the project. A resource with capabilities than the prime contractor? ten criticality implies that insufficient provision of the re- 4. Could resource-based theory explain the ownerÕs source significantly influences the project success. The decisions to include additional contractors? interval, with a lower and a upper value, represents the range of criticality of the resource. A range estimate The methodology used included six steps. Steps 1 is easier for experts to give than a single point because and 2 identified the critical resources required by the they estimate the project requirements with unknown project, based on the consensus of experts. First, all variable possibilities. potentially critical resources were elicited from litera- Professionals were required to answer the question- ture and interviews with professionals not involved naires a second time, if new resources had been added with Taipei 101. Second, 12 senior professionals from since the first round or if some of the resources had the project owner representatives (TFCC and Turner), not received assessment consensus. The assessed interval the construction team (KTRT), and the architect (C.Y. values for each resource r were then transformed into Lee), were selected as the experts for the subsequent triangular fuzzy numbers (TFNs). Appendix A details surveys. These experts were at the management level, the transformation of TFNs and determination of having anywhere from 11 to 29 years of professional convergence. experience. Because some resources did not receive assessment Survey I was sent to the experts in order to assess the consensus and some experts added new critical resources importance of each potential critical resource listed. Be- in Survey I, an additional survey, Survey II, was sent to cause the survey respondents tended to be imprecise the experts along with the information from the first when transforming qualitative impressions into specific assessment result. 486 R.-J. Dzeng, K.-S. Wen / International Journal of Project Management 23 (2005) 483–491

Finally, those resources, whose criticalities were Column 8 shows the convergence test result (see greater than the chosen threshold value (i.e., lower Appendix A for determination of convergence) and bound), were used in the subsequent analysis. The deter- highlights, with an asterisk, the resources whose assess- mination of the threshold value was subjective; it was ments had reached convergence. Note that some re- based on the number of identified resources and how sources (e.g., ability for logistic management) have two much analytical and survey effort could be spared for rows of TMF values; one is underlined and marked with the subsequent steps. In this case, we used 8 (80% of an asterisk (for Survey II) and the other is not (for Sur- the assessment scale of 10) as the threshold by the 80/ vey I). This indicates that these resources did not receive 20 rule. assessment consensus in Survey I, but reached consensus In step 3, Survey III, which contained the identified in Survey II. Column 9 shows the cross point value (see critical resources, was sent to the experts in order to Appendix A) for each of the resources whose assess- determine the prime contractorÕs capacities. Experts as- ments converged. Column 10 identifies and numbers signed a percentage value to each critical resource, rep- the critical resources whose cross point values are resenting the percentage of the resource required by the greater than the threshold of 8. project, which could be provided by the prime contrac- Survey III asked the experts to assess KTRTÕs capac- torÕs current capacity. For example, 100% meant that ity for each critical resource; the result is shown in col- the contractorÕs current resource capacity was just suffi- umn 11. The resources with a capacity deficiency gap cient for the project; 110% meant the capacity was more are shown in bold type. Survey IV asked the experts to than sufficient for project requirements and could assess the capacity of the owner-selected specialist con- accommodate a 10% resource increase due to change or- tractors for their related resources; these results are pre- ders; and 90% meant that the capacity was not sufficient sented in columns 12 and 13. Below, we discuss the and that another 10% of the resource had to be acquired findings of the survey. to fill the gap. A 95% confidence interval was calculated for the expertsÕ assessed percentage values for each re- source, as the result of this step. Because only twelve 7. What were the critical resources required by the subjects (630) participated in this study, the t distribu- project? tion was used [14]. Step 4 identified resource gaps by selecting the re- The resources listed in Survey I were chosen, and sources whose percentage scores in step 3 were less adapted based on a literature review. The experts intro- than 100%. These gaps identified the resources that duced several intangible resources in Survey I that we the owner had to seek from additional specialist con- did not include, such as the operational ability to inte- tractors, in order to ensure project success. Step 5 as- grate and coordinate all contractors. These resources, sessed the percentage of identified resource gaps that along with those that did not receive assessment consen- could be filled by the ownerÕs selected specialist con- sus in Survey I, such as logistics management and con- tractors. A 95% confidence interval was also calculated tract management abilities, were re-assessed in Survey for the expertsÕ assessed values. Step 6 compared the II. All resources received assessment consensus after survey results with the ownerÕs decisions to include spe- Survey II. The threshold of eight resulted in 40 critical cialist contractors. resources (numbered in column 10) out of a total of 45 critical resources identified. Patented construction methods, patented mechanical/ 6. Presentation of results electrical systems, relationship with the media and inno- vation ability in both building technology and mechan- Table 1 summarizes the result of determining critical ical/electrical systems, were five resources identified as resources and identifying resource gaps. Column 1 lists being non-critical. This was different from what we the resources surveyed in steps 1 and 2. Underlined re- had thought before conducting the surveys. We had sources (e.g., integration and coordination of all con- thought that patented technologies and a good relation- tractors) are those added by the experts in Survey I. ship with the media were critical to Taipei 101 as it was Columns 2–7 show the triangular fuzzy numbers, calcu- to become the worldÕs tallest building and would, there- lated based on the expertsÕ interval values, for measuring fore, continuously receive attention from the local media resource criticality in the FDM. Resources with a greater and the citizens of Taipei. interval between mg and Mg imply that their criticalities Most of the resources added by the experts in Survey have more variability depending on the project condi- I focused on the integration of all documents and oper- tions. Resources with a greater interval between lg and ations and the coordination of all parties on the site. The ug,orLg and Ug, suggest that the experts agree less on construction of Taipei 101 gathered together profession- the lower bound, or upper bound, respectively, of the als and companies from different countries, with differ- interval. ent cultures and customs, so coordination of these Table 1 Survey result identifying critical resources and resource gaps Resource items TFNs Convergence C.P Critical resource Resource gap

