an.“
a : .
fi.

am
fine

 

t1
H»:.{.a1n$m.flf.. t J
Md... .uV10.\H.b.;udl..
I. '3‘ .1. . f

   

L .’. .

#1?

 

'u‘

|‘

Ww‘

51% iii a :7

This is to certify that the
dissertation entitled

AN INVESTIGATION OF INFORMATION TECHNOLOGY
INVESTMENTS ON BUYER-SUPPLIER RELATIONSHIP
AND SUPPLY CHAIN DYNAMICS

presented by
$00 Wook Kim

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Business Administration

 

 

 

Major Profess’or’s Signature
4 7/24 / 03

Date

MSU is an Affinnative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DatoDuo.p85-p.15

 

AN INVESTIGATION OF INFORMATION TECHNOLOGY INVESTMENTS
ON BUYER-SUPPLIER RELATIONSHIP AND SUPPLY CHAIN DYNAMICS

By

800 Wook Kim

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Department of Marketing and Supply Chain Management

2003

ABSTRACT
AN INVESTIGATION OF INFORMATION TECHNOLOGY INVESTMENTS
ON BUYER-SUPPLIER RELATIONSHIP AND SUPPLY CHAIN DYNAMICS
By
800 Wook Kim

This research posits that information technology investment and deployment for
efficient supply chain management should be implemented taking into consideration the
interactive, feedback relationships among IT level, product customization level, buyer-
supplier relationship, and supply chain structure.

Three kinds of feedback relationships are derived from prior literature. First is the
mutual, positive feedback relationship among IT level, the number of suppliers, and
buyer’s bargaining power from the perspective of transaction cost theory. In other words,
the continuous deployment of advanced IT ultimately can lead to perfect competition in
an electronic market among a firm’s suppliers by reducing transaction costs. As a result,
such electronic market structure can enable further deployment of more advanced IT by
the buyer. Second is a negative feedback relationship among [T level, the number of
suppliers, and buyer’s bargaining power from the perspective of “incomplete contract
theory,” which leads to a different and divergent view. Succinctly, the continuous
deployment of advanced IT cannot lead to persistent increase in the number of suppliers
due to the increased burden for non-contractible investment by the buyer. Consequently,
this perspective suggests that the continuous improvement of IT level stemming from
declining transaction costs is not possible. Third is a negative feedback relationship

among information sharing by advanced IT, product customization level, and supply

chain dynamics. Such negative feedback relationship suggests that information sharing
by the utilization of advanced IT may not guarantee persistent cost reduction or profit
increase throughout the supply chain, and that the effect of inforrnation-sharing can be
different depending on the detailed characteristics of SCM system including product
customization level. The effect of IT investment in supply chain management can be
different depending on the interaction among the above three feedback relationships
representing IT effect on buyer-supplier relationship and reflecting IT effect on supply
chain dynamics. In other words, depending on how the described feedback relationships
affect each other or which feedback relationship mostly dominates, the effect of IT
investment can be different.

The objective of this dissertation is to identify the nature of these interactive
feedback relationships among IT deployment, buyer-supplier relationship, and supply
chain structure, and also to ascertain whether or not there exists a significant effect from
product customization level on such interactive feedback relationships. Based on this
identification through the utilization of the system dynamics methodology, one of the
dynamic simulation methods, we can attempt to suggest a set of advisable SCM strategies
for effective IT investment and deployment. Investigating the responses of IT level
according to the dynamic changes of product customization level, buyer-supplier
relationship, and supply chain structural issues is a requisite first step. This would be
helpful in suggesting a dynamic E-technology investment and adoption model
appropriate for integrated supply chain management. The insights developed in this
dissertation should provide theoretical foundations and practical guidelines as to the role

and fimction of B-to-B electronic commerce for efficient supply chain integration.

Copyright by
800 WOOK KIM
2003

DEDICATED TO MY PARENTS, MOON H0 KIM AND YOUNG HEE DO

ACKNOWLEDGEMENT

I would like to express my gratitude to my advisor, Professor Ram Narasimhan, for
his support, patience, and encouragement throughout my graduate studies. IT is not often
that one finds an advisor and colleague that always finds the time for listening to the little
problems and roadblocks that unavoidably crop up in the course of performing research.
His technical and editorial advice was essential to the completion of this dissertation and
has taught me innumerable lessons and insights on the workings of academic research in
general.

My thanks also go to other members of my dissertation committee, Dr. Morgan
Swink, Dr. Roger Calantone, and Dr. Kenneth Boyer for reading previous drafts of this
dissertation and providing many valuable comments that improved the presentationand
contents of this dissertation.

I would like to reiterate my gratitude to my parents. Without them being a role model
for the attitudes towards work and life since I was very young, I could not have achieved
what I have done today. I wish also to extend my gratitude to my beloved wife, Ra
Young, for being there to tap my shoulders whenever I felt down towards my work, and
to my kids, Hyun Jung and Hyun Jin, who serve as an inspiration for me to move on
against all odds on my way. The sacrifice they made to support me to pursue graduate
education and the encouragement they provided during the entire period of the doctoral

program will always be appreciated and remembered.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... ix

LIST OF FIGURES ................................................................................. x
CHAPTER 1

INTRODUCTION .................................................................................. 1

1.1 Motivation ...... , .............................................................................. l

1.2 Research Question and Objectives ........................................................ 7

1.3 Methodology ................................................................................. 9

1.4 Significance of the Study ................................................................. 10

1.5 Organization of the Dissertation ......................................................... 13
CHAPTER 2

LITERATURE REVIEW .......................................................................... 15

2.1 Information Technology and Buyer-Supplier Relationship ........................... 15

2.1.1 Buyer-Supplier Relationship ........................................................ 15

2.1.2 Theoretical Backgrounds on Buyer-Supplier Relationship .................... 16

2.1.2.1 Transaction Cost Theory ................................................. 16

2.1.2.2 Resource Dependence Theory .......................................... 17

2.1.2.3 Incomplete Contract Theory ............................................ 19

2.1.3 The Portfolio of Buyer-Supplier Relationship ................................... 20

2.1.4 The Effect of Information Technology on Buyer-Supplier Relationship ..... 23

2.1.4.1 IS Effect in the Perspective of Transaction Cost Theory... ........... 23

2.1.4.2 IS Effect in the Perspective of Incomplete Contract Theory ....... 25

2.2 Information Technology and Supply Chain Dynamics ............................... 28

2.2.1 The Significance of Supply Chain Integration ..................................... 28

2.2.2 The Effect of Information Technology on Supply Chain Dynamics. . . ....31
2.3 Interactive Relationships among IT, Buyer-Supplier Relationship and SC

Structure ..................................................................................... 35
2.3.1 The Alignment among Product Customization Level, Supply Chain
Structure, and Buyer-Supplier Relationship ..................................... 35
2.3.2 The Effect of IT on the Alignment among Product Customization Level,
Supply Chain Structure, and Buyer-Supplier Relationship .................... 41
CHAPTER 3
RESEARCH FRAMEWORK ...................................................................... 46
3.1 Conceptual Framework .................................................................... 46
3.2 Research Scope ............................................................................. 51
CHAPTER 4
RESEARCH DESIGN .............................................................................. 53
4.1 System Dynamics ........................................................................... 53
4.2 System Dynamics Modeling Process ..................................................... 54

vii

4.2.1 Causal Loop Diagram ............................................................... 57

4.2.2 System Dynamics Diagram .......................................................... 63
4.2.2.1 Overview of System Dynamics Diagram ............................. 63
4.2.2.2 System Dynamics Model of this Dissertation ........................ 65
4.2.3 Formulation of Equation ............................................................ 79
4.3 Validity of System Dynamics Model ................................................... 84
4.3.1 The Meaning of Validation ......................................................... 84
4.3.2 Model Validation Test .............................................................. 86
4.2.3.1 Structural Tests ........................................................... 87
4.2.3.2 Behavioral Tests ........................................................... 92
4.2.3.3 Policy Tests ................................................................ 95
CHAPTER 5
RESULTS OF SIMULATION ..................................................................... 96
5.1 The Relationships among IT Level, Buyer-Supplier Relationship, and Supply
Chain Dynamics ............................................................................ 96
5.1.1 The Relationship between IT level and Buyer-Supplier Relationship ....... 97
5.1.2 The Relationship between IT level and Supply Chain Dynamics ........... 105
5.1.3 The Effect of Product Variety Level on the Interactions among Feedback
Loops ................................................................................ 108
5.2 The Suggestion of Policy Leverage ..................................................... 118
5.2.1 The Necessity of Policy Leverage ............................................... 118
5.2.2 Parameter Adjustment Policy ..................................................... 119
CHAPTER 6
IMPLICATION AND CONCLUSION .......................................................... 129
6.1 Contribution ................................................................................ 129
6.1.1 Positive Feedback Relationship between IT level and Buyer’s Bargaining
Power ................................................................................ 130
6.1.2 Change of Information Sharing Level according to IT Level ................ 130
6.1.3 Existence Possibility of Win-Win Strategy ..................................... 131
6.1.4 Validity of the Portfolio of Inter-F inn Transaction Relationship ............ 132
6.1.5 The Availability of Information Technology Itself vs. The Significance of
Managerial capability on Information Technology ............................ 133
6.2 Future Research ............................................................................ 134
APPENDICES ..................................................................................... 137
BIBLIOGRAPHY ................................................................................. 154

viii

LIST OF TABLES

Table 2-1 Asset Specificity ........................................................................ 19
Table 2-2 Contextual Profiles ...................................................................... 22
Table 2-3 Management Profile for Each Contextual Profiles ................................. 23
Table 2-4 The Effect of E-Commerce Adoption on Supply Chain Management ........... 43
Table 4-1 Terminology for Identifying Stocks and Flows ..................................... 65
Table 4-2 Guidelines for formulating Table Functions ......................................... 68
Table 4-3 Initial Stock Points ...................................................................... 83

ix

LIST OF FIGURES

Figure 2-1 Relationship between IT Deployment and Buyer-Supplier Relationship ...... 27
Figure 2-2 Traditional Trade-off Relationship between Inventory and Transportation
Costs ..................................................................................... 34
Figure 3-1 Conceptual Framework ............................................................... 50
Figure 4-1 Causal Loop Diagram vs. System Dynamics Diagram ............................ 56

Figure 4-2 Causal Loop Diagram for IT effect on Buyer-Supplier Relationship in the
Perspective of Transaction Cost Theory ............................................ 58

Figure 4-3 Causal Loop Diagram with IT effect on Buyer-Supplier Relationship in the

Perspective of Incomplete Contract Theory ........................................ 60
Figure 4-4 Causal Loop Diagram with IT effect on Supply Chain Dynamics

(Full Model) ............................................................................. 62
Figure 4-5 System Dynamics Diagramming Notation .......................................... 64
Figure 4-6 Lookup Functions (1) .................................................................. 70
Figure 4-7 Lookup Functions (2) .................................................................. 71
Figure 4-8 Lookup Functions (3) .................................................................. 73
Figure 4-9 Lookup Functions (4) .................................................................. 77
Figure 5-1 Demand Stream ......................................................................... 97
Figure 5-2 The Change of IT Level and No. of Supplier ....................................... 97
Figure 5-3 The Change of Buyer’s Revenue and Cost ......................................... 98
Figure 5-4 The Change of Profit ................................................................... 99
Figure 5-5 The Change of IT Level and Supplier’s Inventory ............................... 100
Figure 5-6 The Change of Supplier’s Revenue and Cost ..................................... 101
Figure 5-7 The Change of No. of Supplier and RM Price .................................... 102
Figure 5-8 The Change of Transaction Cost and Effects of BNCI and AS] ............... 103

Figure 5-9 The Change of Inventory Level ..................................................... 106

Figure 5-10 Sensitivity Analysis on IT level by Product Variety Level .................... 109
Figure 5-11 Sensitivity Analysis on No. of Supplier by Product Variety Level... .......1 1 1
Figure 5-12 Sensitivity Analysis on Demand and Target Inventory ........................ 1 13
Figure 5-13 Key Variables driving the Change of No. of Suppliers ........................ 114
Figure 5-14 Key Variables driving the Change of IT level ................................... 115
Figure 5-15 The Changes of RM price and related variables ................................. 117

Figure 5-16 Sensitivity Analysis on Key Variables by Parameter Adjustment (1)........123
Figure 5-17 Sensitivity Analysis on Key Variables by Parameter Adjustment (2). .......124
Figure 5-18 Sensitivity Analysis on Key Variables by Parameter Adjustment (3).. ......125

Figure 5-19 Sensitivity Analysis on Key Variables by Parameter Adjustment (4). .......126

xi

CHAPTER 1. INTRODUCTION

1.1 Motivation

The cutting edge for business today is electronic commerce. In the face of strong
market forces created by electronic commerce and mounting competition, firms can no
longer plod along historical tracks or seek the preservation of the status quo (Kalakota
and Whinston 1997). Commonly defined, electronic commerce is associated with the
buying and selling of information, products and services via computer networks or via
any one of the myriad of networks that make up the information superhighway (Kalakota
and Whinston 1996). The Automotive Industry Action Group in North America defines it
as the enablement of a business vision supported by advanced information technology to
increase the effectiveness of the business relationships between trading partners (Kim
1999).

Electronic commerce is becoming critical in three interrelated dimensions:
Customer-to-Business interactions; Intra-business interactions; and Business-to-Business
interactions. In particular, from the Business-to-Business or inter-organizational
perspective, electronic commerce facilitates changes in many business applications such
as supplier management, channel management, payment management and so on
(Kalakota and Whinston 1997). In the face of these changes, and in order to survive and
be successful, management has to cope with the changes taking place in the various

market spaces (Kim 1999).

The expansion of business-to-business electronic commerce changes the existing
shape of transaction relationship between companies. This is because the introduction of
B-to-B EC enables the construction of new business model, which was unavailable under
the existing private network or customer-to-business electronic commerce. In other
words, the rapid expansion of electronic commerce has huge potential for enabling
companies, large and small, to gain new marketplaces globally at low cost, or to be
disinterrnediated by others doing so. A totally new competitive environment opening up
new opportunities is upon us. As a result, the electronic commerce is growing quickly
and the importance of electronic marketplace is emphasized.

During the last decade, buying companies have increasingly emphasized the
importance of strategic cooperation and supply-network construction with suppliers, and
systematic supply chain management as a critical success factor for sustainable
competitive advantage. In particular, recently, due to the development of computer and
telecommunication technology, firms have been attempting to improve the efficiency of
transaction between buyer and supplier by information sharing and communication
through advanced information technology. Thus, information technology is recognized as
a way that buying companies can manage more efficiently their supply chain. In the light
of such current change and trend, the deployment and adoption of information technology
between buyer and supplier is highlighted as not only a fresh opportunity to obtain
competitive strategy, but also a facing task both buyer and supplier should challenge and
overcome together.

One of the most important and common findings from the supply chain dynamics

research is the well-known phenomenon of demand amplification and inventory

fluctuation that are more significant in the upstream of supply chain. One of the major
causes of this problem is due to the lack of information sharing between supply chain
partners (Lee et al. 1997a, 1997b, 2000). In such a reason, the utilization of information
technologies and systems in supply chain management has mainly focused on supply
chain integration. Through the utilization of such information systems, companies have
been able to attempt the integration of various functions spread over different areas
within a company and with external suppliers and customers as well as curtail
unnecessary activities. This effort enhances their capability to cope with sophisticated
needs of customers and meet the quality standards of products (Bardi et a1. 1994, Carter
and Narasimhan 1995).

However, from the viewpoint of supply chain management, applying the B to B
electronic commerce to the relationship between firms may be expected to make major
changes. That is, viewed from the perspectives of transaction cost theory and resource
dependence theory, Web-based electronic commerce which has totally different
characteristics with the existing information technologies and systems, may reduce the
excessive dependence on supply chain integration with few key suppliers by reducing
various transaction costs and deriving complete competition in electronic market among
numerous suppliers. This also means that the actual points of supply chain structural
issues related to and focused on just the establishment of SC integration may be changed
depending on the level of E-commerce utilization.

However, studying the role of incomplete contracting considerations in determining
the optimal number of suppliers suggests that, despite recent declines in search costs and

coordination costs due to information technology development, firms do not want to

increase the number of suppliers (Bakos and Brynjolfsson 1993), and IT has a significant
role in shifting inter-firm relationship structure to the middle between market and
hierarchical structures (Clemons et al. 1993). This is because the deployment of advanced
information technology may increase the importance of non-contractible investments by
buyers, such as quality, responsiveness, and innovation in the perspective of incomplete
contract theory, and transaction cost may actually increase, thus bringing out the
necessity of close buyer-supplier relationship to reduce the burden of transaction cost
increase. This argument implies that when such investments are particularly important,
firms will employ fewer suppliers, and this will be true even when search and transaction
costs are very low. This emphasizes that other factors should be accounted for in a more
complete model of buyer-supplier relationship.

The research of Bensaou (1999) suggests conceptually that product customization
level can be a critical factor for explaining such complete model of buyer-supplier
relationship. That is, through the suggestion of the portfolio of inter-firm transaction
relationship, they stressed that the type of market exchange relations in which the
technical and economical dependence of both buyer and supplier on each other is
relatively low, is appropriate for highly standardized products. Meanwhile, the type of
strategic partnership in which buyer-supplier relationship is strongly connected by
considerable transaction-specific assets of both buyer and supplier, is significantly
correlated with highly customized products.

Also, Fisher (1997) asserted that the most important strategic issue, which should be
considered in the design of such supply chain structure, is the characteristics and strategy

of traded products. Specifically, he emphasized that traded products can be classified into

efficient product and innovative product, and supply chain appropriate for each product
category should be discriminated into physically efficient supply chain and market
responsive supply chain. When considering the relationship between buyer-supplier
relationship and product strategy Bensaou suggested, such matching of product
characteristics/strategy and supply chain structure Fisher identified implies that the
strength of buyer-supplier relationship also can be the key strategic issue for constructing
efficient supply chain structures. In fact, Anderson et al. (1994) commented that
depending on the position and level of negotiation power between buyer and supplier, the
overall design and type of supply chain structure could be different, thus supporting the
above argument.

Also, many previous studies (Eisenhardt 1985; Hoecstra and Rome 1991; Bakos
1991; Lassar and Kerr 1996; Fisher 1997; Jones et al. 1997) comment that the
relationship between product variety level and buyer-supplier relationship can influence
supply chain structural issues, and also inversely, depending on such effect on supply
chain structural issues, corporate SCM strategies on product customization and buyer-
supplier relationship can be changed in terms of performance improvement.

The above arguments imply that information technology deployment for efficient
supply chain management should be implemented taking into consideration the
interactive, feedback relationships, among product customization level, buyer-supplier
relationship, and supply chain structure. In other words, the systematic design of supply
chain structure and appropriate application of information technologies can be decided
depending on product customization level and the strength of buyer-supplier relationship.

Also, the above arguments stress that the role of information technology should be

discussed from the perspective that information flows in supply chains can be better
utilized by considering 1) various operating environments and characteristics in a supply
chain, 2) the characteristics of traded products, and 3) buyer-supplier relationship.

Viewed in this perspective, strategic alignment among key SCM strategic and
structural issues, and information technology deployment, should be regarded as the most
significant and urgent research theme for the construction of an effective SCM strategy. It
is necessary to better understand the role and function of information technology for
efficient supply chain management.

However, past research on information technology has mainly focused on the
availability and significance of information technology itself. The assertion of Keen
(1993) that the difference in competitive and economic benefits which firms can obtain
from information technology is dependent on not the difference in technology itself, but
the difference in managerial capability, emphasizes the importance of research on the
adoption of information technology in supply chain management. This is also a reason
that even though the previous works have highlighted the necessity and significance of
information sharing in order to remedy the so-called bullwhip effect, practical
implementations have fallen short of theoretical recommendations. That is, few
companies have succeeded in mitigating or eliminating the bullwhip effect through
information sharing, thus supporting Keen’s observation. According to the above
argument, this study focuses on the suggestion of a set of advisable solutions for the
effective investment and utilization of E-technology, by developing and testing a
framework that posits the relationships among information technology, product

customization level, buyer-supplier relationship, and supply chain structural issues.

1.2 Research Question and Objectives

From the motivation described above, the following research questions emerge:

(1) How does advanced IT deployment affect the balance of bargaining power
between buyer and supplier? In other words, how does advanced IT deployment change
the number of suppliers, the buyer’s bargaining power, and the supplier’s bargaining
power?

(2) Inversely, how much does a changed position of buyer-supplier bargaining power
affect advanced IT deployment?

(3) How much does information sharing by advanced IT deployment alter supply
chain structure? In other words, how much does information sharing by advanced IT
deployment change overall inventory level, order quantity, order penetration point, and
production quantity in the supply chain?

(4) Inversely, how do such changes of supply chain structural issues influence
advanced IT deployment?

(5) How do supply chain dynamics influence the feedback relationship between IT
level and buyer-supplier relationships? In other words, how do the dynamic changes of
supply chain structural issues influence the interactive feedback relationships between
advanced IT deployments, number of suppliers, and buyer’s bargaining power?

(6) Does product customization level significantly affect such interactive feedback
relationships among buyer-supplier relationships, supply chain structure, and advanced

IT deployments exist?

The accumulation of these research questions represents some key points that must
be considered in the creation of a more effective IT investment model. The first important
issue concerns the level of IT deployment. It specifically deals with determining which
level of information technology is the most effective for supply chain management in
terms of performance improvement. As such, it is necessary to consider the relationship
between the levels of various telecommunication/information technologies and the effects
of their utilization. The second issue relates to the formation of an appropriate type of
buyer-supplier relationship and supply chain structure. Of particular concern is which
type of buyer-supplier relationship and supply chain structure is the most effective for IT
investment and deployment in terms of performance improvement. The third issue
concerns the consideration of interactive relationships. This point is based on the
proposition that the advisable IT investment model can be different depending on
interactive relationships among product customization level, buyer-supplier relationships,
and the type of supply chain structure. Thus, examinations on various cases that can be
created from the combination of information technology type and supply chain
configurations should be implemented to identify the most effective feature of the IT
investment model.

Viewed in light of the above arguments, the first objective of this dissertation is to
investigate the response of IT level according to the dynamic changes of buyer-supplier
relationships and supply chain structural issues by testing interactive feedback
relationships among IT deployment, buyer-supplier relationships, and supply chain
structure. The second objective is to identify whether or not product customization level

significantly affects such interactive feedback relationships. Based on the successful

accomplishment of these objectives, we can suggest a set of advisable supply chain
management (SCM) strategies for effective IT investment and deployment for efficient
supply chain management.

Doing so would be helpful in suggesting a dynamic E-technology investment and
adoption model appropriate for integrated supply chain management. Such efforts should
also enable us to provide theoretical foundations and practical guidelines on the role and
function of B-to-B electronic commerce for the efficient construction of supply chain

integration as mentioned previously.

1.3 Methodology

This dissertation develops a dynamic simulation model of advanced IT deployment
considering interactive feedback relationships with product customization level, buyer-
supplier relationship, and supply chain structure. This research utilizes the System
Dynamics method, a method of studying the world around us by viewing the system as a
whole. System Dynamics examines the interaction between all objects or individual parts
of a system and their relationship to one another.

System Dynamics has two typical methodological characteristics. First, it focuses on
a system’s dynamic behavior, that is, the behavioral changes within the system according
to the progression of time. Second, System Dynamics analyzes the fundamental reasons
for dynamic change through a feedback structure. These characteristics precisely relate to
the objectives of this dissertation as mentioned above, which is the identification of the

effect of dynamic changes of IT effect by investigating interactive feedback relationships

with product customization level, buyer-supplier relationships, and supply chain
structure. Thus, this is the reason for using System Dynamics as the methodology for
pursuing the objectives of this dissertation.

Traditionally, System Dynamics researchers have used two modeling approaches.
One is known as the top-down modeling approach, and the other is known as the bottom-
up modeling approach. In the topadown approach that emphasizes ‘feedback loop
thinking’, a causal loop diagram is made, followed by a detailed System Dynamics
diagram. In the bottom-up approach that emphasizes ‘operational thinking’, a System
Dynamics diagram is established first by linking individual stock and flow variables, and
then a causal loop diagram is completed by gradually expanding the System Dynamics
diagram. This research employs the top-down modeling approach. Specifically, System
Dynamics analysis of this dissertation is implemented through a three stage simulation
modeling process of establishing a causal loop diagram, designing a System Dynamics

diagram, and formulating an equation.

1.4 Significance of the Study

From the accomplishment of research objectives through testing the System
Dynamics model, this paper anticipates the following major suggestive contributions.