lg mg ug Lg Mg Ug Contractor capacity Additional Additional contractor contractorÕs capacity (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) Tangible assets Construction equip. for building 6 7.56 8.3 8 9.34 10 * 8.19 1 (106.02,112.24)

structure 483–491 (2005) 23 Management Project of Journal International / Wen K.-S. Dzeng, R.-J. Construction equip. for steel 5.8 7.63 8.6 8 9.41 10 * 8.36 2 (5.02, 7.21) Nippon Steel & (110.41, 114.28) erection China Steel Construction equip. for 5.5 7.41 8.5 7.5 9.02 10 * 8.08 3 (4.86, 7.12) GFC (107. 02, 111.61) elevators/ escalators Construction equip. for curtain 5 7.43 8.5 7.3 9.12 10 * 8.06 4 (6.61, 9.23) Gartner (108.23, 114.25) walls Construction equip. for TMD 5 7.39 8.3 7.8 9.11 10 * 8.1 5 (4.71, 6.92) Motioneering (104.51, 107.33) Construction equip. for 6 7.44 8.4 7.5 8.9 10 * 8.03 6 (105.06, 111.58) underground structure Construction equip. for interior 6 7.53 8.5 8 9.24 9.5 * 8.28 7 (107.23, 113.52) Samsung (106.34, 111.21) work Construction equip. for 6.5 7.61 8.5 8 9.28 9.5 * 8.29 8 (106.86, 112.11) mechanical and electrical systems Construction equip. for 5.6 7.55 8.6 8.2 9.11 10 * 8.39 9 (107.65, 111.68) Da-Hsin (109.25, 111.28) foundation