First, it has been generally recognized that information sharing by advanced IT
deployment can significantly minimize the problems induced by bullwhip effect, which is
defined as distortion in demand information and inventory fluctuation (Bowersox and

Closs 1996; Lee et al. 1997a; Lee et al. 1997b; Closs et al. 1998; Lee et al. 2000).

10

However, Chen (1998) and Grahovac & Chakravarty (2001) suggest a paradox of the
previous arguments on the benefit of information sharing. Such paradox implies that the
effect of information sharing can be different depending on the detailed characteristics of
a SCM system. Moinzadeh (2002) suggests four circumstances in which inforrnation-
sharing can be most beneficial: (a) the supplier’s long lead time (long transit times from
the supplier), (b) not large number of suppliers, (c) not too small or not too large order
quantities, (d) not too small or not too large ratio of the unit holding cost of the
manufacturer to that of the supplier. This dissertation will investigate the validity of
Moinzadeh’s suggestion. Furthermore, this dissertation will check the possibility of the
existence of other system characteristics enduring the benefit of information sharing.
Second, the results of this study can be used for the construction of an E-supply
chain progression paradigm. Traditionally, it has been generally accepted that the
developmental stage of E-networks would shift from the private network through value-
added networks to an open network. Also, EDI, a representative type of inter-
organizational telecommunication system, would move from an intra-EDI through a
value-added EDI to an Intemet-EDI and Web-based EDI. The level of IT development,
which is one of this paper’s construct variables, represents this shift. As mentioned
previously, this study can check the responses of IT deployment level depending on the
dynamic change of product customization level, buyer-supplier relationships, and supply
chain structural issues through the implementation of the System Dynamics simulation
model. Accordingly, if connecting such dynamic changes of SC strategic and structural

issues derived from System Dynamics simulation to the shifi of the IT developmental

11

stage described above, the construction of an E-supply chain progression paradigm is
possible.

Third, this study can identify the existence possibility of a new type of supply chain
structure. Fisher (1997), through the revision of product-process literature of Hayes and
Wheelwright (1979), argues that typically, according to the characteristics of traded
products, supply chain can be classified into two categories; physically efficient and
market-responsive. This paper, by considering buyer-supplier relationships and the effect
of IT investment other than the characteristics of traded products, investigates the validity
of Fisher’s argument and further reaches for the likelihood of creating a new type of
supply chain structure suitable for supply chain management in an E-commerce era.

Fourth, this study can identify the most compatible sets of the characteristics of
traded products and buyer-supplier relationships for effective information technology
investment and adoption. Bensaou (1999) showed that the type of market exchange
relation in which the technical and economic interdependence of buyer and supplier is
relatively low, is appropriate for highly standardized products, while the type of strategic
partnership in which buyer-supplier relationship is strongly connected by considerable
transaction-specific assets of each party, is significantly correlated with highly
customized products. Previous literatures have analyzed the relevance between strategic
issues and structural issues in supply chain management, under the general agreement of
the above combination. However, as mentioned previously, compatibility between
product customization level and buyer-supplier relationship can vary depending on
interactive relationships with IT deployment level and supply chain structure. Through

the analysis on interactive feedback relationships among such four construct variables,

12

this study investigates the possibility that the matching type of product strategy and
buyer-supplier relationship suitable for the E-commerce era may be diverse.

Fifth, this paper examines the likelihood of the implementation of a new dimensional
manufacturing strategy framework. Numerous literatures comment on the trade-offs
among supply chain structural issues (Pine et al. 1993). For example, as mentioned
previously, if lead-time is reduced and distribution time is frequent, overall inventory
level is decreased, while transportation is increased. In this scenario, trade-off
relationships among various supply chain variables can exist (Magee et al. 1985).
However, the deployment of advanced information technology may bring about a change
in the traditionally accepted concept of trade-off relationships among various supply
chain variables. Specifically, the deployment of advanced information technology may
enable buying firms to accept a little longer lead-time relative to the existing lead-time in
terms of minimum total cost. This study tests the possibility that advanced IT deployment

mitigates the trade-off relationships among supply chain structural issues.

1.5 Organization of the Dissertation

This dissertation is organized as follows. Chapter 1 describes the motivation of this
research, main research questions and objectives, methodology, and the significance of
this dissertation. Chapter 2 discusses the various research and literature streams leading
to the development of a research model for testing. Gaps in the literature are identified
and analyzed. Chapter 3 develops a conceptual fiamework based on literature and the

rationale behind the research propositions. Chapter 4 describes research scope and

13

methodology, and develops a causal-loop diagram and System Dynamics simulation
model, with the selection of key constructs and decision parameters. Chapter 5 analyzes
the dynamic effects of IT adoption in various SC strategic and structural perspectives
through the implementation of the System Dynamics model, and discusses theoretical and
managerial implications derived from the analysis. Chapter 6 concludes this research with
a discussion of the research contributions, limitations, and directions for future

investigation.

14

CHAPTER 2. LITERATURE REVIEW

2.1 Information Technology and Buyer-Supplier Relationship

2.1 .1 Buyer-Supplier Relationship

In response to intense global competition and shrinking product life cycles,
organizations have downsized to focus on core competencies and have attempted to
achieve a competitive advantage by forming mutually beneficial relationships with
suppliers to capitalize on their capabilities and technology. SCM evolved when firms
entered into strategic buyer-supplier alliances, and integrated their distribution and
transportation activities in conjunction with logistics providers.

It has been generally recognized that buyer-supplier relationships can provide a
strategic source of efficiency and even competitive advantage if managed appropriately.
Recently, business managers have pursued the quantification of the benefit that might be
extracted from the efficient management of such associations (Cooper and Slagrnulder
1999). Buying firms are actualizing important chances for competitiveness through
transaction with suppliers, which can provide benefit by helping lower a buying firm's
expenses (Cannon and Homburg 2001).

Even from an academic perspective, buyer-supplier relationships have been
emphasized increasingly in the perspective of new management philosophies that indicate
that effective liaisons will open up innovative competitive environments and significantly
contribute to a firrn's strategic success. Many previous articles in both practitioner and

academic journals demonstrate that the buyer-supplier relationship has played a

15

considerably significant role in the success of many organizations over the past few years
(Dowlatshahi 1999).

2.1.2 Theoretical Backgrounds on Buyer-Supplier Relationship

2.1.2.1 Transaction Cost Theory

The focus of transaction cost theory is that the inter-firm relationship should be
established in the perspective of minimizing transaction costs. Even though it was
initiated in an economic background, transaction cost theory provides a fundamental
basis in the research on supply chain management. This is because the most important
factor for supply chain competitiveness is the achievement of the economics of
networking and the basic unit for achieving the economics of networking is the
optimization of the inter-firm transaction relationship.

The expenditures related to the decision to select an optimal inter-firm relationship
include production costs, transaction costs, operating costs, and sunk costs. Optimal inter-
firm relationships should be decided in the viewpoint of minimizing the sum of these
costs. Among these costs, production costs pertain to the costs for preparing and
proceeding with production, while sunk costs are costs related to investment for a specific
transaction. Particularly, high sunk costs can have a significant influence on the level of
inter-firm relationships in that the cost for withdrawing from the existing specific
transaction relationship is very high. Operating costs, (with the exception of the above
two costs in the manufacturing process) are related to business type or the characteristics
of product and process technologies. Transaction costs are the costs for gathering
information from a possible transaction partner, which includes the costs for seeking out

an available transaction partner, negotiating and contracting a specific transaction, and

16

monitoring the implementation of the contract. Therefore, transaction costs increase
under a high level of incomplete information, bounded rationality, environmental
complexity and uncertainty, and opportunistic behavior (Williamson 1985).

The studies of Williamson (1975, 1985, 1981) and Grossman and Hart (1986), which
are the representatives of transaction cost theory, assume that an inter-firm contract under
complexity is always incomplete. Because of such incomplete contractual agreements,
firms with investment in relationship-specific assets may be deprived of additional
benefits from relationship-specific assets of transaction partners. One way for the stable
security of benefits in this case is the internalization of transaction partners. In other
words, by integrating transaction partners, this approach removes opportunistic behaviors
from partners. A less extreme alternative relative to internalization is the reciprocal
buying agreement. In a reciprocal buying agreement, an individual transaction party
should minimize its own opportunistic behaviors and clarify the scope of benefits through
the exchange or partial ownership agreement of hostages. Transaction cost theory
suggests a logical foundation for viewing the relevance and characteristics of such
transaction relationship types in terms of transaction costs. Consequently, transaction cost
theory places the focus of the research on the suggestion of explanations of how to select
transaction partners in order to minimize transaction costs and how to maximize the
effect of investment on the relationship-specific assets.
2.1.2.2 Resource Dependence Theory

The central theme of resource dependence theory is that firms try to secure resources
and reduce environmental uncertainty by establishing a relationship with the environment

or a party that possesses critical resources necessary for surviving in an uncertain

l7

environment. Inter-firm relationships are regarded as the response to uncertainty and
dependence.

In the case that any kind of mutual dependency between firms in a supply chain
exists, a high level of co-specialization in the inter-firm relationship is inevitable. The
level of inter-firm relationship co-specialization will vary according to decision-making
for investment on transaction-specific assets. Resource-based competition theory asserts
that such transaction-specific assets can be a source of competitive advantage in case that
it is impossible to replace and very difficult to imitate. However, investment on un-
imitable transaction-specific assets may become sunk costs, and thus increase the
likelihood of opportunistic behavior by a transaction partner. This means that once
investment on transaction-specific assets is implemented, mutual dependency between
transaction parties becomes high, and the cost for the suspension of the existing
transaction as well as the cost for any new transaction with an alternative partner
increases. Thus, it is emphasized that relevant countermeasures for preventing such
opportunistic behaviors should be carefully arranged according to the level of specific
assets. Williamson (1979) classified specific assets into three types: site asset specificity,
physical asset specificity, and human asset specificity. Dyer (1996) set up and measured
empirically the detailed dimensions of these three specific assets as indicated in table 2-1.

Resource dependence theory is not incompatible with transaction cost theory. But,
the relationship-specific asset addressed in transaction cost theory makes it very difficult
to change transaction partners easily or requires a high burden of cost to change partners.
Accordingly, the level of relationship-specific assets is proportional to the strength of

inter-firm dependency, thus emphasizing the relevance between the two theories.

18

Table 2-1: Asset Specificity

 

 

Definition (Williamson 1979) Measures (Dyer 1996)
Site The situation whereby successive production The physical distance between buyer and
Specificity stages are located in close proximity to one supplier

another to improve coordination and
economize on inventory and transportation cost

 

 

 

Physical Transaction-specific capital investments The percentage of the supplier’s total
Asset (e.g., in customized machinery, tools, dies) capital investments which would have to
Specificity be scrapped if the supplier were
prohibited from conducting transaction
Human Transaction-specific know-how accumulated The ratio of the annual man days that
Asset by transactors through long—standing buyer- spent in face to face contact of buyer-
Specificity supplier relationships (e.g., dedicated supplier supplier to total annual man days
engineers who learn the systems, procedures, The ratio of the average no. of co-located

 

 

and individuals that are idiosyncratic to buyer) or ‘guest’ engineers to total engineers

 

(Source: Dyer, 1996).
2.1.2.3 Incomplete Contract Theory

Transaction cost theory as mentioned above, addresses the type of inter-firm
relationship by using the principle of cost minimization under economic background.
However, when considering that quality, flexibility, and time-based competition beyond
cost dimension are emphasized as the key competitive factors in the recent market,
transaction cost theory cannot flawlessly determine the most suitable type of transaction
relationship, even though it provides the starting point for discussion.

An inter-firm cooperative relationship pursues not only financial benefits such as
cost efficiency, but also the improvement of non-financial aspects such as quality,
responsiveness, defect rate, innovation, technology acquisition, reliability, and
information sharing. In order to improve such non-financial factors, a certain level of
investment by suppliers is required, and such investment may be difficult to describe
accurately in a contract. These non-financial factors are generally referred to as non-
contractible factors. Buyers can pursue the improvement of supply chain performance by
supporting some incentives for investment on non-contractible factors, because buyers

cannot effectively claim the investment to suppliers by way of a contract. With the

19

 

suggestion that an emphasis on quality is the necessary condition for the successful
utilization of advanced manufacturing technologies and facilities, Milgrom and Roberts
(1990) stress that buyers should derive the investment on non-contractible factors from
suppliers by providing incentives for the investment in order to deal with changing
market situations because contracts and operating rules cannot always be altered.

Bakos and Brynjolfsson (1993) assert that the number of suppliers is inversely
proportional to supplier’s incentives for investment on non-contractible factors.
Furthermore, where the importance of quality, responsiveness and innovation is
increased, contracts having a small number of suppliers are the optimal choice. This
argument means that the investment on non-contractible factors can be considered as a
key factor for the decision of an optimal number of suppliers with transaction costs and
relationship-specific assets mentioned previously, thus implying the significant effect of
incomplete contract theory on the buyer-supplier relationship.

2.1.3 The Portfolio of Buyer-Supplier Relationship

The components used to determine the most suitable type of inter-firm transaction
relationships can be largely classified into four segments; the level of relationship-
specific assets, the level of uncertainty on transaction environment or partner’s behavior,
the level of complexity of inter-firm contracts, and the frequency of inter-firm
transactions. However, even though all four factors can be regarded as key dimensions
deciding the type of inter-firm relationship, previous research (Williamson 1979; Dyer
1996) has mainly pointed out that the level of relationship-specific assets is the most
important factor. As mentioned in the preceding sub-section, Williamson (1985) defines

relationship-specific assets as durable investment for supporting a specified transaction

20

relationship, or the opportunity costs of investment when using optimal alternative
transaction rather than the existing transaction or when suspending the existing
transaction. Such relationship-specific assets include mutually specialized physical assets,
human assets, R&D, and knowledge and capability of specified partners.

The type of inter-firm relationship suggested in transaction cost theory also focuses
on relationship-specific assets. One extreme of the relationship type is ‘pure market
structure’ as in the case of the transaction of standardized commodities. In pure market
structure, price is the most powerful tool for inducing the incentives of transaction and
the main criterion for the maintenance or suspension of a transaction. However, in the
case that the relationship-specific asset is strongly required and the size of the supplier
market is small, mutual adjustment on investment is advisable, and co-ownership on
transaction assets may be more effective. The other extreme of relationship type is
vertical integrative or hierarchical structure. This transaction type with unified ownership
is relevant in the case where the prevention of opportunistic behavior on relationship-
specific assets from partners and the close adjustment of decision-making between firms
are required. However, such a hierarchical structure may reduce the incentives of
partner’s profit maximization and induce additional bureaucratic costs relative to
decentralized structures (Milgrom and Roberts 1990). There exist various intermediate
inter-firm relationship types between pure market structure and hierarchical structure.
Such intermediate types include various kinds of contracts and partial ownership
contracts. In such a perspective, Bensaou (1999) suggest the use of a portfolio for the

inter-firm transaction relationship by utilizing the level of investment on relationship-

21

specific assets as classification criteria for the relationship type. Tables 2-2 and 2-3

organize the portfolio of the inter-firm relationship.

Table 2-2: Contextual Profiles

 

Supplier’s Specific Investment

 

 

 

Buyer’s
Specific
Investment

Low High
Captive Buyer Strategic Partnership
W Muct characteristics
OTechnicalIy complex 0High level of customization required
eBased on mature, well-understood 0Close to buyer’s core competency
technology OTight mutual adjustments needed in key
OLittle innovation and improvements to the processes
product OTechnically complex part or integrated
W subsystem

OStable demand with limited market growth
0Concentrated market with few established

eBased on new technology
Olnnovation leaps in technology, product, or

 

 

High players process
OBuyers maintain an internal manufacturing eFrequent design changes
capability OStrong engineering expertise required
met—Cm OLarge capital investments required
°Large supply houses Manic: characteristic;
0Supplier proprietary WMOIOS)’ OStrong demand and high growth market
OFew strongly established suppliers OStrong chry competitive and concentrated market
bargaining power OFrequent changes in competitors due to
Mutamakers heavily depend on these unstable or lack of dominant design
suppliers, their IOChNOIOSY 30d skills OBuyer maintains in-house design and testing
capability
Pager characteristics
OLarge multiproduct supply houses
OStrong supplier proprietary technology
0Active in research and innovation (i.e.,
R&D costs)
OStrong recognized skills and capabilities in
design, engineeringLand manufacturing
Market Exchange Captive Supplier
W95 Muct characteristics
OHighly standardized products OTechnically complex products
OMature technology OBased on new technology (developed by
OLittle innovation and rare design changes suppliers)
ochhnically simple product or well- dmportant and frequent innovations and new
structured complex manufacturing process functionalitics in the product category
OLittle or no customization to buyer’s final OSignificant engineering effort and expertise
product OHeavy capital investments required
0Low engineering effort and expertise MW
L 0W OSmall capital investments required 0High-growth market segment

 

M t characteri i s

OStable or declining demand
0Highly competitive market
OMany capable suppliers
OSame players over time
Supplier Qharacteg'stigs
OSmall ‘mom’ and ‘pop’ shops
0N0 proprietary technology
0Low switching costs

0Low bargaining power
08mm economic reliance on auto-business

 

OFierce competition

OFew qualified players

OUnstable market with shifis between
suppliers

Supplier chmterigics

OStrong supplier proprietary technology
OSuppliers with strong financial capabilities
and good R&D skills

0Low supplier bargaining power

OHeavy supplier dependency on the buyer
and economic reliance on the auto-sector

 

(Source: Bensaou, B.M., I999, Portfolios of Buyer-Supplier Relationships, Sloan Management Review 40(4), 35-44)

22

 

Table 2-3: Management Profile for Each Contextual Profiles

 

Supplier’s Specific Investment

 

Low

High

 

 

Buyer’s
Specific
Investment

Captive Buyer

lnforrnation sharing mechanisms’
0“Broadband” and important exchange of
detailed information on a continuos basis
OFrequent and regular mutual visits

Boundgt spanners’ task characteristics
OStructured task, highly predictable

OLarge amount of time spent by buyer’s
purchasing agents and engineers with
supplier

Strategic Partnership

lnforrnation-sharing mechanisms
0”Broadband,” frequent and “rich media"
exchange

ORegular mutual visits end practice of guest
engineers

Boundgm spanners’ tas_l_< charscteristics
OHighly ill defined, ill structured
0Nonroutine, frequent unexpected events
OLarge amount of time spent with supplier’s

 

 

LOW Climate and process characteristics staff, mostly on coordinating issues
OTense climate, lack of mutual trust Climate and process characteristics
0No early supplier involvement in design OHigh mutual trust and commitment to
OStrong effort by buyer toward cooperation relationship
OSupplier does not necessarily have a good OStrong sense of buyer fairness
reputation OEarly supplier involvement in design
0Extensive joint action and cooperation
OSupplier has excellent reputation
High Market Exchange Captive Supplier
lnfonnation-sharing mechanisms lnfonnation-sharing mechanisms
o“Narrow-bend” and limited information OLittle exchange of information
exchange, heavy at time of contract OFew mutual visits, mostly from supplier to
negotiation buyer

 

OOperational coordination and monitoring
along structured routines

Boundm spanners’ task characteristics
0Limited time spent directly with supplier
staff

OHighly routine aid structured task with little
interdependence with supplier’s staff
Climate and process characteristics

0Positive social climate

0N0 systematic joint effort and cooperation
0No early supplier involvement in design
OSuppIier fairly treated by the buyer
OSupplier has a good reputation and track
record

 

Boundm spanners’ task characteristics
OLimited time allocated by buyer’s staff to
the supplier

OMostly complex. coordinating tasks

Climate and process characteristics

OHigh mutual trust, but limited direct joint
action and cooperation

OGreater burden put on the supplier

 

(Source: Bensaou, BM, 1999, Portfolios of Buyer-Supplier Relationships. Sloan Management Review 40(4), 35-44)

2.1.4 The Effect of Information Technology on Buyer-Supplier Relationship

2.1.4.1 IS Effect in the Perspective of Transaction C ost Theory

The primary goal of buyers in pursuing a market exchange relationship is to

minimize cost and leverage economies of scale through large volumes relying on a large

number of suppliers. Buyers preferring a strategic partnership structure emphasize the

development of a close, long-term relationship with a few key suppliers. This means that

23

 

determining the optimal number of suppliers can be a natural extension of the "make
versus buy" or "markets versus hierarchies" decision. Both questions can be analyzed by
focusing on transaction costs.

Generally, in analyzing the effect of information technology on the number of
suppliers, researchers have centered on, in the viewpoint of transactional considerations,
investigating trade-offs between the increased costs necessary to search out a large
number of suppliers and the increased probability of finding a better price or a superior
product by reviewing a larger number of suppliers. As a result, it has been recognized
that technological developments reducing the cost of achieving information related to
prices and product characteristics should lead to an increase in the number of suppliers,
particularly in markets with differentiated goods (Bakos 1991). In other words,
information technology and inter-organizational information systems should lead to an
increase in the number of suppliers employed by buying companies because they tend to
reduce search costs (Malone 1985; Bakos 1987; Clemons and Row 1989; Bakos and
Brynjolfsson 1993).

Malone et al. (1987) support the above argument. They insist that because the nature
of coordination entails communication and processing information, the utilization of
information technology seems likely to lower coordination costs. Therefore, IT will
facilitate a shift from the single-supplier system within the firm that leads to hierarchical
structure to the multiple-supplier system that drives market structure. Also, they argue
that IT development heightens the capability of controlling the complexity of product
description through the improvement of inter-organizational information processing, and

that the introduction of a flexible manufacturing system increases the ratio of market

24

exchange structure by lowering transaction-specific assets and subsequently, entire
transaction costs.
2.1.4.2 IS Eflect in the Perspective of Incomplete Contract Theory

Even though there are theoretical arguments for a move from a single-supplier
system (hierarchical structure) to a multiple supplier system (market structure) (Johnston
and Lawrence 1988; Brynjofsson et al. 1991), an increase in the number of suppliers has
not been actually observed. On the contrary, there is recent evidence that there has been a
decrease in the number of suppliers, in spite of considerable reductions in transaction
costs due to IT development (Bakos and Brynolfsson 1993). This move to fewer suppliers
in the face of declining transaction costs suggests a paradox in the context of the
arguments relating to IT effect from the perspective of transaction cost theory.

Bakos and Brynjolfsson (1993) present two comprehensible explanations for such
shift to fewer suppliers. One is that transaction costs including search and coordination
costs have actually increased. If predetermined technological and organizational
investments are required when connecting with a supplier, a firm may want to restrict the
number of suppliers in order to cut down on these fixed costs. Similarly, if investments
on IT development are confined to a specific supplier and are not transferable to new
inter-firm relationships, switching costs involved in changing suppliers may restrict the
desirable number of suppliers. For example, investment on an inter-organizational system
for the exchange of component blueprints in a CAD format utilized by a specific supplier
may restrict the capability of the buyer to search out new potential suppliers.

Another explanation focuses on the benefits that smaller and tighter networks of

suppliers have on quality and responsiveness, innovation and technology adoption, defect

25

rates, trust, and information exchanges (Johnston and Lawrence 1988; McMillan 1990;
Cusurnano and Takeishi 1991; Helper 1991a, 1991b). All these characteristics include
investments by suppliers that are difficult or impossible to specify in advance on a
contract. Furthermore, they provide advantages that are confined to a specific buyer-
supplier relationship. Thus, offering incentives for such investments requires particular
tasks. IT is likely to increase the significance of non-contractible factors, such as
innovation, speed, responsiveness, and flexibility, stressing the necessity of these tasks.
When investments are non-contractible and specific, a buyer’s bargaining power
and/or the willingness of their suppliers to share the burden of non-contractible
investments are required. Current literature does not explicitly deal with the question of
how to suggest incentives for supplier investment in quality, responsiveness, and
innovation, and, particularly, how the number of suppliers influences the incentives to
derive such investments (Bakos and Brynojolfsson 1993). However, controlling the
number of suppliers, and thereby increasing supplier’s bargaining power to some degree,
as a competitive strategy, runs counter not only to standard neoclassical economic
models, but to widely used competitive strategy models as well (Porter 1980).
Conclusively, past studies indicate that it can be advantageous to restrict the number
of suppliers a buyer considers when providing incentives to induce suppliers to make
non-contractible investments. This argument implies that if a buyer does not have too
many alternative suppliers, a supplier can accumulate more benefits generated by non-
contractible investment relative to the individual buyer. Accordingly, the supplier will
have greater bargaining power and, therefore, more incentives to make non-contractible

investments, such as quality, responsiveness and innovation. This argument provides a

26

theoretical background for bringing out the necessity of a strategic partnership between
buyers and suppliers (Henderson and Venkatraman 1990; Johnston and Lawrence 1988),

which underscores the advantages of a closer relationship with fewer suppliers.

 

Advanced
IT Deployment
|
I
rm Cur ' W Cm
I /
The Decrease of I The Increase of
Transaction Cost : N on-Contractlble Investment
I
I
I
l
V
I l |
I I I
High # of Middle Low # of
Suppliers Suppliers

Figure 2-1: Relationship between IT Deployment and Buyer-Supplier Relationship

Viewed in this perspective, even if transaction costs decline radically due to

IT development, it can still be advisable for a buyer to restrict the number of

suppliers it employs. The analysis on the role of incomplete contracting

considerations in deciding the optimal number of suppliers provides not only an

27

alternative viewpoint to transaction cost theory, but also an additional argument

that information technology facilitates the "move to the middle." In other words,

the combination of decreased transaction costs and an increased need for non-

contractible investments leads to a shift to the intermediate level of supplier

numbers (Bakos and Brynjolfsson 1993). Clemons et al. (1993) support the

above argument commenting that IT plays a significant role in shifting the inter-

firm relationship to a middle ground between market and hierarchical structures

(Figure 2-1).

2.2 Information Technology and Supply Chain Dynamics

2.2.1 The Significance of Supply Chain Integration
A supply chain includes those activities necessary to produce and deliver products
and services to a customer; it may start with a supplier's supplier and end with a

customer's customer. Chopra and Meindl (2001) define SCM as the management of flows

between the stages in a supply chain to maximize total profitability. Such flows

include not only the flow of overall materials such as the procurement of raw

materials, the transformation into intermediaries, and the distribution of finished

28

products, but also information flow focused on ordering (Bowersox and Close

1996, Disney et al. 1997, Lee et al. 1997, Cooper and Ellram 1993, Hoekstra and

Romme 1991, Lambert et al. 1998). When successfully managed, a supply chain

delivers the right products, at the right time, at the right place, and at a

competitive price (Zheng et al. 2001). Vollmann et al. (2000) suggest that the

term 'supply chain' is no longer appropriate given the recent emphasis on

customer service. They prefer 'demand chain' and identify several management

issues including flawless execution, outsourcing, supply base development, and

partnership implementation. The literature in supply chain management has

addressed a wide spectrum of issues from the analytical design of production-

distribution networks to the management of these processes (Cohen and Mallik

1 997).

Two strategic issues in SCM are integration and coordination (Bowersox and Closs
1996; Lee et al. 1997). Supply chain integration links a firm with its customers, suppliers,
and other channel members by integrating their relationships, activities, functions,
processes, and locations. Such integration supports the current movement from

conventional, arms-length and often conflict-laden relationships to cooperative, long-term

29

business partnerships and strategic alliances (Morash and Clinton 1998). Integration
involves the design of the supply chain network to ensure that the segmented activities
associated with a dispersed network of suppliers can operate as a 'virtual' organization. In
spite of the increasing necessity of integrated SCM, the establishment of integrated SCM
cannot be accomplished in a short period, and thus requires a gradual and stepwise
approach (Stevens 1989; Wikner et al. 1991; Bowersox and Closs 1996). Coordination, in
its role as an intermediary stage for integrated SCM, is a process that manages the flow of
materials, information and funds among supply chain partners and may link decision
making across all nodes of the supply chain network.

Wikner et al. (1991) assert that coordination should be implemented in the first stage
to adjust the role and responsibility of individual supply chain participants, followed by
the integration of manufacturing functions in the second stage, the integration of
distribution functions in the third stage, and the integration of entire supply chains in the
final stage. These four integration stages are consistent with Bowersox’s integration
stages for SCM. Bowersox (1989) asserts that the process of supply chain integration
should progress from the integration of internal logistics processes to external integration
with suppliers and customers, and such internal and external integration can be
accomplished by the continuous automation and standardization of each internal logistics
function and by efficient information sharing and strategic linkage with suppliers and
customers. Stevens (1989), Byme and Markham (1991), and Hewitt (1994) express the
same view as Bowersox in that they assert that the improvement of each internal function
in the internal integration stage should precede the external connection with suppliers and

customers in the external integration stage. Stevens suggests four developmental stages in

30

supply chains; independent operation, functional integration, internal integration, and
external integration. Further, he asserts that a staged approach to SCM can remove the
barriers between ftmctions or organizations, and that IS utilization can strengthen the
linkage among functions and organizations.

Forrester (1961) advocated that the demand-supply unbalance between supply chain
parties is the most critical problem in integrated SCM. Lee et al. (1997) refers to this
distortion in information as the "bullwhip effect". The bullwhip effect is essentially the
phenomenon of demand variability amplification along a supply chain that includes
retailers, distributors, manufacturers, manufacturers’ suppliers, and so on. Lee et al.
characterized this phenomenon as demand distortion, which can create problems for
suppliers, such as grossly inaccurate demand forecasts, low capacity utilization, excessive
inventory, and poor customer service. Price fluctuation, order batching, and shortage
gaming between participants can be recognized as the key explanations of the bullwhip
effect (Lee et al. 1997).

As the demand distortion dilemma expands gradually to become a problem at the
inter-organizational level, supply chains need to be evaluated and managed as a unit in
order to be successful in the long term. This is consistent with the necessity of managing
various demand distortions in the integrative, long-term, and systematic perspectives
(Bowersox and Closs 1996, Hoekstra and Rome 1991, Jones et al. 1997).

2.2.2 The Effect of Information Technology on Supply Chain Dynamics

If a network of customers and suppliers is not effectively integrated and coordinated,

various predicaments such as ordering delays, lower quality, high in-process inventories,

long customer lead-times, and product obsolescence due to lengthy order cycles (Cohen

31

and Mallik 1997; Buxbaum 2000; Shore 2001) can occur. Bowersox and Closs (1996)
assert that the capability and eagerness of a firm's functional areas such as marketing,
purchasing, operations, and logistics to share critical planning and operational
information are preconditions for internal integration. Similarly, supply chain partners
should be able and willing to share important information to accomplish external
integration (Bowersox and Daugherty 1995). By sharing information, all supply chain
members can reduce the dependence on forecasts that too often display significant error.
Instead, firms may get very accurate estimates of customer demand when supply chain
partners successively exchange accurate, timely sales information (Closs et al. 1998).

As clearly described in F orrester's (1961) classic depiction of demand amplification,
the failure to share information with supply chain partners may lead to the substantial
amplification of a slight disturbance in demand at the retail level as it moves through the
channel. This view explains how poor information flow leads to substantial variation in
inventory holdings. That is, the failure to share important sales information with other
supply chain members causes under-stock of inventory in times of peak demand and
over-stock when demand subsides. Lee et al. (1997a, 1997b) define this distortion of
information as the "bullwhip effect". Christopher (1994) suggests that by connecting the
point-of-sale location directly to the point of production via information technology, the
"tidal wave effect" experienced by F orrester's supply chain participants can be softened
eventually in the face of demand variability. This direct linkage of point-of-sale locations
to other supply chain participants can be generally defined as "response-based" or "pull"
logistics systems. On the basis of these arguments, it has been recognized that supply

chain participants can accomplish both cost reduction and better customer service by

32

sharing information in a response-based logistics system (Bowersox and Closs 1996; Lee
et al. 1997a; Lee et al. 1997b; Closs et al. 1998).

Consequently, it is obvious that information sharing maintains a central role in the
integration and coordination of supply chains (Barrett 1986). Furthermore, it has
generally been accepted that such supply chain integration and information sharing may
effectively manage the 'bullwhip' effect in which demand amplification through the
supply chain sequence leads to inaccurate forecasts, low capacity utilization, excessive
inventory, and inadequate customer service, and thus significantly improve overall supply
chain performance (Lee et al. 1997a; Lee et al. 1997b). These two arguments mean that
by substituting inventory with information, the fundamental purpose of information
sharing can be met with the reduction of total costs enjoyed by each supply chain
member. Through analysis on the high technology industry, Lee et al. (2000) assert that
information sharing can significantly minimize the outcomes of the problems induced by
the bullwhip effect, thus supporting the above argument.

However, there exists the argument that, in some cases, it may be necessary to
change the way the supply chain is managed in order to make complete use of the
information flows. Chen (1998) studied the benefits of information flow in a multi-
echelon serial inventory system by computing the difference between the costs of using
echelon reorder points and installation reorder points. He observed that information
sharing benefits decreased with increase in demand variance, increased with increase in
the number of stages, and were lowest at moderate values of penalty cost. Moreover,
Grahovac and Chakravarty (2001) found that sharing and trans-shipment of items often,

but not always, reduces the overall costs of holding, shipping, and waiting for inventory.

33

Unexpectedly, these cost reductions are sometimes achieved through increasing overall
inventory levels in the supply chain. They conclude that even though the ability to
quickly move inventory within the lowest echelon can reduce the overall cost, this
savings may not be always accompanied by a reduction in the overall inventory in the
supply chain.

These opposing trends imply that the effect of information sharing can be diverse
depending on the detailed characteristics of the SCM system. Moinzadeh (2002)
determined that information-sharing is most beneficial in systems that exhibit the
following characteristics: (a) Systems where the supplier’s lead-time is long compared
with other lead-times (i.e., transit times from the supplier to the retailers) in the system;
(b) Systems where the number of retailers is not large (this is clearly a function of system
parameters, such as costs, demand rates, and lead-times); (c) Systems where the order
quantities are of average size; and, ((1) Systems where the ratio of the unit holding cost of
the retailers to that of the supplier is medium sized. This supports the above argument
that SCM characteristics play an important role when considering the effect of

information sharing.

34

   
   

TOTO! Total COIt

_———————- -————_——

 

 

Lot Size (Lead Time)

Figure 2-2: Traditional Trade-off Relationship between Inventory and
Transportation Costs

Cachon and Fisher (2000) also illuminate the dissimilar effects of information
sharing from a different perspective. They find substantial savings from lead-time and
batch size reductions, both of which are facilitated by the implementation of information
technology. They conclude that the observed benefits of information technology in
practice are due more to the impact of information technology on lead-time and batch
size than in facilitating information sharing. Many previous researchers assert that the
reduction of lead-time is an important factor in competitiveness (Stalk and Hout 1990;
Christopher 1992; Krajewski and Ritzman 2000). However, lead-time may also have

influences on other kinds of supply chain related activities. That is, the reduction of

35

supply chain lead-time can lead to the reduction of safety stock, order frequency, and
order batch size. However, that results in an increase in transportation costs. The
reduction of supply chain lead-time may induce trade-off relationships among various
supply chain variables (Magee et al. 1985) and furthermore it may not have any
significant effect on the reduction of total costs.

Even though information sharing is essential to the integration and coordination of
supply chains, it does not guarantee the improvements of all aspects of performance. This
means that the strategy of information sharing should be changed depending on whether
the focus of SCM is just minimizing specific cost or total costs by balancing among
various kinds of cost reduction, which kind of information should be considered for
sharing, whether the underlying demand process is complex or stationary, and what is the
specific structural characteristics of SCM system. The above argument also stresses the
necessity of discussions on the role of information technology and its capacity to change
the operating policies of a supply chain in order to maximize the effect of information
flows within the chain, and the external drivers that have significant impacts on the

efficient utilization of IT for information sharing.

2.3 Interactive Relationships among IT, Buyer-Supplier Relationship and SC
Structure

2. 3. 1 The Alignment among Product Customization Level, Supply Chain Structure, and
Buyer-Supplier Relationship

Even though the necessity of integrated supply chain management has consistently

been emphasized, research on the overall structure of supply chain, which is the starting

36

point for supply chain management, and theoretical relevance between such supply chain
structure and supply chain strategic issues, is at the beginning stage.

The most important strategic issue, which should be considered in the design of such
supply chain structure, is the characteristics and strategy of traded products (Fisher 1997).
Fisher (1997) asserted that traded products can be classified into efficient product and
innovative product, and supply chain appropriate for each product category should be
designed. Lassar and Kerr (1996) examined the relationship between a company’s supply
chain structure and primary product strategy such as differentiation, cost leader, and
focused strategies. Fisher (1997) emphasized the length and thickness of supply chain in
the design of supply chain, while Lassar and Kerr (1996) considered mainly the
partnership between manufacturer and distributor and the regional intensity of
distribution. Bensaou (1999) showed that the characteristics/strategy of traded product
can be closely related to the strength of buyer-supplier relationship, through the
suggestion of the portfolio of inter-firm transaction relationship by the level of
investment on the transaction-specific asset of buyer or supplier. That is, they suggested
that the type of market exchange relations in which the technical and economic
dependence of both buyer and supplier on each other is relatively low, is appropriate for
highly standardized products, while the type of strategic partnership in which buyer-
supplier relationship is strongly connected by considerable transaction-specific assets of
both buyer and supplier, is significantly correlated with highly customized products.
When considering the relationship between a company’s supply chain structure and
product strategy Fisher and Lassar/Kerr identified, such matching of product

characteristics/strategy and buyer-supplier relationship implies that the strength of buyer-

37

supplier relationship also can be the key strategic issue for constructing efficiently supply
chain structure. In fact, Anderson et a1. (1994) commented that depending on the position
and level of negotiation power between buyer and supplier, the overall design and type of
supply chain structure can be different, thus supporting the above argument.

Previous researches on the relevance between the tier of supply chain and strategic
issues (Eisenhardt 1985; Anderson and Gatignon 1986; Miller 1988; Lassar and Kerr
1996) have agreed that in case of standardized/cost-leader focused product appropriate
for market exchange relations structure, the necessity of controlling the number of
participants in a supply chain is relatively low due to the low weights of monitoring and
cooperation on supply chain, while in case of customized/differentiation focused product,
systematic monitoring and cooperation on supply chain is more emphasized in order to
support various types of services to customers. Thus, in case of
customized/differentiation focused product connected to strategic partnership structure,
the possibility that a firm attempts to reduce intermediate supply chain participants and
increase monitoring and trust among supply chain members is higher.

Also, Bakos (1991) suggests a meaningful argument on the relevance between the
number of suppliers and strategic issues. That is, in terms of buyers, in case of
standardized/cost leader focused product appropriate for market exchange relations
structure, the number of transaction-available supplier is expected to increase, because
the change of supply line is relatively easy due to the low level of dependence on
supplier, transaction—specific asset, and changeover cost. In case of customized/

differentiation focused product connected to strategic partnership structure, the

38

significance of increasing the number of supplier may be reduced, because the incentives
for improving non-contractible factors except price is more strongly required.

The necessity of supply chain integration can be also different depending on the type
of product strategy and buyer-supplier relationship. That is, in case of standardized/cost
leader focused product suitable for market exchange relations structure, managerial
independence may be guaranteed, because it is expected that result-based contract among
supply chain members is emphasized and the necessity of monitoring/promotion on
supply chain members is low (Lassar and Kerr 1996). Meanwhile, in case of
customized/differentiation focused product aligned with strategic partnership structure,
product life cycle is relatively short and various kinds of competitive factors (technology,
design, service) except product price should be supported. Also, the demand of finished
product is considerably unstable. Accordingly, buyer should support various types of
promotion activities (Winter 1993; Lassar and Kerr 1996), and accomplish high level of
cooperation or integration with other supply chain members by providing high margin
(Porter 1980; Anderson and Schmittlein 1984; Anderson and Gatignon 1986; Miller
1987, 1988; Lassar and Kerr 1996). Also, systematic training for differentiated
product/service and buying firm’s monitoring on distributor are more strongly required,
because there are a large amount of information which should be processed (Galbraith
1973) and a high possibility of decision-making’s delay (Govindarajan 1985). Therefore,
the possibility of behavior-based contract among supply chain members is very high
(Lassar and Kerr 1996). In other words, there is very much possibility that buyer pursues
high level of cooperation and integration with external supply chain members even with

enduring high burden of integration costs and political conflicts.

39

The approaches for reducing lead time also should be changed depending on the type
of product strategy and buyer-supplier relationship. In case of customized/differentiation
focused product and strategic partnership structure, the type of supplying immediately
products when necessary may be preferred to that of placing a large amount of
inventories on distributors or retailers, because margin rate and value added are high.
Thus, in order to pursue high customer service with low inventory level, quick response
logistics system should be constructed. Also, in case of customized/differentiation
focused product, a wide range of product lines should be completed and the necessity of
consistent product innovation and R&D is high. Because of such characteristics of
product variety, it is very difficult for distributors or retailers to hold and control properly
the inventories of various kinds of products in their warehouses or sales shops (Fisher
1997). Accordingly, in this case, it is advisable to reduce supply chain lead time in order
to deal with effectively immediate demand. However, in case of standardized/cost leader
focused product suitable for market exchange relations structure, it is expected that most
orders are made by the type of batch, because the responsibility of demand uncertainty
may be shifted onto other supply chain members through result-based contract among
supply chain members as mentioned previously. Accordingly, in case of such order with
batch type, supply chain lead time may be relatively long.

A manufacturing firm should try to establish an optimal response point by
considering the characteristics of traded products, product strategy, and related
environmental factors (Hoekstra and Rome 1991; Macbeth and Ferguson 1994; Jones et
al. 1997). In the case of standardized/cost leader focused product, the level of product

variety is not high (Harnbrick 1983), demand is stable (Miller 1988), and demand area is

40

generally wide (Porter 1980; Miller 1987). Also, the strength of buyer-supplier
relationship may be relatively weak because market governance structure is expected to
appear dominantly as mentioned previously. Thus, the likelihood that response point is
built in the upstream of supply chain is high. In the case of customized/differentiation
focused product, the level of product variety is high (Porter 1980; Miller 1987), demand
is unstable (Miller 1988). Also, the level of support and c00peration among supply chain
members for dealing with effectively demand uncertainty is relatively high (Anderson
and Gatignon 1986; Miller and Friesen 1986; Ward et al. 1996; Lassar and Kerr 1996),
because it is expected that hierarchical govemance structure by strategic partnership
between buyer and supplier and behavior-based contract among supply chain members
are mainly shaped (Eisenhardt 1985; Lassar and Kerr 1996). According to these
characteristics, firms with customized/differentiation focused product would try to obtain
information on final demand more quickly than firms with standardized/cost leader
focused product. Thus, the likelihood that response point is established in the downstream
of supply chain is strong.

In case of standardized/cost leader focused product, the characteristics of demand
stability and the availability of multi-suppliers by taking market governance structure can
induce the reduction of safety stock. Meanwhile, in case of customized/differentiation
focused product, safety stock may be maintained consistently as high level, because high
customer service is pursued, demand uncertainty is inherent, and the unexpected accident
of key supplier can lead to the difficulty of immediate response to customer orders when

pursuing excessively strategic partnership with key supplier.

41

2. 3.2 The Eflect of IT on the Alignment among Product Customization Level, Supply
Chain Structure, and Buyer-Supplier Relationship

As discussed above, the key issues for the design of supply chain structure may be
influenced by the characteristics/strategy of traded products and buyer-supplier
relationship. In other words, depending on which type of product strategy and buyer-
supplier relationship are combined with supply chain structural issues, the direction itself
for designing supply chain structure can be totally different, and further the effect of
supply chain structure on performance can be also diverse.

The introduction of B to B E-commerce can contribute to the efficiency
improvement of supply chain structure, because it can provide product information,
inventory level, shipping information, and information on customer requirements, on a
real-time pull basis (Radstaak and Ketelaar 1998). Particularly, because the buying
company can establish cooperative planning on demand forecasting and production
schedule with supplying company through information sharing by E-commerce, the
potential of E-commerce for the efficient design of supply chain structure is tremendous
(Karoway 1997). Additionally, E-commerce makes ‘pull’ supply chain management
possible by linking effectively each function in a supply chain and customer’s demand
information (Kalakota and Whinston 1997).

However, the introduction of E-commerce may somewhat change the type of
relevance between the width of supply chain and strategic issues which Bakos (1991)
argues. That is, like the argument of Malone et al. (1987) that the ratio of electronic
market structure to inter-firm relationship will be increased according to the development

of telecommunication technology, it is predicted that Web-based E-commerce adoption

42

may increase the number of supplier within a limited scope, even in strategically
important customized products. Therefore, we can expect that even though the effort for
maintaining cooperative relationship on strategically important customized products by
reducing the number of supplier will be continued, the intensity will be weakened
according to the introduction of Web-based E-commerce. Such kind of change also may
be indicated in the relationship between the length of supply chain and strategic issues.
That is, in case of customized/differentiation focused product emphasizing more
systematic monitoring and cooperation on supply chain by strategic hierarchical
partnership structure in order to support various types of services to customers, the
likelihood that a firm attempts to reduce intermediate supply chain participants is higher.
However, E-commerce adoption may decrease the necessity of reducing intermediate
participants compulsorily with enduring the burden of related costs by reducing the level
of dependence on physically contacted-monitoring and increasing the capability of real-
time information sharing. Consequently, the introduction of Internet or Web-based E-
commerce enables buyers to pursue moderate electronic market structure even on
strategically important customized products and maintain the proper number of
intermediate supply chain participants.

Such introduction of E-commerce also may have an influence on the relevance
between the level of supply chain integration and strategic issues. That is, E-commerce
adoption may make it possible for buying firms to obtain the effects of integration
simultaneously with guaranteeing somewhat the managerial independence of other chain
members, by supporting the capability of continuous and consistent remote

monitoring/promotion of supply chain members. Therefore, it can be foreseen that buying

43

firms with high level of E-commerce adoption do not need to pursue excessive suppiy

chain integration even on customized/differentiation focused product.

Table 2-4: The Effect of E-Commerce Adoption on Supply Chain Management

 

 

 

 

 

 

 

 

 

Structural The Effect of E-Commerce Related
Issues Literatures
No. 0] Because the ratio of electronic market structure to inter-firm relationship will be Malone et al.

Su liers increased according to the development of telecommunication technology, it is (1987)

pp predicted that Web-based E-commerce adoption may increase the number of
supplier within a limited scope, even in strategically important customized
products.

No. of SC EC adoption may decrease the necessity of reducing intermediate participants Malone et al.
Paths compulsorily with enduring the burden of related costs by reducing the level of (1987)

dependence on physically contacted -monitoring and increasing the capability of
real-time information sharing, even in strategically important customized
products.

The Level By acquiring the capability of continuous and consistent remote monitoring/ Lassar and
of SCI promotion on supply chain members, it can be foreseen that buying firms with Kerr (I996)

high level of E-commerce adoption do not need to pursue excessive supply chain
integration even on customized/differentiation focused product.

Order Even firms with customized/ differentiation focused product may pursue the shift Krajewski and
Penetration of response point into more upstream level of supply chain by improving the Ritzman (2000)
Point capability of remote monitoring and precise prediction on demand information, Berry et aI.

product variety, and market situation through the utilization of E-commerce. (1994)
Lead Time The development of E-commerce technology can lead to the consecutive Fisher (1997)

reductions of total transportation cost and total inventory cost by grafting

advanced information technology onto logistics and inventory processing

activities. Such consecutive reductions should shift the minimum point of total

cost into left and down side. Therefore, the development of EC may enable

buying firms to accept a little longer lead time, even in customized products.

Safety Stock If both the shift of response point into more upstream level and the construction Hoecstra and
of automatic quick-response logistics system can be accomplished successfully Romme (1991)
through the utilization of lntemet or Web-based E-commerce as mentioned Jones et al.
previously, even firms with customized/ differentiation focused product may (1997)
expect the benefits from the reduction of safety stock. Bowersox and

Closs (I996)

 

 

 

Also, the development of E-commerce technology requires the change of concept on
trade-off relationships among lead time, distribution frequency, inventory level, and
transportation cost, which has been accepted traditionally. That is, the development of E-

commerce technology can lead to the consecutive reductions of total transportation cost

44

and total inventory cost by grafting advanced information technology onto logistics and
inventory processing activities. Such consecutive reductions should shift the minimum
point of total cost into left and down. Therefore, the development of E-commerce may
enable buying firms to accept a little longer lead time relative to the existing lead time in
terms of minimtun total cost. According to this logic, it can be anticipated that if the level
of E-commerce adoption is relatively high, the significance of reducing the lead time of
supply chain will be less even in customized/differentiation focused product and strategic
partnership structure.

The utilization of E-commerce may change traditional perspectives on the response
point of supply chain. That is, lntemet or Web-based E-commerce adoption can improve
the capability of remote monitoring and precise prediction on demand information,
product variety, and market situation. Accordingly, it is expected that even firms with
customized/differentiation focused product may pursue the shift of response point into
more upstream level of supply chain through the utilization of E-commerce. This makes
it possible for manufacturing firms to accomplish the proper balance between the shift of
order penetration point into upstream for quick switch to new product for dealing with
effectively demand uncertainty and the maintenance of quick response capability on
customers, which Berry et al. (1994) emphasizes.