Intangible assets Patented construction method 3 5.18 6 5.5 6.12 7 * 5.72 Patented mechanical/electrical 3.6 5.21 6.3 5.5 6.41 7.5 * 5.86 systems Geographic data of the site 5.5 7.32 8.5 7.5 8.89 9 * 8.04 10 (100.85, 102.61) Climate data of the site 5 7.21 8.5 7.7 8.62 9 * 8.03 11 (100.12, 102.56) Technological Building structure 6 8.56 8.7 8.5 9.74 10 * 8.68 12 (107.52, 112.96) management ability Steel erection 6.5 8.52 8.8 8.5 9.71 10 * 8.74 13 (21.08, 26.56) Nippon Steel & (110.36, 113.25) China Steel T&M for elevators and escalators 6.7 8.42 8.8 8.5 9.68 10 * 8.73 14 (22.81, 25.99) GFC (108.24, 111.52) T&M for curtain walls 5.8 8.45 8.7 8.4 9.54 10 * 8.65 15 (21.21, 25.81) Gartner (107.32, 110.93) T&M for TMD 6.5 8.51 8.6 8.5 9.66 10 * 8.59 16 (22.96, 27.52) Motioneering (107.68, 110.35) T&M for underground structure 6.3 8.33 8.6 8.5 9.51 10 * 8.58 17 (106.51, 112.02) T&M for interior work 6 8.31 8.5 8.4 9.45 10 * 8.48 18 (107.53, 113.68) Samsung (106.33, 111.82) T&M for mechanical/electrical 5.8 8.44 8.7 8.5 9.62 10 * 8.66 19 (106.69, 111.85) systems T&M for foundation 6 8.4 8.6 8.5 9.58 10 * 8.58 20 (106.36, 112.58) Da-Hsin (107.33, 112.26) Operational Cash flow management ability 6.5 8.23 8.5 8.4 9.35 10 * 8.48 21 (106.85, 112.06) ability Project schedule management 5.8 8.29 8.5 8.1 9.12 10 * 8.43 22 (106.85, 112.06) ability

Relationship with media 4 5.18 6 5.5 6.05 8 * 5.7 487 (continued on next page) 488

Table 1 (continued) Resource items TFNs Convergence C.P Critical resource Resource gap

lg mg ug Lg Mg Ug Contractor capacity Additional Additional contractor contractorÕs

capacity 483–491 (2005) 23 Management Project of Journal International / Wen K.-S. Dzeng, R.-J. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) Assessment and maintenance of 5 7.75 8.3 7.8 8.51 9 * 8.08 23 (100.03, 102.91) neighboring facilities Logistics management 5.5 7.89 8.5 8.8 9.02 9.5 6 7.81 8.6 8.2 8.82 10 * 8.38 24 (107.81, 111.65) Risk management and emergency 5.3 7.74 8.3 8 8.86 9 * 8.18 25 (102.52, 105.02) response Contract management 6 8.12 8.6 7 9.37 9.5 * 8.33 26 (104.23, 107.88) Integration and Coordination of 5.5 8.41 8.6 8.1 8.93 9.5 * 8.51 27 (106.25, 111.26) all contractors Integration of drawings and 5.8 8.41 8.5 8.3 9.02 9.5 * 8.48 28 (107.69, 112.20) specs. of all contractors Safety and environment 5 8.12 8.4 8.1 8.84 9 * 8.32 29 (101.32, 102.63) management for the entire site Coordination with the owner, 5.6 8.28 8.6 8.3 8.97 9.5 * 8.5 30 (106.36, 111.65) architects and engineers, and contractors Coordination and negotiation 6.3 8.22 8.5 8.3 8.74 9.5 * 8.42 31 (105.20, 107.25) with government

Innovation Building technology 4.5 5.86 6.3 5.8 7.32 7.5 * 6.19 ability Mechanical/electrical systems 4.8 5.71 6.6 5.5 7.28 7 * 6.23 Experience with Building structure 6.5 8.15 8.5 8.3 9.25 9.5 * 8.45 32 (107.52, 112.96) high-rise buildings Steel erection 6.5 8.31 8.5 8.6 9.81 10 * 8.51 33 (20.92, 25.01) Nippon Steel & (108.41, 112.86) China Steel Elevator and escalator 6 8.02 8.5 8.2 9.31 10 * 8.41 34 (21.36, 25.32) GFC (107.26, 111.02) Curtain wall 6.5 8.15 8.8 7.8 9.52 10 * 8.53 35 (21.11, 24.29) Gartner (106.21, 110.11) TMD 6 8.22 8.8 8.4 9.68 10 * 8.68 36 (22.05, 26.92) Motioneering (106.02, 109.62) Underground structure 6 7.44 8.4 7.5 8.9 10 * 8.03 37 (106.01, 111.63) Interior work 5.5 7.56 8.5 7.5 8.5 9 * 8.02 38 (106.12, 113.03) Samsung (105.95, 110.64) Mechanic and electrical systems 6.5 7.61 8.5 8 9.28 10 * 8.29 39 (105.09, 110.15) Foundation 6 7.58 8.5 8 9.11 9.5 * 8.27 40 (105.20, 111.03) Da-Hsin (105.54, 111.84) R.-J. Dzeng, K.-S. Wen / International Journal of Project Management 23 (2005) 483–491 489 entities was challenging, to say the least. The integration of (108.23, 114.25) for the construction equipment for of drawings and specifications for different specialty curtain walls because it offered its in-house designed trades was also important, given the fast-track approach cranes that could move along slabs, and thus provided taken, which necessitated the design and construction of flexible storage of curtain walls. Although the experts Taipei 101 to proceed simultaneously. knew that KTRT was able to provide sufficient capacity The experts also added another two resources. One through outsourcing, specialist contractors who had was the ability to coordinate and negotiate with govern- their own manufacturing plants were able to accommo- ment, because Taipei 101 adopted a BOT approach, date a more flexible design and respond more quickly to making project progress dependent upon some items design changes (e.g., building height and department being furnished by the government. The other was the store layout were changed during construction). ability to maintain safety standards and environment Although there were two critical resources for which integrity; failure to do so could have led to termination KTRT had sufficient capacity, TFCC still included its of the project for inspection, which in turn could have own contractors. They were Da-Hsin for the foundation resulted in significant cost overruns and delayed sales work and Samsung for the interior work. As these two and rental income. This was especially critical, because contractors did not provide significantly better capacity the project was located in Taipei, which has the most than KTRT, why were they included in the team? Fur- rigorous building and construction codes in Taiwan. ther investigation revealed that the reason TFCC chose these two additional contractors was not resource- related as described by the follows. 8. What were the resource gaps? To maximize the net present value within the 70-year concession period of the BOT contract, TFCC chose a As shown by bold numbers in column 11 of Table 1, fast-track approach in order to start the operation and KTRT did not have enough capacity to provide for the receive income as early as possible. They began the construction equipment, technological management foundation construction, which took eighteen months, ability, and high-rise building experience in steel erec- before the design was finalized. Da-Hsin was the first tion, elevators/escalators, curtain walls and wind dam- to be awarded a construction contract, long before per. The resource gaps were significant and required KTRT was selected as the prime contractor. Da-Hsin other specialist contractors to complement the deficien- was also invited to bid for the prime contract, but did cies in capacity. not win because of its pricing. Similarly, TFCC chose Samsung over KTRT for the interior work, mainly be- cause SamsungÕs price was considerably lower than that 9. Were the ownerÕs selected additional contractors able to of KTRT. fill the resource gaps, and provide better resource capabilities than the prime contractor? 10. Could resource-based theory explain the ownerÕs Comparison of the critical resources with gaps (col- decisions to include additional contractors? umn 11), and the resources for which the owner included specialist contractors (column 12), shows that TFCC in- Fig. 1 plots the project teamÕs capacities for each of cluded these contractors for all resource gaps. However, the 40 numbered critical resources on a radar diagram. TFCC also included additional contractors for some re- The bold lines represent KTRTÕs capacities, i.e., the sources that had no gaps. teamÕs capacity without considering the ownerÕs se- TFCC selected a joint venture of Nippon Steel and lected specialist contractors. The bold points inside China Steel, GFC, Gartner and Motioneering for steel the 100-grade dashed circle (e.g., for resources 2, 3, 4 erection, elevators/escalators, curtain walls and wind and 5) indicate the resource gaps where KTRTÕs capac- damper, respectively. Each of these contractors pro- ities are not sufficient for the corresponding resources. vided sufficient capacity, significantly greater than These resources are the directions, suggested by re- KTRT, for their respective resources, because their spe- source-based theory, that the owner should follow to cialties matched the resource requirements and their determine if additional contractors are needed. Re- capacities were greater than the project need. For exam- source numbers with asterisks attached are the re- ple, the joint venture of Nippon Steel and China Steel sources that were sought by the owner, from received a 95% confidence interval of (110.41, 114.28) specialist contractors. The regular line represents the for the construction equipment for steel erection, be- teamÕs capacity, including the specialist contractors. cause it was able to provide 1450MT (Meter-Ton) lifting Thus, if an asterisk appears where a resource gap ex- capacity, compared to 1100MT required by the project. ists, or the regular point significantly exceeds the bold A higher lifting capacity can speed up the lifting process point in terms of capacity grade, then the ownerÕs and make the process safer. Gartner received an interval decisions coincided with the resource-based analysis; 490 R.-J. Dzeng, K.-S. Wen / International Journal of Project Management 23 (2005) 483–491