Such argument on the relationship between E-commerce adoption and the response
point of supply chain also can explain the effect of E-commerce on the relationship
between safety stock and the characteristics of traded product or buyer-supplier
relationship. If order penetration point is shifted to upstream by E-commerce adoption as

mentioned previously, overall inventory in a supply chain is expected to reduce, but

45

average safety stock in a supply chain may increase in order to establish quick-response
logistics system. This is because the weight of push-type management will be decreased,
while the weight of pull-type management is anticipated to increase (Hoekstra and
Rome 1991; Jones et al. 1997). However, if both the shift of response point into more
upstream level and the construction of automatic quick-response logistics system can be
accomplished successfully through the utilization of Internet or Web-based E-commerce,
even firms with customized/differentiation focused product may expect the benefits from
the reduction of safety stock. This means that traditional perspectives on the relationship
between safety stock and strategic issues also can be changed depending on the level of
E-commerce adoption. Of course, in order to derive the above argument, the analysis on
trade-off relationship between additional costs and improved benefits induced by the
change of response point and the reduction of safety stock in the perspective of total cost
minimization should be preceded (Bowersox and Closs 1996). Generally, it is predicted
that under total cost minimum approaches, improved benefits are superior to additional

COSIS.

46

CHAPTER 3. RESEARCH FRAMEWORK

The literature review in the preceding chapter identified the limitations and gaps in
several research streams that are related to the effect of information technology adoption
on buyer-supplier relationship and supply chain dynamics. The gaps and limitations in
literature helped identifying research questions and the potential solutions to these
questions in this dissertation. This chapter describes a research fiamework for the study

derived from literature review.

3.] Conceptual Framework

As discussed in the literature review, there are two contradictory arguments in the
effect of IT on buyer-supplier relationship. Transaction cost theory suggests that
information technology and inter-organizational information systems can reduce
transaction-specific asset and the cost of achieving information about prices and product
characteristics, and such reduction should lead to an increase in the number of suppliers,
particularly in markets with differentiated products. However, studying the role of
incomplete contracting considerations in determining the optimal number of suppliers
provides that, despite recent declines in search costs and coordination costs due to
information technology development, firms do not want to increase the number of
suppliers (Bakos and Brynjolfsson 1993). This is because information technology
development may increase the importance of non-contractible investments by suppliers,

such as quality, responsiveness, and innovation in the perspective of incomplete contract

47

theory, thus bringing out the necessity of close buyer-supplier relationship. Accordingly,
when such investments are particularly important, firms will employ fewer suppliers, and
this will be true even when search and transaction costs are very low. Such conflict
between transaction cost theory and incomplete contract theory implies that IT has a
significant role in shifting inter-firm relationship structure to the middle between market
and hierarchical structures (Clemons et al. 1993). This also emphasizes that other forces
should be accounted for in a more complete model of buyer-supplier relationship.

Bensaou and Venkatraman (1996) suggest the likelihood that product customization
level can be a critical factor for explaining such complete model of buyer-supplier
relationship. They emphasized that the type of market exchange relationship in which the
technical and economical dependence of both buyer and supplier on each other is
relatively low, is appropriate for highly standardized products. Meanwhile, the type of
strategic partnership in which buyer-supplier relationship is strongly connected by
considerable transaction-specific assets of both buyer and supplier, is significantly
correlated with highly customized products.

Also, Fisher (1997) asserted that such product customization level should be
considered as one of the most important strategic issue for the design of supply chain
structure. That is, he emphasized that traded products can be classified into efficient
product and innovative product, and supply chain appropriate for each product category
should be also classified into physically efficient supply chain and market responsive
supply chain. Lassar and Kerr (1996) examined the relationship between a company’s
supply chain structure and primary product strategy such as differentiation, cost leader,

and focused strategies. When considering the relationship between buyer-supplier

48

relationship and product strategy Bensaou and Venkatraman suggested, such matching of
product characteristics/strategy and supply chain structure Fisher identified implies that
supply chain structure can be also an important factor for explaining the complete model
of buyer-supplier relationship. In fact, Anderson et al. (1994) commented that depending
on the position and level of negotiation power between buyer and supplier, the overall
design and type of supply chain structure can be different, thus supporting the above
argument.

Also, much prior literature (Eisenhardt 1985; Hoecstra and Rome 1991; Bakos
1991; Lassar and Kerr 1996; Fisher 1997; Jones et al. 1997) notes that the relationship
between product customization level and buyer-supplier relationship can influence supply
chain structural issues, and also inversely, depending on such effect on supply chain
structural issues, corporate SCM strategies on product customization and buyer-supplier
relationship can be changed in terms of performance improvement.

All of the above arguments imply that information technology deployment for the
establishment of efficient supply chain structure should be implemented under the
consideration on the interactive feedback relationships with the characteristics of traded
products and buyer-supplier relationship. This stresses that the role of information
technology should be discussed from the perspective that information flows in supply
chains can be better utilized by considering 1) various operating environments and
characteristics in a supply chain, 2) the characteristics of traded products, and 3) buyer-
supplier relationship.

Also, the above argument suggests that firms can acquire the competitive and

economic benefits of advanced information technology, specifically, E-technology not

49

from the effect of technology itself, but from managerial capability on such technology
(Keen 1993). This provides a persuasive reason on why there exists a gap between
practical implementations and theoretical recommendations about the effect of
information sharing on supply chain dynamics. That is, it has generally been accepted
that information sharing by advanced information technology can manage effectively the
'bullwhip' effect and minimize the outcomes of demand distortion due to such bullwhip
effect (Lee et al. 1997a; Lee et al. 1997b; Lee et al. 2000). However, according to some
previous researchers, the benefits of information sharing decrease with increase in
demand variance and may not be always accompanied by a reduction in the overall
inventory in the supply chain. Also, the observed benefits of information technology in
practice are derived not from facilitating information sharing itself, but from the impact
of information technology on supply chain structural issues such as lead time and batch
size (Chen 1998; Cachon and Fisher 2000; Grahovac and Chakravarty 2001). These
opposing trends imply that the effect of information sharing can be different depending
on the detailed characteristics of a SCM system and how such characteristics are
managed, thus emphasizing the significance of managerial capability on information
technology (Moinzadeh 2002).

Viewed in this perspective, strategic alignment among key SCM strategic and
structural issues, and advanced IT deployment, should be regarded as the most significant
and urgent research theme for the construction of an effective SCM strategy. This study
attempts to suggest the shape of such effective SCM strategy through the development
and testing of a framework for investigating the relationships among information

technology, product customization level, buyer-supplier relationship, and supply chain

50

structural issues. Figure 3-1 indicates the conceptual framework based on the selected

arguments discussed above.

   
    

 
 

Eisenhardt (1985),
Hoecstra & Rome (1991),
Bakos (1991), Jones et al. (1997)

Information Firm

TeChnOIOQ Transaction Cost Bakos & Performance

Deal 5 cut Theory, Brynjolfsson (1993),

Incomplete Contract Clemons et 31 (1993)
Theory

 
 

  
    
 
 
      
  
   

  
    

. . Chen (1988),
:' E Buyer-Supplier Cachon &
EMasec 1 Fisher (2000)
? et ,1, 2 Relationship Gr ahovac &
3 (1985) 5 Lee Chakravarty
‘ : (1997a, 1997b), Anderson (2001).
3 Lee et a1. (2000) et al. Mmmdeh
, ;' (1994) (2002)
i :: Bensaou &
" Venkatraman (1996) ‘
Product S
upply Chain
Customization Level Structure
Fisher (1997),
Lassar & Kerr (1996)

Figure 3-1 : Conceptual Framework

One thing to be noted is that, as can be seen in the above figure, the relationship
between information technology deployment and product customization level is indicated
not by a solid, but by a dotted line. This is because it was relatively difficult to find

related previous research supporting the relationship. However, this relationship also

deserves further investigation.

51

As mentioned previously, previous literature comments that there exists the trade-
offs among supply chain structural issues (Magee et al. 1985; Pine et al. 1993). For
example, as mentioned previously, if lead-time is reduced and distribution occasions are
frequent, overall inventory level is decreased, while transportation is increased. However,
the deployment of advanced information technology may change traditional concept on
such trade-off relationships. That is, the deployment of advanced IT may enable buying
firms to accept a little longer lead-time relative to the existing lead-time in terms of
minimum total cost. Accordingly, if a firm’s IT adoption level is relatively high, the
firm’s effort for reducing the lead time of supply chain may be less required even in
customized/differentiation focused product. This suggests the existence probability of the
relationship between information technology deployment and product customization

level. This research will investigate such probability.

3.2 Research Scope

Supply chain management is concerned with "the flow of material, work-in-process,
and finished inventory." (Bowersox and Closs 1996) Thus, the demand for logistics
services is connected to the shift of inventory throughout the supply chain. From the
manufacturer's perspective, finished good orders stimulate the need for outbound logistics
services from manufacturers to customers. Raw materials, components, or resale goods
create the need for inbound logistics services to manufacturers.

This paper confines the scope of this study to inbound logistics, because this study

focuses on Business-to-Business IT adoption with the consideration of inter-

52

organizational buyer-supplier relationships. There are many reasons, both operational and
strategic, why the concentration on inbound logistics is so important. First, the firm is
taking a holistic view of its own role in the supply chain. But, further, it enables the firm
to seize initiative in the supply chain. Information technology's effect on inventory
management can also be observed in the perspective of effective inbound logistics. The
technology is utilized to manage the acquisition of the right amount of parts at the right
place at the right time. Such efforts can result in the elimination of the disadvantages of
carrying excessive inventory or stocking out of raw materials. The capability to
effectively manage inbound logistics by information technologies can lower the need for
safety stocks, inventory cost, and disturbances in production due to a lack of raw
materials. Indeed, it leads to a higher level of service and greater customer satisfaction.
This study does not consider detailed manufacturing capability characteristics such
as production planning, capacity, process type, and workforce as input variables for the
System Dynamics model. In addition, beyond the analysis on the benefit of IT adoption
just in terms of cost savings, this study captures the benefits in various perspectives such
as inventory or safety stock reductions, demand amplification, order penetration point,

lead-time of supply chain, product price, and investment.

53

CHAPTER 4. RESEARCH DESIGN

4.1 System Dynamics

In order to fully address the above mentioned research questions and objectives,
more sophisticated models must be constructed and more comprehensive simulations
must be carried out. Thus, this dissertation will develop a dynamic simulation model of
advanced IT deployment considering interactive feedback relationships with product
customization level, buyer-supplier relationship, and supply chain structure.

This research uses the System Dynamics method. In the 19605, Massachusetts
Institute of Technology Professor Jay W. Forrester (1961) created Systems Dynamics, a
method of studying the world around us by viewing the system as a whole. Systems
Dynamics examines the interaction of all objects or individual parts of a system and their
relationship to one another. Basic system structures are assessed to better understand the
cause and effect that may be produced. Many of these systems can be built as
computerized models and are able to perform reliable calculations at a much greater
speed than the human-mind based model.

Even though the application scope and focus of System Dynamics have changed
continuously, the typical methodological characteristics of System Dynamics have still
remained strong during the last 40 years. The first characteristic is that System Dynamics
focuses on the dynamic behavior of systems, that is, the behavioral changes of systems
according to the progression of time. This implies that System Dynamics emphasizes

practical perspectives such as the change, evolution, development, and decline of

54

systems. The second feature of System Dynamics is that it analyzes the fundamental
reasons of dynamic change through feedback structure. Feedback structure means that a
closed loop is established by linking causal relationships among variables (Richardson
1991). Emphasizing feedback loop indicates that the dynamic change of a system is
analyzed through endogenous variables rather than exogenous variables. Explaining the
change of a system by exogenous variables makes it difficult to alter the behavior of a
system strategically. However, the utilization of endogenous variables makes it possible
to change the behavior of a system within a model. Another strong point of the feedback
structure is to analyze the change of a system in terms of the overall structure of the
system rather than the change of parameter related to a specific variable. These
characteristics precisely correspond to the objective of this dissertation as mentioned
above, which is the identification of dynamic changes of IT effect by investigating
interactive feedback relationships with product customization levels, buyer-supplier
relationships, and supply chain structure. This provides the reason for using System

Dynamics as the methodology to pursue the objectives of this dissertation.

4.2 System Dynamics Modeling Process

The simulation modeling process by the System Dynamics method can be largely

classified into three stages; establishing a causal loop diagram, designing a System

Dynamics diagram, and formulating an equation.

Stage 1: Establishing a causal loop diagram

55

A causal loop diagram is the map identifying the feedback structure of a system and
organizes the cause and effect that may be produced in a system by indicating the
feedback structure on a two-dimensional diagram. A causal 100p diagram is established
by a set of causal propositions or hypotheses on the shape of relationships among selected
construct variables.

Stage 2: Designing a System Dynamics diagram

A System Dynamics diagram is the map quantitatively actualizing the feedback
relationships derived from a causal loop diagram through the consideration of two
concepts on system behavior such as system state and system activity. System state is
indicated as the values of all variables constructing a system at time t. Such values are the
changing values obtained as a result of activities occurred between time t-l and t. Also,
information on such system state leads to the change of future activity by feedback.

For example, consider the flow of a product through a supply chain from producer to
customer. The product shifts through a network of stocks (inventories) and flows
(shipment and delivery rates). System dynamics diagram for this process is indicated in
the bottom of Figure 4-1. As can be seen in the figure, production starts add to the stock
of work in process (WIP) inventory. The production completion rate decreases the stock
of WIP and increases the stock of finished inventory. Shipment to customers decreases
finished inventory. In other words, WIP inventory increases when production start rate
exceeds production completion rate, and finished inventory also increases when
production completion rate exceeds shipment rate. Causal loop diagram indicated in the
top of Figure 4-1 makes it difficult to check such physical flow of product through the

system and the conservation of material in the stock and flow chain. The establishment of

56

system dynamics diagram can supplement such pitfall of causal loop diagram to indicate

the stock/flow distinction.

+ -

Production Start Work in Process Production
Rate Inventory Completion Rate

+-

F in'shed Invemory Shipment Rate

Causal Loop Diagram for a Manufacturing Process

 

 

 

 

 

CD
as

 

E} E > Big esfls1 E" $ IFrmshed #8
Production Start Inv ‘7 Production nventory Shipment Rate
Rate Completion Rate

 

 

 

 

 

 

System Dynamics Diagram for the Manufacturing Process

(Source: Sterman, 2000)

Figure 4-1: Causal Loop Diagram vs. System Dynamics Diagram

Stage 3: Formulating an equation

In this stage, equations corresponding to system state and system activity indicated in
the System Dynamics diagram are formulated by computer simulation language. Such
formulation of equations is completed through supplementation by several test runs of the

simulation model.

57

4.2.1 Causal Loop Diagram

The explanation on the causal loop diagram in this dissertation starts from the impact
of IT on buyer-supplier relationships. In the perspective of transaction cost theory,
Malone et al. (1987) and Bakos (1987) insist that IT will facilitate a move from single-
supplier arrangements within the firm ("hierarchies") to multiple supplier arrangements
("markets") because it reduces transaction costs with suppliers. According to this logic,
technological developments lowering the cost of acquiring information about prices and
product characteristics in a given market may reduce the excessive dependence on few
key suppliers by reducing transaction-specific assets and subsequently entire transaction
costs, and this should lead to an increase in the number of suppliers considered.

Such a shift to multiple supplier arrangements can provide greater ex post bargaining
power to a buyer, because a buyer can have many alternative suppliers. Also, such a
buyer’s increasing bargaining power can deduce the decrease of material price, thus
leading to the increase of a buyer’s profit. If a buyer’s profit is increased, the buyer can
increase the investment for advanced IT deployment, thus making a connection to the
increase of IT level. Figure 4-2 indicates a feedback loop reflecting the above argument.
The feedback loop in Figure 4-2 shows that there is a mutual positive feedback
relationship among IT level, the number of suppliers and buyer’s bargaining power in the
perspective of transaction cost theory. In other words, continuous advanced IT
deployment ultimately can derive complete competition in the electronic market among
numerous suppliers, and such an electronic market structure can provide the capability

enabling pursuit of the deployment of more advanced IT to the buyer.

58

+

k—\
Investment
' ft
Ca ability for ITD Buyer s Pro 1
+
Information Technology Raw Material Price
Deployment

67‘

Investment on Asset
Specificity

+

Transaction Cost

Buyer's Bargaining
Power

+

No. of Suppliers

V

Figure 4-2: Causal Loop Diagram for IT effect on Buyer-Supplier Relationship
in the perspective of Transaction Cost Theory

However, the perspective of incomplete contract theory suggests a paradox of such

positive feedback relationship among IT level, the number of suppliers, and buyer’s

bargaining power under the viewpoint of transaction cost theory. That is, the decrease of

material price induced by the increase of buyer’s bargaining power obviously leads to the

increase in buyer’s profit, but simultaneously drives a decrease in supplier’s profit. Such

decrease in supplier’s profit may reduce incentives for suppliers to make non-contractible

investments in areas such as quality, innovation, speed, responsiveness, and flexibility.

The reduction of supplier’s incentives to make non-contractible investments increases the

burden on the buyer to make non-contractible investments. As a result, transaction costs

may actually increase in spite of the decline of investment on transaction-specific assets

59

due to IT development. Accordingly, buyers must depend on close relationships with
suppliers to reduce the burden of increasing transaction costs, and this should lead to a
decrease in the number of suppliers considered. Also, such shift to fewer key supplier
arrangements provides great bargaining power to suppliers, and this increase can induce
an increase in material price, subsequently followed by a decrease in buyer’s profit, the
decrease of investment for IT deployment, and ultimately the reduction of desired IT
level. Conclusively, the above argument means that there is a negative feedback
relationship among IT level, the number of suppliers, and buyer’s bargaining power in
the perspective of incomplete contract theory unlike in the perspective of transaction cost
theory. In other words, continuous IT deployment cannot lead to a persistent increase in
the number of suppliers, and inversely, such stagnation in the number of suppliers
prevents the continuous improvement of IT level, and thus each factor consisting of
feedback loop is self-regulated and stabilized.

Actually, Bakos and Brynjolfsson (1993) assert that IT is likely to increase the
importance of non-contractible factors, and Clemons et al. (1993) note that a positive
relationship between IT level and non-contractible investment shifts the structure of
buyer-supplier relationship to the middle ground between market and hierarchical
structures, thus supporting the above argument (See Figure 4-3).

Negative feedback loop can also be found in the effect of IT on supply chain
dynamics. Information sharing by the utilization of advanced IT can reduce distortion in
demand information between buyer and supplier, which is referred to as the "bullwhip
effect" (Lee et al. 1997; Christopher 1994). Such decrease of demand variability enables

accurate forecasting, and this can lead to the decrease of target raw material (RM) level.

60

This lowering of target RM level can deduce the reduction of periodic RM order quantity,
subsequently followed by the decline of RM inventory level and finished good (F G)
inventory level. This logic is consistent with the argument of previous researches (Lee et
al. 1997a; Lee et al. 1997b; Lee et al. 2000) asserting that information sharing can

significantly minimize the problem of excessive inventory.

+

M

lnvestrnent Buyer’s Profit

Capability for ITD

Supplier's +
9, Non-contractible ‘\
+ -

Investment Supplier's Profit
.4.

Information Need for Buyer's

Technology‘_> Non-contractible Raw Material Pu.“
Deployment\_> Investment

C,+ B r’ Economies of Scale
uye s
Non-contractible - Ba Buy Cf;
Investment rgarnmg ower

Investment on

Asset Specificity +

4.

Transaction Cost 'N‘L 0fSUPI’IICI‘S

Figure 4-3: Causal Loop Diagram with IT effect on Buyer-Supplier Relationship
in the perspective of Incomplete Contract Theory

However, as Grahovac and Chakravarty (2001) mentioned, decreasing overall

inventory levels in the supply chain may not always be beneficial. In other words, in

61

order to reduce overall costs and maximize profit, increasing inventory levels may be
necessary. This is due to the fact that there is a probability that too low a level of
inventory results in under-stocks of inventory in times of peak demand. Such probability
increases the need for safety stock. Particularly, in the case of customized/differentiation
focused product in which the level of product customization and variety is high (Porter
1980; Miller 1987), and demand is unstable (Miller 1988), the need for safety stock may
be higher, because high customer service and responsiveness is required. Such need for
safety stock brings the increase of production quantity, and subsequently this should lead
to the increase of RM backorder necessary for the production of required quantity. When
considering that this sequence is initiated from a low inventory level, the increase of RM
backorder means that demand for RM is greater than the supply capability of RM, thus
leading to the increase in RM pricing. As mentioned in the feedback loop for IT effect on
buyer-supplier relationships, the increase of RM price subsequently leads to the decline
in buyer’s profit, the reduction of investment capability for IT development, and
ultimately the reduction of desired IT level (See Figure 4-4).

Negative feedback loop representing IT effect on supply chain dynamics in Figure 4-
4 indicates that information sharing by the utilization of advanced IT may not guarantee
persistent cost reduction or profit increase through entire supply chain sequence. As
mentioned previously, these opposing trends imply that the effect of inforrnation-sharing
can be different depending on the detailed characteristics of the SCM system. Chen
(1998) observed that information sharing benefits decreased with an increase in demand

variance, thus supporting the above argument.

62

:33: :55 mom—:23: 52—0 baasm a: Beta .5 5?» Banana nee.— ..aaaU 319 9...»:—

   
     

_o>3 .0053 Snook—
/ “\III _o>o.._ boEo>£ 2m
+ /
b58683 BEER. $4.3 b25>£ on. + b.5550
EEO SE 28.5..
/ I O /
+
xoofi I
>8me com 302 boa—.02: SE .0th
+

+

€230 cocoavoi
am can

:2 5953 scams—8.2.
- used .8 SE35 2:

‘llllll .
+ + eaaaam do .2 a8 8.83:5 + .
+
hop—038m £530.“ m:==a /

{gem i 22:32,:— bBEooam .82

03.68.5952 8 .5585. .9829.

28m .6 8_Eocoom {exam 23:60 .5 £3.35

+ - \ + - 5.3 msazm cos—«8.8:—
ootm 3.632 33. _ +
22582.
escoggoéoz
+ $5.95 com 302 +
E256. on
Eon—~82: . Do 0: oo
oaco>om {flan—2m 03:35—89:02 cosh—phat".
. {1' $2396 .
+

+

E2; {can 0

o: be £388
Ens—Eng:—

63

The effect of IT investment in supply chain management can differ depending on the
interaction among feedback loops representing IT effect on buyer-supplier relationships
and reflecting the effect of IT on supply chain dynamics. In other words, depending on
how the described feedback loops affect each other or which feedback loop is most
dominant, the effect of IT investment can vary.

4.2.2 System Dynamics Diagram
4. 2. 2.1 Overview of System Dynamics Diagram

In order to simulate the causal loop diagram described above by computer, two kinds
of computer program are required. First is simulator, which runs system dynamics model
by computer and suggests the results by table or graph. Second is simulation program,
which represents the specific operating principles of system as a system dynamics model.
Generally, simulation program is made under the simulator. This simulation program is
called the system dynamics diagram.

This dissertation uses VENSIM, one of system dynamics software, in order to make
system dynamics diagram based on the causal loop diagram described above. System
dynamics uses a particular diagramming notation for variables (Figure 4-5) (Sterman
2000). Variables represented by rectangles are stocks. Stocks are accumulations. They
describe the state of the system and create the information on which decisions and actions
are based. Stocks give systems inactivity and provide them with memory. Stocks generate
delays by accumulating the difference between the inflow to a process and its outflow.
Pipes pointing into and out of the stock indicate inflows and outflows respectively.
Valves manage such flows. By decoupling rates of flows, stocks are the source of

disequilibrium dynamics in systems.

64

 

A Stock variable (Level variable)

 

 

 

A Auxiliary variable
——§ Flow of Material
————-v Flow of Information

X Valve (Flow Regulator)
{2) Source or Sink

(Stocks outside model boundary)

Figure 4-5: System Dynamics Diagramming Notation

The difference between stocks and flows can be recognized in many regulations and
various areas (Table 4-1). In order to discriminate more clearly between stocks and flows,
we can use the snapshot approach which takes a picture on a system. Stocks would be
those things we can count or measure in the picture, including psychological states and
other intangible variables. For example, if taking a picture on river, something we can
quantify in the picture is the stock of water. We cannot check whether the water level is
increasing or decreasing. Similarly, if taking a picture on workers within a firm, we can
count the number of workers, but cannot count the number of hired or fired workers. Like
this, stocks represent the state of system at a particular point, while flows mean the
change of system state.