1 ceed the capacity of a single contractor, and therefore 39 40 2 3 the project owner may decide to include additional con- 38 4 37 5 tractors, to work as a team, with the prime contractor. 36 6 The intention of this study was to explain, in retrospect 100 35 7 and using resource-based theory, the decisions made to 34 8 include additional contractors, by the project owner of 33 9 Taipei 101. The survey results suggest that resource-

32 10 based theory may help an owner determine whether to include additional contractors, by identifying resource 31 11 0 gaps resulting from the insufficient resource capacity 30 12 of the prime contractor. However, it did not explain 29 13 all of the ownerÕs decisions related to selecting addi- 28 14 tional contractors, especially when an additional con- 27 15 tractor was employed due to project timing, bid 26 16 pricing, or because resources were unable to be shared 25 17 among contractors (e.g., cash). Therefore, this analytical 24 18 result, using resource-based theory, can be used as an 23 22 20 19 21 important guide in identifying resource gaps requiring Team's capacities without subcontractors additional resource providers; it should not, however, be used as the sole basis for determining the inclusion Team's capacity with subcontractors of additional contractors. Fig. 1. Comparison of resource gaps and TFCCÕs decisions regarding selected additional contractors. Appendix A. Triangular fuzzy numbers and convergence otherwise, the decisions were not able to be explained To determine if the experts have consensus on the by resource-based analysis. criticality of each resource, the assessed interval values The diagram justifies TFCCÕs decisions for including for each resource r can be transformed into triangular specialist contractors for resources 2, 3, 4, 5, 13, 14, 15, fuzzy numbers (TFNs), namely N~ r (the TFN of the 16, 33, 34, 35 and 36 because gaps existed for these re- upper-bound degree of criticality) and n~r (the TFN of sources. However, it does not justify TFCCÕs decisions the lower-bound degree of criticality), as shown in for including additional contractors for resources 7, 9, Fig. 2, using the following equations: 18, 20, 38 and 40, because gaps did not exist for these N~ r ¼ðLr; M r; U rÞ; ð1Þ resources, and the additional contractorsÕ capabilities Q 1 n n did not significantly exceed those of KTRT. As ex- where Lr – Min(Xri), i =1 n; Mr – i¼1X ri ; Ur – plained earlier, the inclusion of these contractors was Max(Xri), i =1 n. n – the number of experts; Xri – mainly due to reasons of project timing and pricing. the upper-bound degree of criticality assigned by the th During the survey, some experts suggested that, i expert for resource r. Mr – The geometric mean of Xri. although cash flow management ability was a critical re- n~r ¼ðlr; mr; urÞ; ð2Þ source, it should be excluded because oneÕs inability Q 1 n n could not be complemented by others; i.e., no contrac- where lr – Min(xri), i =1 n; mr – ð i¼1xriÞ ; ur – tors would be willing to provide a temporary cash loan Max(xri), i =1 n. xri – the lower-bound degree of to the prime contractor. Instead, it was felt that cash flow management ability should have been used as a pre-qualification criterion, and only those with sufficient cash flow management ability should have been further evaluated for their resource capabilities. This evaluation 1.0 could include the contractorÕs credit record, transaction records with banks, financial reports, contract volumes, etc.