Clouds in Figure 4-5 indicate the sources and sinks for the flows. A source indicates

the stock from which a flow generating outside the boundary of the model comes up. A

65

sink indicates the stocks into which flows leaving the model boundary. It is assumed that
sources and sinks have infinite capacity and cannot restrict the flows they support.
Auxiliaries are the variables which are used as a intermediate variable in order to clarify
and simplify the calculation of flow variables. Such auxiliary variables consist of
functions of stocks (and constants or exogenous inputs) (Sterrnan 2000). These variables

set the boundaries for the simulation model.

Table 4-1: Terminology for Identifying Stocks and Flows

 

 

 

 

 

 

 

 

 

 

 

Area Stocks Flows
Mathematics, Physics Integrals, states, state Derivatives, rates of change,
and Engineering variables, stocks flows
Chemistry Reaction products Reaction rates
Manufacturing Buffers, inventories Throughput
Economics Levels Rates
Accounting Stocks, balance sheet items Flows, cash flow or income
statement items
Biology, physiology Compartments Diffusion rates, flows
Medicine, Prevalence, reservoirs Incidence, infection,
epidemiology morbidity and mortality rates

 

(Source: Sterman, 2000)

4. 2. 2.2 System Dynamics Model of this Dissertation
Appendices 1-3 indicate the System Dynamics diagram of this dissertation.
Appendix 1 is a System Dynamics diagram representing the effect of IT investment
on buyer-supplier relationships, while Appendix 2 reflects the effect of IT investment and
information sharing on supply chain dynamics. Also, Appendix 3 shows the diagram for
analyzing the effect of IT investment derived from both diagrams in Appendices l and 2,
in terms of profit. Even though three System Dynamics diagrams are detailed in

Appendices 1-3, they are all connected to each other by the shadow variable, which

66

 

means the copied variable of the original variable. For example, the IT level can be found
in both appendices I and 2. The IT level in Appendix 2 is the shadow variable of that in
Appendix 1, or in other words, the identical variable. Thus, both diagrams in Appendices
l and 2 are connected by this IT level variable. Similarly, the F G sales in Appendix 1 are
the shadow variable of that in Appendix 2, thus connecting both diagrams in Appendices
1 and 2. In addition, the profit variable in Appendix 1 is the shadow variable of that in
Appendix 3, and buyer’s revenue, RM price, RM shipping, F G sales, FG inventory, and
RM inventory in Appendix 3 are the shadow variables of those in Appendix 1 and 2. By
connecting the diagrams by shadow variable, a complete feedback loop is established.

The diagram in Appendix 1 representing the effect of IT investment on buyer-
supplier relationships was totally created by this dissertation based on previous literatures
(Bakos and Brynjolfsson 1993; Clemons et al. 1993) as mentioned above. In the case of
the diagram in Appendix 2 reflecting the effect of IT investment and information sharing
on supply chain dynamics, this dissertation modified Kim’s two-level information
sharing model (1998) on the basis of Sterman’s (2000) and Coyle’s (1977) supply chain
dynamics models. This is because Kim’s model was interpreted as the model representing
most effectively the sequence and effect of information sharing in the supply chain
among various supply chain dynamics models. Also, the diagram in Appendix 3
analyzing the effect of IT investment in terms of profit was modified from the model
suggested in Sterman’s ‘Business Dynamics’ (2000).

One thing to be noted is that cost variables in the diagram of Appendix 3 are not
indicated in the causal loop diagram. This is because such variables in this research are

not mainstream, which can be backed up theoretically by previous literatures. So, for the

67

sake of brevity, we omit presentation of such cost variables in the causal loop diagram.
However, we know that a buying firm’s profit can be influenced by not only raw material
price, but also selling price of finished products and operating costs for manufacturing
such finished products. Accordingly, for purposes of realism, this research considers such
variables in the System Dynamics diagram in order to investigate more precisely the
effect of IT deployment in the perspective of profit.

As shown in Appendix 1, this dissertation assumes that IT deployment level is
determined by lookup function with investment budget for IT deployment (Figure 4-6).
Lookup function can specify an arbitrary nonlinear relationship between two construct
variables. Put simply, a Lookup is a list of numbers representing an x axis and a y axis.
The inputs to the Lookup are positioned relative to the x axis, and the output is read from
the y axis. Lookups can be used to create your own specialized functions. Lookups are
also referred to as "Lookup Functions," "Tables," "Lookup Tables," "Table Functions,"
and sometimes "Graphical Functions." Lookups can be declared as x,y pairs or by
specifying the x-axis followed by the y-axis. The format for declaring a Lookup is:

LOOKUP NAME([(Xmin,Xmax)-(Ymin,Ymax), (Xrefl,Yref1), (Xref2,Yref2),

(Xrefn, Yrefn)] (X1, Y1), (X2, Y2), . . .(Xn, Yn) ) ~ Units ~ Description |

or:

LOOKUP NAME(X1, X2, X3,....Xn, Y1, Y2, Y3,.... Yn) ~ Units ~ Description |

Table 4-2 gives guidelines for specifying the shapes and estimating the values of
table functions. This dissertation tried to specify the shapes of table functions according
to the guidelines in the table. Sensitivity analysis in stage 9 will be noted specifically in

the subsection for model validation test in this chapter.

68

Table 4-2: Guidelines for formulating Table Functions

 

Stag

Description

 

Normalize the input and output. Instead of Y = le), normalize the function so that the input is
the dimensionless ratio of the input to a reference value X“ and the output is a dimensionless
effect modifying the reference value Y“, Y = Y‘KX/X").

 

Identify the reference points where the values of the function are determined by definition. For
example, in normalized functions of the form Y = Y‘flX/X“), the function usually must pass
through the point (1,1) so that Y = Y“ when X = X”.

 

Identify reference policies. Reference policies are lines or curves corresponding to standard or
extreme policies. The reference policy f(X/X*) = 1 represents the policy that X has no effect on
Y. The 45 degree line represents the policy that Y varies 1% for every 1% change in X and is
often a meaningful reference policy. Use the reference policy curves to rule out infeasible
regions.

 

Consider extreme conditions. What values must the function take at extremes such as -oo, 0, and
+00? If there are multiple nonlinear effects in the formulation, check that the formulation makes
sense for all combinations of extreme values and that the slopes of the effects at the normal
operating points conform to any reference policies and constraints on the overall response of the
output.

 

Specify the domain for the independent variable so that it includes the full range of possible
values, including extreme conditions, not only the normal operating region.

 

Identify the plausible shapes for the function within the feasible region defined by the extreme
conditions, reference points, and reference policy lines. Select the shape you believe best
corresponds to the data (numerical and qualitative). Justify any inflection points. Interpret the
shapes in terms of the physical constraints and policies.

 

Specify the values for your best estimate of the function. Use increments small enough to get the
smoothness you require. Examine the increments between values to make sure there are no kinks
you cannot justify. If numerical data are available you can often estimate the values statistically.
If numerical data are not available, make a judgmental estimate using the best information you
have. Often, judgmental estimates provide sufficient accuracy, particularly early in a project, and
help focus subswent modelinggnd data collection efforts.

 

Run the model and test to make sure the behavior of the formulation and nonlinear function is
reasonable. Check that the input varies over the appropriate range (e.g. that the input is not
operating off the ends of the function at all times).

 

 

 

Test the sensitivity of your results to plausible variations in the values of the function. If
sensitivity analysis shows that the results change significantly over the range of uncertainty in the
relationship, you need to gather more data to reduce the uncertainty. If the results are not
sensitive to the assumed values, then you do not need to spend additional resources to estimate
the function more accurately.

 

(Source: Sterman, 2000)

Investment budget for IT deployment is secured from a buyer’s cumulative profit,

which is the sum of the weekly profits, as described in the causal loop diagram, by

multiplying the fraction variable indicating the ratio of profit invested for IT deployment

to buyer’s cumulative profit. A buyer’s weekly profit is calculated by subtracting the total

cost from the buyer’s revenue. Total cost consists of production costs, inventory costs,

RM purchasing costs, logistics costs including transport and processing costs, transaction

69

 

costs, investment for IT deployment, buyer’s non-contractible investment, and investment
on asset specificity (See Appendix 3). The RM purchasing cost is computed by
multiplying RM shipping quantities with the RM unit price. The value obtained from
multiplying the RM shipping quantities with the RM unit price then becomes the
supplier’s revenue. By subtracting the supplier’s total cost from the supplier’s revenue,
the supplier’s weekly profit and the supplier’s cumulative profit are obtained. The
supplier’s total costs consist of production and inventory costs. The RM shipping cost is
not included in the supplier’s total cost, because it is assumed as the burden of the buyer.
The supplier’s cumulative profit dictates the level of the supplier’s non-contractible
investment by using the fraction variable indicating the ratio of the supplier’s profit
invested for non-contractible investment to the supplier’s total profit. As Bakos and
Brynjolfsson (1993) mentioned, the supplier’s non-contractible investment has a
significant effect on the buyer’s non-contractible investment. In this perspective, the
buyer’s non-contractible investment is determined by lookup function with the supplier’s
non-contractible investment (Figure 4-6). The buyer’s non-contractible investment is
determined by multiplying the buyer’s cumulative profit with the fraction variable
indicating the level of the buyer’s profit invested to non-contractible investment as well
as the effect of the supplier’s non-contractible investment by lookup function. Investment
on asset specificity is determined by lookup function with IT level (Figure 4-6).
Transaction cost is decided by lookup function with the buyer’s investment on asset
specificity and non-contractible investment, and the desired number of suppliers is
dictated by lookup function with such transaction cost (Figure 4-7). The desired number

of suppliers adjusts the current number of suppliers through adjustment time.

70

 

 

 

 

 

 

 

 

 

 

mWWIUII’DWm

 

 

Effect of
Societal.

”' \

 

 

 

 

 

 

 

 

 

 

Eflectot
I'I'onfilo

......... \
\

0,1 \_..

01 to 2.0
"Level

 

 

 

 

 

 

 

 

Figure 4-6: Lookup Functions (1)

71

 

 

Effectot

 

ASIon 1.0
TC

/

 

 

 

 

 

 

1500
lnvestnert on Meet Specificity

 

EMU!

/

 

/

 

BNCI onl .0
TC

0.5

20

 

 

 

\

 

 

 

3000
Buyer: Harm Writ

 

\

 

 

\

 

 

 

\

 

 

 

5030
rumpus

72

Figure 4-7: Lookup Functions (2)

 

 

 

The actual number of suppliers adjusted from the desired number of suppliers
determines the levels of bargaining power of both buyer and supplier by lookup function
(Figure 4-8). By the ratio of supplier’s bargaining power to buyer’s bargaining power, the
effect of bargaining power is weighed. This bargaining power effect can be reflected in
the modification of the RM price as mentioned previously. The RM price is decided by
the effect of economies of scale and the effect of the RM demand-supply balance other
than the effect of bargaining power. The effect of economies of scale is determined by
lookup function with the number of suppliers (Figure 4-8). The effect of the RM demand-
supply balance is measured by relative inventory coverage (the ratio of buyer’s RM
demand quantity to supplier’s inventory quantity). Prices are assumed to respond to the
balance of demand and supply, without specifying how supply and demand are perceived
by market participants. Inventory coverage (the ratio of shipments to available inventory)
is an excellent measure of both inventory-carrying costs for producers and the ability of
buyers to receive reliable, timely deliveries. Consistent with many commodity models
and substantial empirical evidence, price is adjusted above (below) the expected
equilibrium level as inventory coverage rises (falls) relative to a normal, or reference
level. Also, price depends on perceived coverage, not instantaneous coverage, because
the instantaneous shipment rate is unknown. It takes time to gather and report data on
inventory and shipments. For simplicity, perceived coverage is modeled with first-order
smoothing. The coverage perception time would be short in markets with very good data
or high sensitivity of storage costs to inventory levels and longer in markets with poor
quality data or less sensitivity to storage costs (Sterrnan 2000). The indicated RM price

decided by the above three effects adjusts the current RM price through adjustment time.

73

 

 

Buyer's

 

0.1

We

 

 

 

 

 

 

10
Madanoliers

 

 

 

Bum“

 

 

 

 

 

 

10
No 01W

 

 

Effect of
Eoonomiestc

 

015680

\

 

 

\

 

 

 

 

10
knotSIDDliU

Figure 4—8: Lookup Functions (3)

74

 

 

 

Similarly, the indicated FG item price is decided by the effect of the F G demand-
supply balance that is measured by relative inventory coverage (the ratio of PG demand
quantity to F G inventory quantity). Reference inventory coverage level is determined at
the point leading to equilibrium in which the indicated FG price is equal to FG price, in
other words, the net rate of change of PG price must be zero, when it is assumed that
demand is stationary. The explanation on equilibritun situation will be addressed at the
end of this chapter.

As can be seen in Appendix 2, we consider a two-level supply chain that consists of
RM supplier and manufacturer. In the model, manufacturer determines its order quantity
as follows: 1) Manufacturer uses a periodic inventory control policy (order-up-to S.
policy). 2) In each period, manufacturer realizes demand. If manufacturer has enough
stock, the demand will be satisfied immediately. The unmet demand will be backordered.

Based on the actual demand (D) the manufacturer forecasts the demand for the next
period by an adaptive expectation (i.e., exponential smoothing). When the smoothing

time constant is T units of time, the forecast for the next calculation time, F n.1,. will be:

F(+d(:F1+ dt/T(D('F‘)
Thus, the forecast for the next time unit, t + 1, will be:

I/dt
17,... = Zdt/T(1-dt/T)"1D, + (1 can’t” F,
i =1

= [1- (1 -dt/T) 1("'1 D, + (1 -dt/T) ”d' F,

Note that d is the interval of time between calculations. Therefore, our simulation

model recalculates its numerical values every I/dt of unit time.

75

The target FG inventory level at time t, S,, is determined by:

S. = F.L + mm)”
= F. [L + za/F.(L)"”J
= F. [L + 2(1)”)

where F t is the mean demand forecast, o. is the standard error of demand forecast, and L
is the sum of the inventory review period, the transit lead time, and the manufacturing
lead time. The manufacturer assumes that the coefficient of variation of the demand
remains constant. Thus, she determines the amount of safety stock by a pre-specified
constant 2. The safety stock must cover demand uncertainty for a longer period of time.
When using a normal probability distribution, we multiply the desired standard deviations
to implement the cycle-service level, 2, by the standard deviation of demand during the
protection interval, oP+L(P: inventory review period, L: lead time). The value of z is the
same as for a continuous inventory review system with the same cycle-service level
(Krajewski and Ritzman 2000).

The manufacturer places an order for RM with the supplier and plans the production
quantity to bring his FG inventory level to the target level. Also, the manufacturer will
receive the shipment of the order for RM after a certain transit lead-time. The transit lead-
time is fixed. However, the manufacturer will not receive the same amount as it ordered,
if the supplier does not have enough stock to fill the order. The production rate is
adjusted continuously. There is a yield loss in production. The number of non-defective
finished items follows a binomial distribution (n, 1 - 8), where n is the number of items
started at the manufacturing facility and 8 is the defect rate. Since the manufacturer
knows the defect rate, the number of items started is adjusted (or divided by 1 - 5) so that

the expected number of non-defective finished items is equal to the amount to bring the

76

F G inventory level to the target level. Manufacturer determines weekly production
quantity according to the following equation:

Average F G Production Qty + (Target F G Inventory-F G 1nventory)/F G Inventory
Adjustment Time + (Demand Backlog-Desired Demand Backlog/Time to adjust Demand
Backlog + (Desired Work In Process -Work In Process)/F G WIP Adjustment Time
(Note: Desired Demand Backlog=A verage Demand Rate *Norrnal F G Delivery Delay,
Desired Work In Process =A verage Sales Rate *Mf Lead Time)

The supplier handles the manufacturer’s orders as follows. First, the manufacturer’s
order will be immediately shipped from the supplier’s inventory if the supplier has
enough stock. The unmet order will be backordered, and delivery delay is considered.
Second, the supplier forecasts the order for the next period by an adaptive expectation
(i.e., exponential smoothing) based on day-to-day demand information obtained from
information sharing with manufacturer. Generally, it can be recognized that the effect of
information sharing can be different depending on the level of information technology. In
other words, if more advanced IT is deployed, more accurate and timely sharing of
manufacturer’s forecast on daily demand becomes possible. Inversely, if IT level is
relatively low, the possibility of demand distortion due to a lack of information sharing
may increase. In order to reflect such argument, this dissertation creates the fraction
variable indicating the level of accurately reflecting manufacturer’s sales forecast to the
decision of supplier’s target inventory level by lookup function with IT level (Figure 4-
9). That is, the target RM inventory level is determined by multiplying the above fraction

variable with the same equation as the manufacturer determines the target FG inventory

level to be.

77

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/
//
° 0,// 1.0
/
//
00 // 5.00. 1000
\

 

 

 

 

 

 

 

 

2.0
IT Level

Figure 4-9: Lookup Functions (4)

78

 

Also, similarly with the manufacturer, the supplier plans the production quantity to
bring his RM inventory level to the target level. The production rate is adjusted
continuously. Weekly production quantity is decided by the same equation as the
manufacturer determines as follows:

Average RM Production Qty + (Target RM Inventory-Sp Inventory)/Sp Inventory
Adjustment Time + (RM Backlog-Desired RM Backlog)/Time to adjust RM Backlog +
(Desired Sp Work In Process —Sp Work In Process)/Sp WIP Adjustment Time
(Note: Desired RM Backlog= Average RM Order Rate *Normal RM Delivery Delay,
Desired Sp Work In Process =A verage RM Shipment Rate *Sp PD Lead Time)

We assume an infinite and instant production of raw material by the supplier. Also,
we consider a case when no member in the supply chain has the perfect information
about the demand. (i.e., no one knows the type of demand distribution or its parameters.)
Therefore, every member should “forecast” the demand. We believe our model will more
adequately capture the benefits of information sharing since, in real situations, neither the
manufacturer nor the supplier has the perfect information about the demand and the real
demand distributions are often non-stationary in many industries.

Since the demand forecasting is one of the critical factors in analyzing the benefits of
information sharing, we consider various types of demand distribution and their impact
on the benefits. Bourland et al. (1996) and Gavimeni et al. (1996) consider a stationary
demand distribution whose parameters (mean and variance) are known to every member
in the supply chain. Lee et al. (1997) consider a non-stationary (autoregressive) demand
distribution whose parameters are known to every member in the supply chain. They
suggest that the benefits of information sharing occur in almost every case since the

demand distribution is non-stationary. In this dissertation, we consider non-stationary

demand distributions.

79

4.2.3 Formulation of Equation

In order to implement the System Dynamics model described above, formulating its
associated computer (mathematical) equations is required. This is followed by selecting
‘reasonable parameter and stock values’ for initial starting values for the simulation.

As mentioned previously, one of the most important points in system dynamics
modeling is how to set reasonable initial starting points for parameter values. Thus, the
estimation of parameter values receives a great deal of attention in SD modeling. The
basic choice is formal statistical estimation from previous numerical data. However, it’s
not easy to get numerical data appropriate for all parameter values. Such limitations on
numerical data availability mean it is often impossible to estimate all parameters in a
model. So, in order to estimate parameters more realistically and find reasonable initial
starting point, this dissertation collected the qualitative and quantitative data from the
case study, because the case study can provide the insights necessary to develop the
analytical models that are needed in the simulation. Case studies were implemented on
several US manufacturing companies in North America and the North America
corporations of Korea’s large manufacturing companies, which are engaged in
automobile, electrics/electronics, food, and telecommunication industries. Equations for
stock, flow, and auxiliary variables used in the model are listed in Appendix 4.

As can be seen in Appendix 4, this dissertation sets the initial FG item price as $350
and the RM unit price as $100. If the demand remains constant, then bullwhip effect
scarcely occurs. So, in order to make a more realistic model, it is assumed that FG
demand is irregular within mean demand :I: standard deviation multiplied by product

customization level. In this dissertation, mean and standard deviation are set as 200 and

80

20 respectively. Standard deviation is multiplied by product customization level in order
to consider the positive influence of product customization level on demand uncertainty
discussed in the literature portion.

The level of information technology ranges from 0.1 to 2.0, and the fraction of
cumulative profit to investment on IT deployment is defined as 0.05, which means that
5% of cumulative profit per week is assigned as investment budget for IT deployment.
Product customization level is also assumed to increase by increments of 0.1 from 0.1 to
2.0. Such scale for product customization level is based on the approach of Kraljic (1983)
who classified product characteristics into two factors such as the level of market
complexity and the level of strategic dependence on product, and using the sum of
weights for these two factors as the level of product variety. The amount of normal
investment on asset specificity is set as $1500 per week and investment on asset
specificity is determined by multiplying such normal investment on asset specificity to
the effect of IT level on asset specificity ranged from 0.1 to 2.0. The fractions of
supplier’s and buyer’s cumulative profit to non-contractible investment are defined as
0.01 respectively, which means that 1% of supplier’s and buyer’s cumulative profit per
week is budgeted for non-contractible investment respectively. Normal transaction cost is
set as $5000 per week and changes according to the effects of buyer’s non-contractible
investment and asset specificity investment on transaction cost. The ntunber of suppliers
is decided within the range of 1 and 20. Buyer’s bargaining power, supplier’s bargaining
power, and effect of economies of scales are set to have a range from 0.1 to 2.0 in order

to assume the equal effect of those three variables on the indicated RM price.

81

Supplier’s and manufacturer’s unit costs for producing and holding the RM are
defined as $13 and $16 respectively. These lead to the initial equilibrium point at which
supplier’s and manufacturer’s revenues are equal to their costs respectively, that is, the
net rate of profit is zero. Equilibrium point will be noted again in the later part of this
chapter. The rationale behind manufacturer’s ratio of item cost to initial item price being
set as less than the supplier’s ratio is to reflect that FG manufacturing requires greater
efforts in fulfilling customer’s demands than RM manufacturing, and entire inventory
management after RM shipping from supplier and investment on IT deployment are
assumed as the burden of manufacturer in this dissertation. Both RM and F G holding
costs remain constant regardless of time flow, but the FG manufacturing cost is assumed
to change according to product customization level (Figure 4-9) by considering that, in
reality, cost or time saving (loss) may occur when changing production according to the
decrease (increase) of product customization level. Transport cost, one of two variables
composing logistics costs, is determined by lookup function with transport late (Figure 4-
9), and processing cost is assumed to remain constant at $20. These two definitions were
modified from the model suggested by Sterman (2000).

RM manufacturing lead-time, FG manufacturing lead-time, and transit lead-time
from RM supplier to manufacturer are each set at 2 weeks. One of the options for
determining the manufacturer’s inventory review period (P) is to base P on the cost trade-
offs of the economic order quantity (EOQ). In other words, P can be set equal to the .
average time between orders for the EOQ. Because demand is variable, some orders will
be larger than the EOQ and some will be smaller. However, over an extended period of

time, the average lot size should equal the EOQ. Using such logic, we can divide the

82

EOQ by the annual demand, D, and use this ratio as P (Krajewski and Ritzman 2000).
From the above procedure, the inventory review period is set at 2 weeks. Inventory
adjustment and backlog adjustment times are defined as 8 weeks respectively, and WIP
adjustment time is set at 4 weeks. Times to average production, shipment, and order rates
are defined at 4 weeks. As mentioned previously, safety stocks for both RM supplier and
manufacturer are decided by the equation zo,(L)'/2. Standard deviation is set at 20 as
mentioned previously, and z is defined as 1.28 for a 90 percent cycle-service level. The
exponential smoothing time constant is defined at 4 weeks. Defect rate is set as 0%, and
both supplier’s and manufacturer’s production capacities are assumed to be infinite.

As mentioned previously, this dissertation utilizes many lookup functions in order to
arbitrarily specify a nonlinear relationship between two construct variables that cannot be
indicated quantitatively. Lookup functions utilized in this dissertation are indicated in
Figure 4-6, 4-7, 4-8, and 4-9.

One of the important formulation techniques that will be used in the System
Dynamics model is the initialization of the levels to an equilibrium or steady state value.
This is needed to test the sensitivity to perturbations, that is, sensitivity to changes in
variables and in loop structure. A stock is in equilibrium when it is unchanging (a system
is in equilibrium when all its stocks are unchanging). For a stock to be in equilibrium the
net rate of change must be zero, implying the total inflow is balanced by the total outflow
(Sterman 2000). In other words, in equilibrium, the total inflows and outflows of each
stock must be equal. We can use this logic when trying to compute an initial value for
each stock variable. That is, initial stock value is decided at the point in which the total

inflows and outflows are equal. However, as mentioned previously, this dissertation

83

assumed that demand is randomly distributed. As such, it is very difficult to locate an
initial stock point leading to equilibrium. Accordingly, in order to find an initial stock
point, we observed the dynamic behavior of variables when demand remained constant at
215 units - that is the initial starting demand rate as well as the annual average demand
rate in a non-stationary demand model. Key initial stock points derived from the above

procedure are indicated in Table 4-3.

Table 4-3: Initial Stock Points

 

 

 

 

 

 

 

 

 

 

 

 

 

Stock Variable Initial Point Stock Variable Initial Point
IT level 1 No. of supplier 10
RM price $100 FG price $350
Manufacturer’s $300,000 Supplier’s $200,000
cumulative profit cumulative profit
Supplier’s WIP Average RM Manufacturer’s Average Sales
inventory Shipment Rate *Sp WIP inventory Rate *Mf Lead Time
PD Lead Time (Desired WIP)
(Desired WIP)
Supplier’s Target RM Manufacturer’s Target F G
inventory inventory inventory inventory
Average RM 200 Average FG 200
production production
quantity quantity
Average RM RM shipping Average RM Manufacturer’s RM
shipment rate order rate order in
RM backlog Desired RM FG backlog Desired demand
backlog backlog
RM in transit 400 RM inventory 400
Average sales rate FG sales Average demand Demand
rate
AR: Last demand 0

 

 

 

 

 

It is unnecessary to initialize structure in equilibrium, but it can be very helpful. It is

very common for the early part of a simulation run to be dominated by adjustments from

84

imbalances in levels and rates. The dynamics generated by this imbalance do not tend to
have intrinsic importance and can severely hamper the understanding of model dynamics.
Running time for simulation is set as 150 weeks. Unit for time is week, and time step

is 1 week.

4.3 Validity of System Dynamics Model

4.3.1 The Meaning of Validation

This section describes model-testing process by which we can build confidence that
the SD model mentioned above is valid for the purpose of this dissertation. In other
words, this section explains specific tests and procedures you should follow to test the
validity of a model for this dissertation’s purpose, uncover flaws, and improve the reality
of this dissertation’s SD model.

Many modelers talk about model “validation” or argue to have “verified” a model. In
fact, validation and verification of models is impossible. In other words, no model can
ever be verified or validated, because all models are wrong. All model, mental or formal,
are limited, simplified representations of the real world. They differ from reality in ways
large and small, infinite in number. The only statements that can be validated-shown to
be true-are pure analytic statements, propositions derived from the axioms of a closed
logical system. The impossibility of validation and verification is also applied to
computer models. Any theory that refers to the world relies on imperfectly measured
data, abstractions, aggregations, and simplifications, whether the theory is embodied in a

large-scale computer model, consists of the simplest equations, or is entirely literary. The

85

differences between analytic theories and computer simulations are differences of degree
only (Sterrnan 2000).

If validation is impossible and all models are wrong, why do we should build them?
We must recognize that our important decision to use a model-mental or formal- is never
whether to use a model but only which model to use. That is, our responsibility is to use
the best model appropriate for the purpose at hand in spite of its inevitable limitations.
The decision to delay action in the vain quest for a perfect model is itself a decision, with
its own set of consequences. Experienced modelers likewise recognize that the goal is to
help their clients make better decisions, decisions informed by the best available model.
Instead of seeking a single test of validity models either pass or fail, good modelers seek
multiple points of contact between the model and reality by drawing on many sources of
data and a wide range of tests. Instead of viewing validation as a testing step after a
model is completed, they recognize that theory building and theory testing are intimately
intertwined in an iterative loop. Instead of presenting evidence that the model is valid,
good modelers focus the client on the limitations of the model so it can be improved and
so clients will not misuse it.

Once we recognize that all models are wrong and abandon the black and white
dualism of truth and falsification, we can focus on the important questions: Is the model
useful? Do its shortcomings matter? To answer these questions you must first ask: Useful
for what purpose? Matter to whom?

Model users must critically assess the model’s boundary, time horizon, and level of
aggregation in light of their purpose. The model boundary determines which variables are

treated endogenously, which are treated exogenously, and which are excluded altogether.

86

Factors relevant to the purpose must be captured endogenously. Treating a concept as
exogenous, or omitting it, cuts all feedbacks involving that variable. Models with narrow
boundaries don’t capture the system’s responses to policies, leaving the clients to
discover them as unforeseen side effects in the real world. Narrow model boundaries are
the single greatest source of policy resistance in systems.

This section discusses tests to help answer questions about model purpose and
boundary, physical and decision-making structure, and sensitivity analysis. This section

focuses on the practical and political issues of modeling.

4.3.2 Model Validation Test

System dynamics modelers have developed a wide variety of specific tests to
uncover flaws and improve models (e.g., Forrester 1973; Forrester and Senge 1980;
Barlas 1989, 1990, 1996). Levine and Fitzgerald (1992) established three types of tests to
verify system dynamics models: structural tests, behavioral tests, and policy tests.
Structural tests include structural verification, parameter verification, dimensional
consistency, and boundary adequacy. Behavioral tests include behavior reproduction,
behavior prediction, behavior anomaly, surprise behavior, extreme policy,
generalizability, and boundary sensitivity. Policy tests include changed behavior
prediction, boundary commission, and system improvement.

The purpose of these tests is to determine the robustness of the model to normal and
extreme situations and to ensure that no relevant paths have been omitted from the model

nor have extraneous paths been added. Internal and external validity involves

87

adding/deleting exogenous and endogenous variables to the model to determine their
effects on the system (J anszen 2000).

4. 2. 3. 1 Structural Tests

(1) Boundary Adequacy Test

Boundary adequacy tests assess the appropriateness of the model boundary for the
purpose at hand. These tests consider whether there are potentially important feedback
omitted from the model. Of course, the list of omitted concepts and variables is infinite.
So, the focus of boundary adequacy test should be placed on whether any feedbacks
omitted from the model, if included, might be important given the purpose of the model.
With the appropriateness of the model boundary, we should check the validity of the
model feature itself. That is, we should ask whether or not the overall feature of the
model makes sense and the process and hypothesis for constructing feedback loops are
valid. Review of the relevant literature and archival materials, interview with outside
experts, and direct experience with the system may suggest some significant points for
validating model boundary and feature.

This dissertation made three key feedback loops representing IT effect on buyer-
supplier relationship and supply chain dynamics on the basis of previous key articles
(Bakos and Brynjolfsson 1993; Clemons et al. 1993; Lee et al. 1997a, 1997b, 2000;
Sterman 2000; Grahovac and Chakravarty 2001) suggesting direct causal relationship
among variables constructing the causal loop diagram described previously. The
boundary validation of this dissertation’s SD model was checked and modified from few
US. and Korean supply chain managers who have direct experience with SCM system

and top-level executives of sales, production, or planning department who was well

88

acquainted with supply chain policies and corporate strategy of the firm. Also, the
committee members of this dissertation screened the boundary validation of the model.
(2) Relationship Validity Test

Relationship validity test check whether the direct relationship between variables
constructing the feedback loops in this dissertation’s SD model was appropriately
established. Such relationship validity test can be largely classified into two parts. First is
to test whether the direction of causal links between variables was appropriately defined
and moving direction between variables assigned as positive or negative was properly
indicated. Second is to examine the validity of equations indicating specific relationship
between variables.

The validity of causal links can be identified from previous researches (Bakos and
Brynjolfsson 1993; Clemons et al. 1993; Lee et al. 1997a, 1997b, 2000; Sterman 2000;
Grahovac and Chakravarty 2001) suggesting direct causal relationship among variables
constructing the causal loop diagram similarly with boundary adequacy test. The validity
of equations is also verified by previous supply chain dynamics models (Coyle 1977;
Narasimhan 1979; Kim 1998; Sterman 2000).

However, as mentioned in the causal loop diagram, this dissertation’s SD model
includes the diagram representing the effect of IT investment on buyer-supplier
relationship which was totally created by this dissertation based on previous literatures
(Bakos and Brynjolfsson 1993; Clemons et al. 1993) other than the diagram reflecting the
effect of IT investment on supply chain dynamics. Also, such diagram includes many
non-linear relationships between variables. As mentioned previously, this dissertation

used lookup function in order to specify an arbitrary nonlinear relationship between two

89

construct variables. It is very important to test the sensitivity of the results in order to
identify plausible variations in the assumed shape and values of all non-linear functions.
If sensitivity analysis shows that the results change significantly over the range of
uncertainty in the relationship, we need to spend additional resources to estimate the non-
linear function more accurately. In order to identify whether or not the diverse shapes of
non-linear relationships lead to different simulation results, sensitivity analyses with
various types of lookup fimctions were implemented, and the variance among the
simulation results was investigated by the ANOVA and Duncan multiple range tests. The
results showed that there was no significant difference in either the magnitude of
fluctuation or the trend of behavioral change.
(3) Structure Assessment Test

Structure assessment test considers whether the model is consistent with knowledge
of the real system relevant to the purpose. Structure assessment focuses on the level of
aggregation, the conformance of the model to basic physical realities such as
conservation laws, and the realism of the decision rules for the agents (Sterman 2000).

Violations of physical laws such as conservation of matter or energy usually arise
because the model does not appropriately capture the stock and flow structure of the
system. As mentioned previously, in order to discriminate more precisely between stocks
and flows, this dissertation used the snapshot approach that takes a picture on a system
and distinguish stocks and flows depending on whether we can count or measure in the
picture. Another approach for discriminating between stocks and flows is to suppose the
situation that all business transactions are suddenly stopped. Stocks would be those firings

we can identify subsistence. For example, if business transaction is stopped, buyer’s and

90

supplier’s profits at that moment should be zero. But, cumulative profit subsists.
Similarly, when business transaction between buyer and supplier is stopped, we can count
the number of existing suppliers, but the number of currently desired suppliers is zero.
Through double screens by the two approaches described above, this dissertation
captured the stock and flow structure of the system.

Other common violations of physical law involve stocks that can become negative.
Real quantities such as inventories, populations, and cash balances cannot be negative.
Therefore the outflows from all such stocks must approach zero as the stock approaches
zero. This means that there must be a first-order negative feedback loop that restricts all
the outflows from real stocks so that the flow is zero when the stock is zero. These loops
must be first-order because any time delay in the loop could cause the rate to continue
even after the stock reaches zero, a physical impossibility. We can check for the presence
of first-order control by direct inspection of the equations. Also, partial model tests can
demonstrate the intended rationality of the individual decision rules. In a partial model
test, each organizational firnction or decision point is isolated from its environment until
environment is consistent with the mental model that underlies the decision rule. The
subsystem can then be challenged with various exogenous patterns in its inputs (Sterman
2000). Through equation inspections and partial model tests, this dissertation checked
whether or not the abovementioned violations exist among total 21 stock variables in this
dissertation’s SD model. As a result, any abnormal violation was not found.

(4) Dimensional Consistency Test
Dimensional consistency is one of the most basic tests that we should do first. We

should always specify the units of measure for each variable as you build your models.

91

More often, units error reveal important flaws in your understanding of the structure or
decision process you are trying to model.

Vensim, which is one of simulation software package for system dynamics,
implements automated dimensional analysis. Thus, we can test for dimensional errors
with a single command. This dissertation investigated the existence of dimensional errors
by the command, and as a result, no error messages were discovered.

(5) Parameter Assessment Test

Model constancy arises either because dynamics affecting the state of the system are
so slow that change is imperceptible or because there are powerful negative feedback
processes keeping the state of the system nearly constant even in the face of
environmental disturbances. The uncertainty in parameter values is important and must
be tested. Sensitivity analysis checks whether our conclusions change in ways important
to your purpose when assumptions are varied over the plausible range of uncertainty.

In assessing sensitivity to parametric assumptions we should first identify the
plausible range of uncertainty in the values of each parameter or nonlinear relationship.
Then, we should test the sensitivity to those parameters over a much wider range. Vensim
includes automated sensitivity analysis tools. First, we specify which parameters to vary,
and then provide a range of values for each. Vensim then runs the model as many times
as we like, using the specified values for each parameter, either one at a time (univariate
testing) or all at once (multivariate testing).

This dissertation identified the robustness of the model through the implementation
of sensitivity analysis. One common method for sensitivity analysis is to define extreme

situations represented by highest and lowest values. This involves manipulating the rate

92

and exogenous variables from normal to extreme situations represented by highest and
lowest values. Models should be robust in extreme conditions. Robustness under extreme
conditions means that the model should behave in a realistic and consistent fashion no
matter how extreme the inputs or policies imposed on it may be. Sensitivity analysis
checks whether models behave appropriately and consistently when the inputs take on
extreme values. Thus, sensitivity analyses were conducted on all parameters in the model.
As a result, the behavioral patterns of endogenous variables in the cases of highest and
lowest values for each parameter were the same with base case even though a little
difference in the magnitudes of variables among two extreme cases and base case exists.
This means that the change of parameter value does not have a significant influence on
the behavioral change of overall system, even though the implication in terms of
magnitude can be different. In other words, this testing process indicates that parameter
does not significantly change the constructs or the propositions that are to be tested in this
dissertation, thus verifying the robustness of the model.
(6) Time-step Error Test

The results of SD model should not be sensitive to the choice of time step. In other
words, the wrong time step can introduce spurious dynamics into SD model. One method
for testing such time step (DT) error is to cut the time step in half and running the model
again. If the results change in ways that matter, the time step should be modified, and
such modification continues until the results are no longer sensitive to the choice of time
step (Sterman 2000). By the abovementioned procedure, this dissertation confirmed that

the change of time step does not change simulation results.

4. 2. 3.2 Behavioral Tests

93

(1) Dynamic Behavior Test

There are many available methods to evaluate a model’s ability reproducing the
behavior of a system. Most common method is to use descriptive statistics to assess the
point-by-point fit. Point-by-point metrics calculate some measure of the error between an
actual data series and the model output at every point for which data exist and then report
some sort of average over the relevant time horizon (Sterman 2000). The most
representative measures of point-by-point fit are the Coefficient of Determination (R2),
Mean Absolute Error (MAE), Mean Absolute Precent Error (MAPE), Mean Absolute
Error as a fraction of the mean (MAE/Mean), Root Mean Square Error (RMSE), and
Theil’s inequality Statitics decomposing MSE into three components : bias (U M), unequal
variation (Us), and unequal covariation (UC); UM + US + UC = 1.

Unfortunately, this dissertation does not have actual data series available to assess
point-by-point fit. In such a reason, it’s impossible to implement behavior reproduction
test. Instead, we can think about an alternative in which the SD models suggested in
previous research related to supply chain dynamics are used as reference model for
assessing the validity of system behavior. In other words, through the comparison in
system behavior between this dissertation’s SD model and those of previous researches,
the validity of system behavior within this dissertation’s SD model may be indirectly
identified. The researches of Narasimhan (1979), Christopher (1994), and Lee et al.
(1997, 2000) can be suggested as the most representative reference models. Those
researches emphasize consistently that information sharing by the utilization of advanced
IT can reduce distortion in demand information between buyer and supplier, specifically,

inventory and production fluctuations are softened gradually according to time flow in

94

case of perfect information sharing between buyer and supplier. Figure 5-5 (See p.100)
indicates that the system behavior of SD model suggested in this dissertation is consistent
with the previous research’s emphasis described above. This provides the validity of
system behavior of this dissertation’s SD model.

However, as can be seen in Figure 5-5, not only supplier’s inventory fluctuation is
softened, but also supplier’s entire inventory level is reduced, as the level of IT increases.
This is clearly different from the results of previous research. This is because this
dissertation assumes that the effect of information sharing can be different depending on
the level of information technology, unlike previous research that analyzed just the
difference in effect between two extreme situations such as perfect information sharing
and imperfect information sharing. Such difference between this dissertation and
previous research brings out the necessity of point-by-point comparison with actual data
series to assess more precisely the validity of system behavior. This deserves further
investigation in the future.

(2) Behavior Anomaly Test

The abovementioned data limitation makes it difficult to test the statistical
significance of important relationships or formulations. Behavior anomaly test examines
the significance of relationships or formulations by considering whether anomalous
behavior arises when the relationship is deleted or modified. Anomalous behavior derived
from the deletion of a relationship provides the evidence for its significance (Sterman
2000). From the implementation of such behavior anomaly test, the significance of each
relationship established in this dissertation was verified.

(3) Extreme Condition Test

95

Models should be robust in extreme conditions. Robustness under extreme
conditions means that the model behave in a realistic fashion mo matter how extreme the
inputs or policies imposed on it may be. For example, inventory cannot drop below zero,
and production cannot occur without raw materials. Extreme condition test checks
whether models behave appropriately when the inputs take on extreme values such as
zero or infinity (Sterman 2000). This dissertation carried out extreme condition test by
automated “reality check” tool which Vensim provides.

4. 2. 3.3 Policy Tests
(1) System Improvement Test

System improvement test checks whether the modeling process helped change the
system for the better. To pass the test, the modeling process must identify policies that
can lead to improvement. In system dynamics, such policies are called as ‘policy
leverage’. Policy leverage means the point that can produce better performance within a
limited and insignificant change of the magnitude with keeping the shape derived from
base run. Such policy leverage can be classified into parameter adjustment policy and
structural adjustment policy. Parameter adjustment policy means changing artificially
parameter values in a SD model. Structural adjustment policy means changing feedback
loops in a SD model. In chapter 5, this dissertation suggests policy leverage leading to
better performance of the system (See pp.118-126), thus providing the validity of the SD

model suggested in this dissertation in terms of system improvement.

96

CHAPTER 5. RESULTS OF SIMULATION

5.1 The Relationships among IT Level, Buyer-Supplier Relationship, and Supply
Chain Dynamics

In this section, we first investigate the results of simulation run on the System
Dynamics model that was discussed in the preceding chapter. From now, we will refer to
this as the base model. As mentioned in the preceding chapter, in this dissertation, we
consider non-stationary demand distributions. To generate non-stationary demand
streams, we use a simple AR(1) process. The demand at time t at the retailer is

D,=d+pD,-1+e,
where d> 0, -l < ,0 <1, and 3, follows a normal distribution with mean zero and standard
deviation 0. Note that p = 0 corresponds to the special case where the demand for each
period is an i. i.d. normal distribution with mean d and standard deviation 0. Since the
mean and the variance of the above AR(1) process depends not only on d and 0, but also

on the autoregression coefficient, p, we adjust d and 0 according to p as follows:

d = (I - p) E(D)
o" = (1 — p’) Var(D)

In this section, we observe the results of experiments when setting ,0 = 0.

As mentioned previously, mean and standard deviation are set as 200 and 20
respectively. Standard deviation is multiplied by product customization level.

Figure 5-1 shows demand stream derived from a simulation run on the base model

under the above assumption for non-stationary demand distributions.

97

Demand

 

400

350

300

250

 

 

200

J;

0 20 40 I 60 80 100 120 140
Time (week)

 

 

Demand : Current units/week

Figure 5-1: Demand Stream

5.1.1 The Relationship between IT level and Buyer-Supplier Relationship

Figure 5-2 indicates the changes of IT level according to time flow.

IT level and No. of Suppliers

2 Dmnl
20 Persons/week

 

0.8 Dmnl
0 Persons/week

 

 

 

 

0 3o 60 90 1201 150
Time (week)

 

IT level ‘i i fi4 1 ‘i ‘i 4 Dmnl
"No. of Suppliers" "fi 2 2 2 2 Persons/week

 

Figure 5-2: The Change of IT Level and No. of Suppliers

As can be seen in the figure, both IT level and the number of suppliers continuously
increase after going through three transient recession periods. This is an unexpected

result in the perspective of incomplete contract theory emphasizing that continuous IT

98

deployment cannot lead to persistent increase in the number of suppliers, and inversely,
such stagnation in the number of suppliers prevents the continuous improvement of IT
level. The reason of such unexpected result can be explained by the interactive
relationship between buyer’s profit and supplier’s profit. As mentioned previously, in this
dissertation, it was assumed that IT level is decided by investment budget for IT
deployment, and such investment budget is obtained from buyer’s cumulative profit. This
means that IT level is influenced by buyer’s cumulative profit. In such instance, we need

to check the dynamic change of buyer’s cumulative profit.

Buyer's revenue and cost

100,000 dollars/week ’
100,000 dollars/week

 

60,000 dollars/week
60,000 dollars/week

 

 

 

0' 30 60 90 120 150
Time (week)

 

Buyer's Revenue : Current 1 ‘i ‘r dollars/week
Total Operating Cost : Current 2 2 2 dollars/week

 

Figure 5-3: The Change of Buyer’s Revenue and Cost

As can be seen in Figure 5-3, even though this dissertation ran simulation at the point
leading to equilibrium in which revenue and cost are equal, buyer’s revenue becomes
slightly greater than buyer’s cost according to time flow. This is because a slight
difference exists between buyer’s revenue and cost at initial stages due to randomly

distributed demand variations according to time flow, and such slight difference is

99

accumulated through a continuous feedback process with other construct variables. Thus,
this leads to the gradual increase of buyer’s cumulative profit (Figure 5-4), and is

subsequently followed by the increase of IT level as shown in Figure 5-2.

Buyer's Profit and Supplier's Profit

600,000 dollars
400,000 dollars/week

 

 

200,000 dollars
0 dollars/week

 

 

 

0 3O 60 90 120 150
Time (week)

A A A

Buyer's Cumulative Profit : Current 1 i i dollars
Supplier's Cumulative Profit: Current —-2-—2- dollars/week

Figure 5-4: The Change of Profit

Such increase of IT level ultimately leads to the continuous increase of buyer’s profit
through the consecutive reductions of investment on transaction specific assets and
transaction costs, an increase in the number of suppliers, the increase of buyer’s
bargaining power, and the decline of raw material price, as discussed in the causal loop
diagram and SD model. Accordingly, the result in Figure 5-2 shows the existence of a
mutual positive feedback relationship among IT level, number of suppliers, and buyer’s
profit.

The above result is inconsistent with the perspective of incomplete contract theory
based on the concept of the zero sum game between buyer and supplier. The key
argument of incomplete contract theory is that the increase of buyer’s profit by the
decrease of RM price cannot be persistent due to subsequent responses such as the

decrease of supplier’s profit, the reduction of supplier’s incentives to make non-

100

contractible investments, the actual increase of buyer’s transaction costs, a decrease in the
number of suppliers, the increase of supplier’s bargaining power, and the increase of
material price. However, unlike anticipation, supplier’s profit does not decrease as much
as buyer’s profit increases as can be seen in Figure 5—4. In the figure, supplier’s
cumulative profit is shifted inversely with buyer’s cumulative profit. That is, when
supplier’s cumulative profit arrives at the point of diminishing return, buyer’s cumulative
profit is at the point of increasing return, and vice versa. However, such diminishing or
increasing return point in supplier’s cumulative profit also increases in a similar fashion
with buyer’s cumulative profit, thus indicating that supplier’s cumulative profit increases
continuously.

This implies that supplier’s decrease in profit does not coincide exactly with the
buyer’s increase in profit. We can explain the reason of this result from the effect of IT

on supply chain dynamics.

IT level and Supplier's Inventory

2 Dmnl
2,000 units

 

 

0.8 Dmnl
800 units

 

 

 

0 30 60 90 120 150
Time (week)

 

IT level : Current 4. 1 Li + ‘i ‘i Dmnl
Sp Inventory : Current 2 2 2 2 2 units

 

Figure 5-5: The Change of IT level and Supplier’s Inventory
Figure 5-5 indicates that the increase of IT level drives not only the decrease of

inventory fluctuation, but also the decrease of the entire inventory level. That is,

101

information sharing by the utilization of advanced IT can reduce distortion in demand
information between buyer and supplier, which is referred to as the "bullwhip effect"
(Lee et al. 1997; Christopher 1994). Such decrease in demand variability enables accurate
forecasting, and this can lead to the decrease of supplier’s target raw material (RM) level.
This lowering in supplier’s target RM level may deduce the decline of supplier’s
inventory level. Such decline in supplier’s inventory can lead to the decrease of supplier’s
holding cost and total cost, and such decrease may offset the decline of supplier’s
revenue. Actually, as can be seen in Figure 5-6, the diminishing return point of supplier’s
revenue decreases continuously. This is because RM price decreases gradually due to the
increase of buyer’s bargaining power by the increase in the number of suppliers, as can
be seen in Figure 5-7. However, supplier’s total cost also decreases gradually due to the
decrease of supplier’s holding cost, and Figure 5-6 indicates that such decrease of
supplier’s total cost may offset the decrease of supplier’s revenue, thus leading to the

gradual increase of supplier’s cumulative profit.