Membership grade 0 11. Conclusions lr mr Lr ur Mr Ur Degree of criticality A super-large project, such as Taipei 101, requires many resources. These resource requirements often ex- Fig. 2. Transformation of assessed values into TFNs. R.-J. Dzeng, K.-S. Wen / International Journal of Project Management 23 (2005) 483–491 491

th criticality assigned by the i expert for resource r. mr – [11] Hoskisson RE, Hitt MA, Wan WP, Yiu D. Theory and research in strategic management: swings of a pendulum. J Manage The geometric mean of Xri. Whether the experts have reached convergence can be 1999;25(3):417–56. [12] Ishikawa A, Amagasa M. The max-min Delphi method and fuzzy determined by the overlapping of the upper-bound and Delphi method via fuzzy integration. Fuzzy Set Syst lower-bound degrees of criticality. The convergence is 1993:241–53. determined by comparing the gray area, defined by [13] Kaufmann A, Gupta MM. Fuzzy mathematical models in engineering and management science. Amsterdam: North-Hol- (ur Lr). Convergence was reached when 0 6 (ur L ) 6 (M m ). The cross point of two TFNs, equal land; 1988. r r r [14] Keller G, Warrack B. Statistics for management and econom- to (ur · Mr mr · Lr)/(Lr + Mr mr + ur), was used ics. CA: Thomson Learning Inc.; 2003. as the assessment consensus of the criticality of resource [15] KTRT. Available from: http://www.ktrt.com.tw/english/ r. index.htm (last updated, 2004). [16] Mahoney JT, Pandian JR. The resource-based view within the conversation of strategic management. Strategic Manage J 1992:363–80. References [17] McNulty PJ. Economic theory and the meaning of competition. Quarterly J Econ 1968;82:639–56. [1] Bain JS. Industrial organization. New York: John ; 1968. [18] Murray TJ, Pipino LL, van Gigch JP. A pilot study of fuzzy set [2] Barney J. Firm resources and sustained competitive advantage. J modification of Delphi. Hum Syst Manage 1985(5):76–80. Manage 1991;17:99–120. [19] Nelson RR, Winter SG. An evolutionary theory of economic [3] Bubshalt AA, Al-Gobali KH. Contractor prequalification in change. Cambridge (MA): Belknap Press; 1982. Saudi Arabia. J Manage Eng 1996;12(2):50–4. [20] Palaneeswaran E, Kumaraswamy MM. Contractor selection for [4] Chang PT, Huang LC, Lin HJ. The fuzzy Delphi method via design/build project. J Constr Eng Manage 2000;126(5):331–9. fuzzy statistics and membership function fitting and an [21] Reina P. Tall buildings: megastructure supports TaipeiÕs 508- application to the human resources. Fuzzy Set Syst 2000;112: meter ÔmegatowerÕ: team is on track to garner ÔworldÕs tallestÕ title 511–20. late next year. Eng News-Record, Nov. 24, 2003. [5] Conner KR. A historical comparison of resource-based theory [22] Porter ME. Competitive strategy: techniques for analyzing and five schools of thought within industrial organisation industries and competitors. New York: Free Press; 1980. economics: do we have a new theory of the firm. J Manage [23] Russell JS. Model for owner prequalification of contractors. J 1991;17:121–54. Manage Eng 1990;6(1):59–75. [6] Dalkey N, Helmer O. An experimental application of the Delphi [24] Sackman H. Delphi assessment, expert opinion, forecasting, and method to the use of experts. Manage Sci 1963;9(3):458–67. group process, R-1283-PR. Santa Monica: Rand Corp; 1974. [7] Demsetz H. Industry structure, market rivalry, and public policy. [25] Sonmez M, Holt GD, Yang JB, Graham G. Applying evidential J Law Econ 1973;16:1–9. reasoning to prequalifying construction contractors. J Manage [8] Grant RM. The resource-based theory of competitive advantage: Eng 2002;18(3):111–9. implications for strategy formulation. Calif Manage Rev [26] TFCC, Taipei Financial Center Corp. Available from: http:// 1991;33(3):114–35. www.tfc101.com.tw/english/index.htm (last updated, 2004). [9] Grant RM. Contemporary strategy analysis/concepts, techniques, [27] Williamson OE. Markets and hierarchies: analysis and antitrust applications. Cambridge (MA): Blackwell; 1995. implications. New York: Free Press; 1975. [10] Hitt MA, Ireland RD, Hoskisson RE. Strategic management: [28] Yasemin YK, Mahoney JT. Edith PenroseÕs (1959) contributions competitiveness and globalization with InfoTrac college edi- to the resource-based view of strategic management. J Manage tion. Cincinnati (OH): SouthWestern College Publishing; 2002. Stud 2004;41(1):183–91.