Supplier's Revenue and Cost

40,000 dollars/week
40,000 dollars/week

 

 

0 dollars/week
0 dollars/week

 

 

0 30 60 90 120 150
Time (week)

 

Supplier's Revenue : Current 1 ‘r ‘fi dollars/week
Supplier's total operating cost : Current '2—2- dollars/week

Figure 5-6: The Change of Supplier’s Revenue and Cost

102

No. of Suppliers and RM Price

 

20 Persons/week
200 dollars/unit

0 Persons/week
O dollars/unit

 

 

0 30 60 90 120 150
Time (week)

A L A

"No. of Suppliers" : Current i i I Persons/week
RMPrice:Current 2 2 2 2 J2 dollars/unit

 

Figure 5-7: The Change of No. of Suppliers and RM Price

Such increase of supplier’s profit leads to the increase of supplier’s incentives for
non-contractible investment, and such increase of supplier’s incentives may reduce
buyer’s burden for non-contractible investment.

According to incomplete contract theory, the increase of buyer’s profit by advanced
IT deployment cannot be persistent because the increase of burden for buyer’s non-
contractible investment offsets the reduction of transaction cost due to IT development.
As shown in Figure 5-8, the effect of asset-specificity investment (ASI) on transaction
cost (TC) decreases radically after week 18, and does not indicate any significant
increase. This is an expected result according to the continuous increase of IT level.
However, the effect of buyer’s non-contractible investment (BNCI) on TC does not
increase unlike our expectation, and on the contrary, after week 100, stabilizes at a very
low level. We can recognize easily that the reason for this is due to the decrease of
buyer’s burden for non-contractible investment by the unexpected subsequent increases

of supplier’s cumulative profit and supplier’s non-contractible investment.

103

Transaction Cost

 

6,000

3,000

 

 

 

 

0 20 40 60 80 100120 140
Time (week)

Transaction Cost : Current4—-1—-1 dollars/week

 

    

Effect of BNCI and AS1 on TC
2
2
+ a
0.5 . -

 

 

 

 

0 20 40 60 80 100 120 140

Time (week)

 

Effect of BNCI on TC : Current t t ‘i t 1
Effect of ASl on TC : Current 2 2 2 2 2

 

Figure 5-8: The Change of Transaction Cost and Effects of BNCI and AS]

The decrease of RM price inducing the decrease of supplier’s revenue and the
increase of buyer’s profit can be derived from the low level of transaction cost (the first
figure in Figure 5-8) through the above interaction between the effects of buyer’s non-
contractible investment and asset specificity investment on transaction cost. That is, the
decrease of transaction cost by the interaction between the effects of buyer’s non-

contractible investment and asset specificity investment on transaction cost subsequently

104

leads to an increase in the number of suppliers and an increase of buyer’s bargaining
power, thus ultimately connected to the decrease of RM price.

The continuous increase of buyer’s cumulative profit shown in Figure 5-4 can be
explained by the decrease of supplier’s revenue due to the decrease of RM price.
Supplier’s revenue in Figure 5-6 is the purchasing cost in terms of buyer, which is one of
the buyer’s cost factors. This means that the decrease of supplier’s revenue mentioned
previously may lead to the decrease of buyer’s cost, because purchasing costs combined
with holding costs place a large part of the total cost.

Conclusively, all of the above results indicate that the improvement of IT level
reduces supplier’s revenue by the decrease of RM price due to an increase in the number
of suppliers as expected. However, such decrease of supplier’s revenue is offset by the
decrease of supplier’s inventory cost due to the improvement in the level of IT. Thus,
supplier’s cumulative profit does not decrease as much as the reduction of supplier’s
revenue even during a recession period. Such a less than anticipated decrease prevents
too great a reduction in the number of suppliers and in buyer’s profit. This drives a
gradual increase in the amount of supplier and buyer profits, thus leading to the
continuous improvement of IT level. This means that continuous advanced IT
deployment can ultimately derive complete competition in the electronic market among
numerous suppliers, and such electronic market structure can provide the capability to
persistently deploy more advanced IT to the buyer. That is, there is a mutual positive
feedback relationship between IT level and buyer’s bargaining power, thus supporting
transaction cost theory. However, such positive feedback relationship is not derived

solely fi'om the relationship between IT level and buyer-supplier relationship. The above

105

results show that the decrease of supplier’s inventory level and inventory cost due to
more accurate and timely information sharing by advanced IT plays a significant role in
the positive feedback relationship between IT level and buyer’s bargaining power. This
implies that supply chain dynamics may have a significant influence on the relationship
between IT level and buyer-supplier relationship, thus supporting this dissertation’s key
proposition that interactive feedback relationships among IT level, buyer-supplier
relationship, and supply chain dynamics should be considered for IT investment and

deployment for efficient supply chain management.

5.1.2 The Relationship between IT level and Supply Chain Dynamics

Previous literatures (Lee, et al. 1997, 2000; Christopher 1994) emphasize that
information sharing by the utilization of advanced IT can reduce distortion in demand
information between buyer and supplier, which is referred to as the "bullwhip effect", and
thus can significantly minimize the outcomes of the problems induced by bullwhip effect.
This dissertation agrees with previous literatures concerning the effect of information
sharing on bullwhip effect. But, the perspective of this dissertation on information
sharing is a little different from those of previous literatures. Previous literatures analyzed
only the difference in effect between two extreme situations such as when information
sharing is achieved and when information sharing is not implemented. That is, previous
literatures did not comment that the level of information sharing might be different
according to IT level. This dissertation assumes that the effect of information sharing can
be different depending on the level of information technology. In other words, if more

advanced IT is deployed, more accurate and timely sharing of manufacturer’s forecasts

106

on day-to-day demand may be possible. Inversely, if IT level is relatively low, the
possibility of demand distortion due to the lack of information sharing may increase. As
mentioned previously, this dissertation’s SD model reflects such an argument, and the
result is indicated in Figure 5-5.

Figure 5—5 indicates that as the level of IT increases, not only is supplier’s inventory
fluctuation softened, but supplier’s entire inventory level is reduced. This means that the
improvement of information sharing capability by continuous deployment of advanced IT
can reduce supplier’s total cost more than anticipated by not only minimizing substantial
variation in inventory holdings induced by bullwhip effect, but also lowering the absolute
level of supplier’s inventory. Such reduction of supplier’s absolute inventory level can be
the critical factor driving the gradual increase of both buyer’s and supplier’s profits, IT

level, and buyer’s bargaining power as mentioned previously.

Inventory Level

 

1,700 units
1,700 units
1,700 units

950 units
950 units
950 units

200 units
200 units W
200 units
0 20 40 60 80 100 120 140
Time (week)

 

 

 

 

Sp Inventory : Current 1 1 1 1 1 1 1 1 1 1 units
RM Inventory : Current 1 2 2 2 1 er a 2 A; 11 units
FG Inventory : Current 3 a s 3 a s s

 

 

Figure 5-9: The Change of Inventory Level

107

However, the benefit of inventory reduction due to the improvement of IT level is
not applied to the buyer. As can be seen in Figure 5-9, buyer’s RM and FG inventory
levels do not change substantially. This means that there is no significant direct
relationship between IT level and buyer’s inventory level, thus implying that the buyer
may not anticipate the effect of advanced IT deployment in terms of buyer’s inventory
reduction. Nevertheless, we can assert that making investment for advanced IT
deployment is needed, because it is still beneficial in terms of the buyer. As mentioned
previously, the improvement of IT level induces the reduction of supplier’s revenue due
to an increase in the number of suppliers and the decline of RM price, and also drives the
reduction of supplier’s inventory cost, which is more than enough to offset the decline of
supplier’s revenue. Accordingly, even if a supplier accepts the fall of RM price, the
supplier’s ex ante incentives to make non-contractible investment may not be declined
significantly, and thus the buyer can enjoy the benefit of RM price reduction derived
from low transaction cost and high buyer’s bargaining power in the long run.

Conclusively, the above argument implies that the buyer-supplier relationship may
not necessarily be a zero-sum game. John Gossman, Vice-President of materials
management at AlliedSignal, has emphasized that competition is no longer company to
company, but supply chain to supply chain (Gossman 1997), and such an observation
brings out the importance of a win-win strategy in which both buyer and supplier can co-
exist. The above result suggests the existence possibility of such a win-win SCM
strategy. One point to be identified from the results of base run is that a win-win strategy
in which both buyer and supplier can co-exist is not derived just from the relationship

between IT level and buyer-supplier relationship. That is, the buyer-supplier relationship

108

in this dissertation was modeled under the perspective of transaction cost theory and
incomplete contract theory, which are based on the concept of a zero-sum game between
buyer and supplier. Viewed in this perspective, the possibility of a win-win strategy can
be suggested when feedback loop representing IT effect on supply chain dynamics affects
or further dominates feedback loop reflecting IT effect on buyer-supplier relationships.
The results of base run discussed previously support such an argument. This also
provides the likelihood that the change of supply chain structural issues and the role of IT
in such change of supply chain structural issues may alter the feature of buyer-supplier
relationship based on the zero-sum game. Such possibility will be examined in the

suggestion of policy leverage, which will be noted later.

5.1.3 The Effect of Product Variety Level on the Interactions among Feedback Loops

As observed in the preceding section, the effect of IT on buyer-supplier relationships
cannot be explained only by the conflict between transaction cost theory and incomplete
contract theory. This suggests that other forces like supply chain dynamics should be
accounted for in a more complete model of buyer-supplier relationship. Product
customization (PC) level can also be an important factor for explaining such a complete
model of buyer-supplier relationship.

In order to examine such possibility, we implemented sensitivity analysis on the
responses of key variables by product customization level, which is a parameter value in
this dissertation’s SD model. Figure 5-10 indicates the result of sensitivity analysis on IT
level by PC level. The solid line indicated inside the bounds in the first figure is the result

of the base run.

109

 

Current

50% 75% .95%-100%-
IT level
2

 

1.5

 

 

 

 

 

0.5
0
0 37.5 75 1 12.5 150
Time (week)
(Overall)
Current
50% 75%.95%-100%-
IT level
2

 

1.5

 

 

 

 

 

 

 
    

 

 

 

0.5
0
0 37.5 75 112.5 150
Time (week)
(Low Product Variety Level)
Current
50% 75%.95%-100%-
IT level
2
1.7
1.4 -
1.1 M
0.8 1’ ‘
0 37.5 75 112.5 150
Time (week)
(High Product Variety Level)

Figure 5-10: Sensitivity Analysis on IT level by Product Variety Level

110

A graph indicates confidence bounds for all the output values of IT level when PC
level is randomly varied about their distributions. The outer bounds (100%) mean
maximum and minimum values of PC level’s distribution. 50% indicates the medium
value of the distribution. As shown in the figure, according to the change of PC level, IT
level and number of suppliers have different magnitudes even though they show similar
shapes. In order to identify more clearly the effect of PC level, we divided the first figure
into two parts based on the result of the base run.

As can be seen in the second and third figures, as PC level decreases relative to base
model, the magnitude of IT level increases while keeping the shape derived from the base
nm. Meanwhile, as PC level increases relative to the base model, the magnitude of IT
level decreases with the same shape as that in the base nm, and at the highest level of
product variety, IT level fluctuates continuously within the scope of 1.0 and 1.3. We need
to compare the above results with the results of sensitivity analysis on the number of
suppliers by product variety level. Figure 5-11 indicates the result of sensitivity analysis
on the number of suppliers by product customization level. As can be seen in the figure,
as PC level decreases relative to the base model, number of suppliers changes immensely,
similar to the case of the base run, but its magnitude is a little higher than that of the base
run. Meanwhile, as PC level increases relative to the base model, the number of suppliers
continues to show large fluctuations after week 50, and increasing return point decreases
gradually by small steps. The above two results indicate that as PC level increases, it
becomes increasingly difficult to realize any mutual positive feedback relationship among
IT level, the number of suppliers, and buyer’s bargaining power. As well, the IT level and

number of suppliers tends to return to the initial middle level.

111

 

Current
50% 75%-95%-100%=

"No. of Suppliers"

 

 

 

 

 

 

5
0
0 37.5 75 1 12.5 150
Time (week)
(Overall)
Current

50% 75%-95%-100%-

"No. of Suppliers"

 

 

 

 

 

 

5
0
0 37.5 75 112.5 150
Time (week)
(Low Product Variety Level)
Current

50% 75%fi95%-100%-

"No. of Suppliers"

 

    

 

 

 

20

. x '  \/
10

5

0 O 37.5 75 112.5 150

Time (week)
(High Product Variety Level)

Figure 5-11: Sensitivity Analysis on No. of Supplier by Product Variety Level

112

We already know from the discussion in the preceding section that the above
responses of IT level and number of supplier in the case of high product customization
level were derived from the interaction with the inventory level of the entire supply
chain. Figure 5-12 indicates the result of sensitivity analysis of demand and target
inventory level in the case of high product customization level. As can be seen in the
figure, the decreasing trend of target RM inventory level slows down gradually. This is
because the need for safety stock increases due to high demand uncertainty in
customized/differentiation focused products in which the level of product customization
and variety is high (See the first figure in Figure 5-12). The entire level of F G inventory
also increases even though the shape is very similar to that of the base model. Such
gradual increase of target inventory level leads to the increase of supplier’s inventory
level and buyer’s FG inventory level as shown in Figures 5-13 and 5-14 indicating the
changes of key variables driving the responses of IT level and number of suppliers in the
case of high product customization level. That is, after week 40, as product variety level
increases, supplier’s inventory level shows large fluctuations without any significant
change of increasing and diminishing return points. Such response of supplier’s inventory
level may induce the blunting of supplier’s profit growth due to the increase of supplier’s
inventory cost (See the second figure in Figure 5-13). As mentioned previously, the
blunting of supplier’s profit grth prevents the increase of supplier’s non-contractible
investment, and subsequently increases the buyer’s burden for non-contractible
investment. Consequently, actual transaction cost increases relative to the base model
(See the third figure in Figure 5-13), thus slowing down the increasing trend in the

number of suppliers shown in the third figure in Figure 5-11.

113

 

Current
50% 75%

Demand
400

95%-100%=

 

 

350

300 *

 

" :’ I ‘ ' . I '5

 

 

200 0 37.5 75 112.5 150

Time (week)

 

 

Current
50% 75% 95%.100%E

Target RM Inventory
2.000

 

 

1,700

l ,400

 

1,100

 

 

 

800 0 37.5 75 112.5 150

Time (week)

 

Current
50% 75%m95%-100%-

Target FG Inventory
2,000

 

1,750

1,500

l ,000
0

1,250

 

 

 

37.5 75 1 12.5 150
Time (week)

Figure 5-12: Sensitivity Analysis on Demand and Target Inventory
(In case of High product variety level)

114

 

Current
50% 75%-.95%-100%

Demand
400

 

350
300

5 i
II I . g... f, “ .
250 Il‘. \j\,“' i .J is V \Ilil 1 Q! I k!‘:
I ' ' I / I \ I ‘A\’: . ‘3 V

I l

 

 

 

200 0 37.5 75 , 112.5 150

Time (week)

 

Current
50% 75%n95%-100%

Target RM Inventory
2.000

 

 

1,700
l ,400

1,100

 

 

 

 

800
0 37.5 75 112.5 150

Time (week)

 

Current
50% 75%m95%-100%-

Target FG Inventory
2,000

 

1,750

1,500

l ,000
0

1,250

 

 

 

37.5 75 1 12.5 150
Time (week)

Figure 5-12: Sensitivity Analysis on Demand and Target Inventory
(In case of High product variety level)

114

 

Current
50% 75%-95%-100%-

Sp Inventory
2,000

 

1,500

 

l ,000

500

 

 

 

0 37.5 75 112.5 150
Time (week)

 

Current
50% 75%-95%-100%-

Supplier's Cumulative Profit
400,000

 

300,000

200,000

 

1 00,000

 

 

 

0 0 37.5 75 112.5 150

Time (week)

 

Current
50% 75%-95%-100%-

Transaction Cost
8,000

 

6,000

4,000

2,000

 

 

 

0

 

0 37.5 75 1 12.5 150
Time (week)

Figure 5-13: Key Variables driving the Change of No. of Suppliers
(In case of High product variety level)

115

 

Current
50% 75%

FG Inventory
2,000

 

m 95%-100%=

 

1,750
1,500

1,250 N

l ,000
0

 

 

 

37.5 75 1 12.5 150
Time (week)

 

Current
50% 75%B9S%-100%
Total Operating Cost
100,000 ;

 

 

90,000

80,000

 

70,000 ~~

 

 

 

60,000 ' "
0 37.5 75 112.5 150

Time (week)

 

Current
50% 75% u— 95%-100%-
Buyer's Cumulative Profit

600,000

 

 

500,000
400,000

300,000

 

 

200,000

 

0 37.5 75 112.5 150
Time (week)

Figure 5-14: Key Variables driving the Change of IT level
(In case of High product variety level)

116

As can be seen in the first figure in Figure 5-14, as product customization level
increases, the entire level of buyer’s F G inventory increases while maintaining the shape
of the base run. Such increase in F G inventory level leads to the substantial increase of
buyer’s total cost due to the increase of inventory cost, and such increase of total cost is
connected to the subsequent reduction in buyer’s profit growth (See the third figure in
Figure 5-14) and IT level as shown in Figure 5-10.

One thing to be noted is that slowing down the increasing trend in the number of
suppliers is not connected to the substantial increase of RM price and supplier’s revenue,
which is enough to offset the increase of total cost due to the increase of inventory level.
Figure 5-15 suggests an explanation for this trend. As can be seen in the figure, as
product customization level increases, supplier’s bargaining power gradually increases
according to the reduction in the number of suppliers. However, as product customization
level increases, the ratio of RM shipping to supplier’s inventory decreases according to
the increase of supplier’s inventory level. Accordingly, RM price does not show a
significant increase relative to that of the base run as much as expected from limiting the
increasing trend in the number of suppliers. Also, due to such stagnation of RM price, the
buyer’s profit may not decrease below the initial point, and then fluctuate continuously
within a certain limited range.

Conclusively, in the case of low product variety level, mutual positive feedback
relationship among IT level and buyer’s bargaining power in the perspective of
transaction cost theory dominates, while in the case of high product variety level,
negative feedback relationship among IT level and buyer’s bargaining power in the

perspective of incomplete contract theory dominates.

117

 

Current
50% 75%-95%-100%n

Supplier's Bargaining Power
2

 

1.5

 

 

 

 

0.5
0
O 37.5 75 112.5 150
Time (week)
Current

 

50% 75%-95%-100%fl

Relative Sp Inventory Coverage
2

 

1.5

 

 

 

 

0 37.5 75 1 12.5 150
Time (week)

 

Current

50% 75%-95%-100%-
RM Price
200

 

150

100

 

50

 

 

 

0 37.5 75 112.5 150
Time (week)

Figure 5-15: The Changes of RM price and related variables
(In case of High product variety level)

118

This argument is consistent with the suggestion of Bensaou and Venkatraman (1996)
emphasizing that the type of market exchange relationship in which the technical and
economical dependence of both buyer and supplier on each other is relatively low, is
appropriate for highly standardized products. Meanwhile, the type of strategic partnership
in which buyer-supplier relationship is strongly connected by considerable transaction-
specific assets of both buyer and supplier is significantly correlated with highly
customized products.

Conclusively, all of the above results indicate that the change of feedback
relationship between IT level and buyer’s bargaining power is influenced by interactive
relationships among information sharing by advanced IT, product customization level,
and supply chain dynamics. This means that depending on how the described feedback
relationships affect each other or which feedback relationship is dominant, the effect of
IT investment can vary. This provides the validity of this research’s key proposition that
information technology investment and deployment for efficient supply chain
management should be implemented taking into consideration the interactive feedback
relationships among IT level, product customization level, buyer-supplier relationship,

and supply chain structure.

5.2 The Suggestion of Policy Leverage

5.2.1 The Necessity of Policy Leverage

As discussed in the preceding section, as product customization level increases, it

becomes difficult to realize mutual positive feedback relationships among IT level, the

119

 

number of suppliers, and buyer’s bargaining power relative to the base model, and IT
level and number of suppliers tends to return to the initial middle level. In other words, in
the case of highly customized products, it is difficult to bring out the existence of a win-
win strategy in which both the buyer and supplier can co-exist. The source of such result
is the increase in supplier’s total cost due to the increase of supplier’s inventory level,
which slows down the increasing trend in supplier’s profit. Consequently, such increase
of supplier’s inventory level prevents the continuous improvement of IT level. Inversely,
due to such stagnation of the IT level, the effect of information sharing between buyer
and supplier cannot be demonstrated properly.

Viewed in this perspective, the question of how to reduce inventory level effectively
can be the most critical factor that enables pursuing a win-win strategy even in highly
customized products. Accordingly, policies for reducing inventory level effectively in
highly customized products should be considered. In System Dynamics, such policies are
referred to as ‘policy leverage’. Policy leverage means a policy interruption point that can
produce considerable effects with little input of limited policy resources. That is, policy
leverage is a variable or the relationship between variables that enables the entire system
to demonstrate considerable dynamics with little effort. Policy leverage can be classified
into parameter adjustment policy and structural adjustment policy. Parameter adjustment
policy means altering artificial parameter values in a SD model. Structural adjustment

policy means altering feedback loops in a SD model.

5.2.2 Parameter Adjustment Policy

120

As discussed in the results of sensitivity analysis by product customization level in
the previous section, the beginning of the result leading to increasing trends of IT level
and the number of suppliers becoming blunt is the increase of target RM and F G
inventory levels, which comes from the increase in the need for safety stock due to high
demand uncertainty in customized/differentiation focused products. This means that the
effort for finding policy leverage should start from the consideration on policies reducing
target inventory levels. Policies for reducing target inventory levels can be inferred from

equations for target RM inventory level and target FG inventory level.

Target RM Inventory Level = Fraction of reflecting Mfg Sales Forecast to Target RM
Inventory*(Mfg Sales Forecast *(Sp Production Lead Time + Mf Review Period)+
"SF.'Sp"*sqrt(Sp Production Lead Time + Mf Review Period))

Target F G Inventory Level = Mfg Sales F orecast*(T ransit Lead Time + Mf Lead Time +
Mf Review Period)+"SF: Mf"*sqrt(Transit Lead Time + Mf Lead Time + Mf Review
Period)

The above equations indicate that transit lead-time from RM supplier to
manufacturer may have significant influences on target inventory level. Actually, Fisher
(1997) suggests that the approaches for reducing lead-time should also be changed
depending on the type of product strategy and buyer-supplier relationship. That is, in case
of customized/differentiation focused products and strategic partnership structure, the
type that immediately supplies products upon demand may be preferred to that of placing
a large amount of inventories on distributors or retailers, because the margin rate and
added value are high. Thus, in order to pursue high quality customer service with low

inventory level, a quick response logistics system should be constructed. Also, in case of

121

15“ ____

customized/differentiation focused products, a wide range of product lines should be
completed and the necessity of consistent product innovation and R&D is high. Because
of such characteristics of product variety, it is very difficult for distributors or retailers to
hold and control properly the inventories of various kinds of products in their warehouses
or sales shops. Accordingly, in this case, it is advisable to reduce supply chain lead-time
in order to deal effectively with immediate demand. The above argument means that the
reduction of parameter values related to transit lead-time might be considered as the
policy for reducing target inventory level.

Another thing to be noted in the equation for target RM inventory level is the
fraction variable for reflecting manufacturer’s sales forecast to the decision of target RM
inventory. As mentioned in chapter 4, this variable reflects that the effect of information
sharing can be different depending on the level of information technology. That is, if
more advanced IT is deployed, more accurate and timely sharing of manufacturer’s
forecasts on daily demand may be possible. Inversely, if IT level is relatively low, the
possibility of demand distortion due to the lack of information sharing may increase.
When considering that IT level is directly influenced by investment budget for advanced
IT deployment, and such investment budget is decided by multiplying the fraction of
buyer’s profit invested for IT to buyer’s cumulative profit in this dissertation’s SD model,
the above argument means that the increase of parameters related to the fraction of
buyer’s profit invested for IT may have a significant effect on the reduction of target
inventory level, thus deserving another policy for reducing target inventory level.

The comparison between the above two kinds of policies is another interesting issue.

The change of parameter for the fraction of buyer’s profit to IT investment is needed to

122

test the importance of the improvement of information technology itself, while the
change of parameter values for transit lead-time reflects the analysis on the significance
of managerial capability on information technology. From such comparison between two
types of policies, we can identify the validity of Keen’s (1993) assertion that the
difference in competitive and economic benefits, which firms can obtain from
information technology, is dependent not on the difference in technology itself, but the
difference in managerial capability.

Figures 5-16, 5-17, 5-18, and 5-19 indicate the result of sensitivity analysis on the
responses of key variables by the reduction of transit lead-time and the increase in
fraction of buyer’s profit to IT investment. In order to compare more precisely the effects
of the two policies, we set the rate of change for the two parameters related to the two
policies as identical. Figures 5-16 and 5-17 consists of key variables driving the change
in the number of suppliers, while Figures 5-18 and 5-19is composed of key variables
driving the change of IT level. As can be seen in Figures 5-16 and 5-17, both the
reduction of transit lead-time and the increase in fraction of buyer’s profit to IT
investment lead to the increase in supplier’s profit. As discussed previously, such
increase in supplier’s profit is derived fi'om the reduction of inventory holding cost due to
the decrease of supplier’s RM inventory level as shown in the first figure. Also, such
increase in supplier’s profit ultimately leads to the increase in the number of suppliers
through the increase of supplier’s incentives for non-contractible investment, the decrease
of buyer’s burden for non-contractible investment, and the decrease of transaction cost

(See the third figure in Figures 5-16 and 5-17).

123

Current
50% 75%-95%-100%-
Sp Inventory

2,000

 

 

1,500

1,000

 

 

 

 

 

500
0
0 37.5 75 112.5 150
Time (week)

Current
50% 75%-95%-100%-
Supplier's Cumulative Profit
400,000 1

 

300,000

200,000

 

 

 

 

100,000
0 L
O 37.5 75 1 12.5 150
Time (week)
Current

 

50% 75%-95%-100%-

"No. of Suppliers"
20

 

15

 

 

 

 

0 37.5 75 1 12.5 150
Time (week)

(by the reduction of transit lead-time)

Figure 5-16: Sensitivity Analysis on Key Variables by Parameter Adjustment (1)

124

 

C urrent

50% 75%-95%-roo%-

Sp Inventory
2,000

 

1,500

1,000

 

 

 

 

 

500
0
0 37.5 75 112.5 150
Time (week)

Current
50% 75%-95%-100%-
Supplier‘s Cumulative Profit
400,000

 

300,000

 

 

 

 

100,000
0
0 37.5 75 1 12.5 150
Time (week)
Current

 

50% 75%-95%-100%-

"No. of Suppliers”

 

 

 

 

 

20
15
10
5
O
0 37.5 75 112.5 150
Time (week)

(by the increase of fraction of profit to IT)

Figure 5-17: Sensitivity Analysis on Key Variables by Parameter Adjustment (2)

125

Current

50% 75%-95%-100%-

FG Inventory
2,000

 

 

1.750

1 .250

 

 

l ,000

 

0 37.5 75 112.5 150
Time (week)

Current

50% 75%-95%-100%-

Buyer's Cumulative Profit
800,000

 

 

600,000

 

 

 

 

o 37.5 75 1 12.5 150
Time (week)
Current
50% 75%-95%-100%-
IT level
2

 

 

0.5

 

 

 

0 37.5 75 I 12.5 150
Time (week)

(by the reduction of transit lead-time)

Figure 5-18: Sensitivity Analysis on Key Variables by Parameter Adjustment (3)

126

 

Current
50% 75%-95%-100%-

FG Inventory
2.000

 

1.750

1,500

1.250

 

 

 

0 37.5 75 112.5 150
Time (week)

 

Current

50% 75%-95%-100%-
Buyer‘s Cumulative Profit
600,000

 

500,000

400,000

300,000

 

 

200,000

 

 

o 37.5 75 1 12.5 150
Time (week)
Current
50% 75%-95%I-100%-
11‘ level
2

 

 

0.5

 

 

 

0 37.5 75 112.5 150
Time (week)

(by the increase of fraction of profit to IT)

Figure 5-19: Sensitivity Analysis on Key Variables by Parameter Adjustment (4)

127

However, unlike anticipated, it is very difficult to find a significant difference
between the effects of both policies. When considering that the sources of the results in
Figures 5-16 and 5-17 started from supplier inventory level, this means that in terms of
supplier, there does not exist a significant difference in benefit between two policies.
However, in terms of buyer, there exists a substantial difference in effect between the two
policies. As can be seen in Figures 5-18 and 5-19, even though both the reduction of
transit lead-time and the increase in fraction of buyer’s profit to IT investment lead to the
subsequent increases of buyer’s profit and IT level, there exists a clear difference in the
magnitude of effect between the two policies. That is, the effect of transit lead-time
reduction indicates to be greater than that of the increase of fraction of profit to IT
investment. Such result can be also explained by the interaction with inventory level. As
can be seen in the first figure in Figures 5-18 and 5-19, in case of transit lead-time
reduction, the magnitude of buyer’s FG inventory level decreases substantially while
keeping the shape derived from the base model, while in the case of the increase of
fraction of profit to IT investment, the magnitude of buyer’s FG inventory does not
change at all relative to the result of the base model. This result is reasonable because we
already discussed from Figure 5-9 that there does not exist a significant direct
relationship between IT level and buyer’s inventory level, and thus the buyer may not
anticipate the effect of advanced IT deployment in terms of buyer’s inventory reduction.
Such difference in buyer’s FG inventory level drives the substantial difference in the
effect on buyer’s profit and IT level between transit lead-time reduction and the increase
of fraction of profit to IT investment. This means that in terms of the buyer, the change of

supply chain structural issue such as the reduction of transit lead-time may be more

128

beneficial and effective than the increase of investment for the improvement of IT itself.
This supports the argument of Keen ( 1993) emphasizing the significance of appropriate
adoption of information technology rather than the availability of information technology

itself.

129

CHAPTER 6. IMPLICATION AND CONCLUSION

6.] Contribution

This research starts from the following fundamental research question. Is the
development of advanced information technology always beneficial for efficient supply
chain management? The assertion of Keen (1993) that the difference in competitive and
economic benefits which firms can obtain from E-commerce technology is dependent on
not the difference in technology itself, but the difference in managerial capability,
emphasizes the importance of research on IT adoption in supply chain management rather
than the availability of IT itself. In order to test the validity of this argument, this research
posits that information technology investment and deployment for efficient supply chain
management should be implemented taking into consideration the interactive, feedback
relationships among IT level, product customization level, buyer-supplier relationship,
and supply chain structure. Through dynamic simulation analysis by system dynamics on
three kinds of feedback loops representing the effect of IT on buyer-supplier relationship
and supply chain dynamics, this research disclosed that the effect of IT investment in
supply chain management can be different depending on the interaction among the above
three feedback relationships. In other words, depending on how the described three
feedback relationships affect each other or which feedback relationship mostly
dominates, the effect of IT investment can have a different shape. This provides the
validity of this research’s key proposition above described. This important insight bears
on several significant implications relating to effective IT investment and deployment for

efficient supply chain management.

130

6.1.1 Positive Feedback Relationship between IT Level and Buyer’s Bargaining Power
Specifically, simulation results indicate that there is a mutual positive feedback
relationship between IT level and buyer’s bargaining power. In other words, continuous
advanced IT deployment ultimately can derive complete competition in electronic market
among numerous suppliers, and inversely such electronic market structure can provide
the capability for persistent deployment of more advanced IT to buyer. This is consistent
with the perspective of transaction cost theory. However, such positive feedback
relationship is not derived just from the relationship between IT level and buyer-supplier
relationship. The results of this research show that the decrease of supplier’s inventory
level and inventory cost due to more accurate and timely information sharing by
advanced IT plays a significant role in the positive feedback relationship between IT
level and buyer’s bargaining power. This implies that supply chain dynamics may have a
significant influence on the relationship between IT level and buyer-supplier relationship,
thus supporting empirically this dissertation’s theoretical proposition that interactive,
feedback relationships among IT level, buyer-supplier relationship, and supply chain
dynamics should be considered for IT investment and deployment for efficient supply

chain management. This is the first contribution of this dissertation.

6.1.2 Change of Information Sharing Level according to IT Level
Previous literatures (Lee et al. 1997, 2000; Christopher 1994) did not comment that
the level of information sharing might be different according to IT level. That is, previous

literatures analyzed just the difference in effect between two extreme situations such as

131

when information is shared and when information is not shared. This dissertation
assumes that the effect of information sharing can be different depending on the level of
information technology. In other words, if more advanced IT is deployed, more accurate
and timely sharing of manufacturer’s forecast on day-to-day demand may be possible.
Inversely, if IT level is relatively low, the possibility of demand distortion due to the lack
of information sharing may increase. This dissertation’s SD model reflects such
argument, and from such effort, suggests that the improvement of information sharing
capability by continuous deployment of advanced IT can reduce supplier’s total cost
more than anticipated by not only minimizing substantial variation in inventory holdings
induced by bullwhip effect, but also lowering the absolute level of supplier’s inventory.
This suggestion discriminates the value of this dissertation from previous researches.
However, one thing to be noted is that this dissertation used two-level information
sharing model, while some previous researches deal with three or four-level information
sharing model. It is very interesting to examine whether this dissertation’s result
described above can be shown even in three or four-level model. This deserves further

investigation.

6.1.3 Existence Possibility of Win-Win Strategy

This research also reveals that the reduction of supplier’s absolute inventory level
above described can be the critical factor driving the gradual increase of both buyer’s and
supplier’s profits. This implies that buyer-supplier relationship may not be necessarily
zero-sum game, thus emphasizing the significance of win-win strategy in which both

buyer and supplier can live together. The results of this research suggest the existence

132

possibility of such win-win SCM strategy. However, win-win strategy in which both
buyer and supplier can live together is not derived from the relationship between IT level
and buyer-supplier relationship under the perspective of transaction cost theory and
incomplete contract theory, which are based on the concept of zero-sum game between
buyer and supplier. The results of base run discloses that the possibility of win-win
strategy can be suggested when feedback loop representing IT effect on supply chain
dynamics affects or further dominates feedback loop reflecting IT effect on buyer-

supplier relationship. Such finding deserves another contribution of this dissertation.

6.1.4 Validity of the Portfolio of Inter-Firm Transaction Relationship

As mentioned previously, the effect of IT on buyer-supplier relationship cannot be
explained only by the conflict between transaction cost theory and incomplete contract
theory. This suggests that other forces like supply chain dynamics should be accounted
for in a more complete model of buyer-supplier relationship. The results of sensitivity
analysis on the responses of key variables by product variety level find that product
variety level can be also an important factor for explaining such complete model of
buyer-supplier relationship. That is, as product variety level decreases relative to base
model, the magnitude of IT level and the number of supplier increase with keeping the
shape derived from base run. Meanwhile, as product variety level increases relative to
base model, the magnitude of IT level and the number of supplier decrease with the same
shape as that in base run. The above result means that in case of low product variety
level, the establishment of electronic market structure can be still derived from mutual

positive feedback relationship among IT level and buyer’s bargaining power. However,

133

in case of high product variety level, we cannot expect the persistence of such electronic
market structure due to the domination of negative feedback relationship among IT level
and buyer’s bargaining power. This argument is logically consistent with the portfolio of
inter-firm transaction relationship of Bensaou and Venkatraman (1996) emphasizing that
the type of market exchange relationship in which the technical and economical
dependence of both buyer and supplier on each other is relatively low, is appropriate for
highly standardized products. Meanwhile, the type of strategic partnership in which
buyer-supplier relationship is strongly connected by considerable transaction-specific
assets of both buyer and supplier, is significantly correlated with highly customized
products. However, the portfolio of Bensaou and Venkatraman (1996) is based on
theoretical suggestion, while this dissertation tested empirically the validity of Bensaou

and Venkatraman (1996)’s argument. This is another contribution of this dissertation.

6.1.5 The Availability of Information Technology Itself vs. The Significance of
Managerial Capability on Information Technology

The possibility of win-win strategy in which both buyer and supplier can live
together above described also provides the likelihood that the change of supply chain
structural issues and the role of IT in such change of supply chain structural issues may
alter the feature of buyer-supplier relationship based on zero-sum game, even in highly
customized and differentiation-focused product. Specifically, the results of this research
reveals that depending on how to manage more efiiciently supplier’s and buyer’s

inventory level, the effect of IT investment on buyer-supplier relationship and supply

134

chain dynamics can be different. As a way for managing inventory level, we can consider
the change of lead-time and IT investment budget.

The comparison between the above two ways is very interesting issue. Keen (1993)
insists that the difference in competitive and economic benefits which firms can obtain
from information technology is dependent on not the difference in technology itself, but
the difference in managerial capability. This research identifies the validity of Keen
(1993)’s assertion. The result of sensitivity analysis on the responses of key variables by
the reduction of transit lead-time and the increase of fiaction of buyer’s profit to IT
investment indicates that even though both the reduction of transit lead-time and the
increase of fraction of buyer’s profit to IT investment lead to the subsequent increases of
buyer’s profit and IT level, the effect of transit lead-time reduction is greater than that of
the increase of fraction of profit to IT investment. This means that the change of supply
chain structural issue such as the reduction of transit lead-time may be more beneficial
and effective than the increase of investment for the improvement of IT itself. This
supports the argument of Keen (1993) emphasizing the significance of appropriate
adoption of information technology rather than the availability of information technology

itself.

6.2 Future Research

In this dissertation, we considered just the changes of transit lead-time and fraction
of buyer’s profit to IT investment as the subjects of parameter adjustment policy.
However, we cannot deny the possibility that according to the changes of other parameter

values, we may get different results and more advisable suggestions relative to previous

135

discussions. Also, the behavioral changes of key variables according to the interactions
among parameter values also cannot be ignored. This should be addressed in future
research.

As mentioned previously, policy leverage can be classified into parameter
adjustment policy and structural adjustment policy. Parameter adjustment policy means
changing artificially parameter values in a SD model, as we already discussed above.
Structural adjustment policy means changing feedback loops in a SD model.

As an approach for managing inventory in the dimension of structural adjustment
policy, we can think about the change of inventory review system. This research assumes
that both buyer and supplier review inventory level by periodic inventory review system
in which the item's inventory position IP every P time periods is reviewed (not
continuously), and an order equal to (T —IP), where T is the target inventory, that is, the
desired 1P just after placing a new order is placed. However, we can consider the case
that both or either buyer or supplier use continuous inventory review system in which
whenever a withdrawal brings IP down to the reorder point (R), an order for Q (fixed)
units is placed. It will be very interesting to compare the results under the assumption of
both inventory review system.

Another interesting research theme in structural adjustment policy is the effect of
information sharing in a response-based logistics system. As mentioned in the literature
review part, Christopher (1994) suggests that “response-based” or “pull” logistics
systems which represent the direct connection of point-of-sale location to the point of
production via information technology, can reduce demand distortion and inventory

fluctuation. This argument emphasizes that supply chain participants can accomplish both

136

cost reduction and better customer service by sharing information in a response-based
logistics system (Bowersox and Closs 1996; Lee et al. 1997a; Lee et al. 1997b; Closs et
a1. 1998). Such possibility also deserves further investigation in policy experiment.

The model of this dissertation is confined to two-level serial supply chain model
between RM supplier and F G manufacturer. The expansion of model into three-level
supply chain model (RM Supplier —> Manufacturer —> Retailer) and non-serial supply
chain model (lzN, Nzl, NzN) will suggest more impressive results. Also, the simulation
analysis on the case that demand is randomly distributed with increasing or decreasing
trends will be addressed in future research.

System dynamics models rest on the theory of nonlinear dynamics, an area in which
tremendous progress has been made over the past two decades (Sterman 2000). This
dissertation’s SD model includes many nonlinear relationships among the construct
variables. The issue on how do we represent such nonlinear relationship more reliably
and realistically is a big challenge for the generalization and future development of this
dissertation’s proposed model. The cross-validation process of applying the model to data
from empirical survey with a sample of US. and European firms and evaluating its
goodness of fit by analyzing the structural relationships can generalize the proposed

model. This should strengthen the external validity of this study’s results.

137

APPENDICES

138

APPENDIX 1:
System Dynamics Diagram (1)

   
 

Reference FG
Inventory Coverage

Fraction of Profit to
Investment

  

<FG Inventory>
FlC Perception Time

FG Inventory
Cove

Relative FG
Inventory Coverage

 

Supplier's
Curnulatrve
Profit Perceived FIC

Investrrrent Budget
for IT Developrrrent

 

 

 

Supplier‘s Profit

  
  
 

 

Supplier’s total
operating cost

<FG Sales> .
FG Pnce

 

 

 

Suppliers Revenue

/

Supplier’s Buyer‘s Revenue

Supplier's ,
rrrventory cost

production cost
<RM Shipping>

 

RMPrice

 

angeinRMPrice

 

 

Gian Producmg> <Sp Inventory>

+

RM Price
Adjustment Time

Target [T level Fraction of Supplier‘s

Profit to NC Investment /
Indicated RM Price
Fraction of Buyer’s .
Profit to NC
Investnrent

IT Adjustment Tune

Suppliers
Non-contractrble
Investment

+

Effect of
<Buyer’s Bargaining Power
Cumulative Profrt> ' +

 

IT level Effect of Econonaes

of Scales

 

 

Change in IT level

 

Buyer‘s .
Bargaining Power+ Suppliers

Bargarnrrrg Power
\x

No. of
Suppliers

Buyer‘s
Non-contractible
_ lnvestrnent

Effect of

Effect of IT on AS SUPPUCVS NC]

Investment

 

 

 

 

Normal Investnrent on
Asset Specrfrcrty

Grange II No of
/Suppliers

/

Sp Inventory
Coverage

- Effect of BNCI on TC

Supplier
\ Adjustment Tme

Transaction CostJ

lnvestnrent on
Asset Specificity
Desired No of

Suppliers

EITCCI OIASI on TC <RM Shipping)

+

139

   

<Buyer‘s
Cumrlatrve Profit>

Effect of F0 DS
Balance

+

Indicated FG Pnce

Orange 'rr FG Price

V

FG Price
Adjustment Tire

Effect of RM DS
Balance

Relative Sp Inventory
Coverage

Perceived SIC

Reference Sp
Inventory Coverage

SIC Perception Time

<Sp lnventory>

Eng 955 «means. Ezv
.232 ¢<v

52 3:52
.205 as am

      

AB: on: .55.?
a: 93 2 am

w Acne: ¥<v
A 838 uSEoOV
>38 but—om 85% 39:2 2am .085
O... EEOZ
/18\u¢m nap—2
mEmVImIouEm :95". wsmmouflwfi\ 52' CE 2 «SEE 5
I253: I... b . . .Id b9:o>=_ am
A05 :20
2M quo><v
\Ila 5.2.8: am
5:052;
EM omEo><
HE. Eu:_m=€<

«Sum 9350
omEu><v /
be 5.385:
AEHEmad<
mac—=02: 2M "um—E. bo~=u>5 am
max own—96 o. 2:;

flit—mice. I.I.II.II'
mozuum 9.5.x
magnum 3v

boEoZ: on. 59:

  

2:632; :Sm

            

332n—

mseeuE is...
‘
5 S3 cam

     
   

 

             
          
          

          

   

               

cassava:
on. omEo><

    
 

”Eu own—oi o. HEP
Anon»; 32>uEEv

   

 

  

05: Efiazv
:2 Unmr\‘
2.; Enema:
E? On.
332$ 33m 82 “E; .:u=_m_.._.c< A85. 33 On nwv
boEoZ: O..—
voeea 33:52 a: Enema?
as am
3% E
2.: 5% 2 mm: mesa». 2.; 33 55;
Ce 5:8:
\ A_o>u_ .:v
A25: 55815.25 9.988... so
6:va seam. .._.:o 952:

£33 cots—.226 AuSEoOV
:5va
2 S gamma momEmFAD 806$

N X—DZm—mm<

140

APPENDIX 2 (cont’d)

System Dynamics Diagram (2-2)

 

 

 

 

 

 

 

 

 

 

 
  
    
   
 
 
 
  
 
 

  
   
  
 
  
 

 

 

 

 

 

<RM Shipping> QEgBSfS)
<Norrml RM (PG 5819?
Delivery Delay>
Dermnd ‘
mom RM Backlog #53 {BE-E—b Ramos ~
er In Mf Order Out Den'and Fulfillrmnt
. . Dest RM Backlog Desired Dermnd
Penodrc RM Backlog
Order Qty
RM Backlo <AR Current Dmd>
8
. Adjustment Demnd Backlog
<MfProductron Qty> Adjustment
. . Time to adjust
Trnre to adjust Dermnd Backlog
RM Backlog
Average RM
Average ‘==%€3
Order Rate Change in Average Demand Rate Change in
Dennnd Rate

 

 

 

Order Rate

 

 

 

 

  

Tirrre to Average
Order Rate Time to Average
Dermnd Rate
AR‘ errent Dmi
/’ €3=I—-3:’ AR Last Dmd
AR' Const AK Updg'z/
AR‘ Error \
Tirrre to Update
Dermnd
AR SD Enor
AR Mean AR: Rho AR StDev AR RN Seed
Product

Customimtion Level

141

APPENDIX 3
System Dynamics Diagram (3)

<Buyer's
Non-contractible
Investment>

       
 
      

<Investment on
Asset S ecificity>

<lnvestrmnt Budget
for IT Development>

 

Buy er‘s
—:

<Buyer's Revenue>

<l‘ransaction Cost>

<Start Process ing>
<RM Price> Total Operating Cost
Purchasing Cost
<RM Shipping> Transmit Cost Holding Cost «item cos,
V Processing Cost
<Product
Customization Level>
Unit Order

     

Transport Rate

<FG <RM
Inventory> Inventory>

142

(001)

(002)

(003)

(004)

(005)

(006)

(007)

(008)

(009)

(010)

(011)

(012)

(013)

(014)

APPENDIX 4
Equation

"AR: Const"= (1-"AR: Rho")*"AR: Mean"
Units: units/week

"AR: Current Dmd"= "AR: Const"+"AR: Rho"*"AR: Last Dmd"+"AR: Error"
Units: units/week

"AR: Error"= RANDOM NORMAL(O,100,0,"AR: SD Error","AR: RN Seed")
Units: units/week

"AR: Last Dmd"= INTEG ("AR: Updating", 0)
Units: units/week

"AR: Mean"= 200
Units: units/week

"AR: Rho"= 0
Units: Dmnl

"AR: RN Seed"=1.5
Units: Dmnl

"AR: SD Error"=sqrt(1-"AR: Rho""2)*"AR: StDev"
Units: units/week

"AR: StDev"= 20*Product Customization Level
Units: units/week

"AR: Updating"=("AR: Current Dmd"-"AR: Last Dmd")/Time to Update Demand
Units: units/week/week

Average Demand Rate= INTEG (Change in Demand Rate, Demand)
Units: units

Average F G Production Qty= INTEG (Change in FPR, 200)
Units: units

Average RM Order Rate= INTEG (Change in Average Order Rate, Mf Order In)
Units: units

Average RM Production Qty= INTEG (Change in RPR, 200)
Units: units

143

(015)

(016)

(017)

(018)

(019)

(020)

(021)

(022)

(023)

(024)

(025)

(026)

Average RM Shipment Rate= INTEG ((RM Shipping-Average RM Shipment
Rate)/Time to average RM shipment, RM Shipping)
Units: units

Average Sales Rate= INTEG ((FG Sales-Average Sales Rate)/Time to average
sales, F G Sales)
Units: units

Buyer's Bargaining Power = WITH LOOKUP ("No. of Suppliers", ([(1,0.1)-
(20,2)],(1,0.l),(3.61468,0.191667),(5.76453,0.325),(7.74006,0.5 16667),(9.07645,
0.716667),(10,1),(l1.4587,1.39167),(l3.0856,1.65833),(15.2936,].81667),(17.61
77,1 .925),(20,2) ))

Units: Dmnl

Buyer's Cumulative Profit= INTEG (Profit, 300000)
Units: dollars

"Buyer's Non-contractible Investment"=IF THEN ELSE( Buyer's Cumulative
Profit<0, 0, Buyer's Cumulative Profit*Fraction of Buyer's Profit to NC
Investment“Effect of Supplier's NCI )

Units: dollars/week

Buyer's Revenue=F G Sales‘F G Price
Units: dollars/week

Change in Average Order Rate=(Mf Order In-Average RM Order Rate)/Time to
Average Order Rate
Units: units/week

Change in Demand Rate=
(Demand-Average Demand Rate)/Time to Average Demand Rate
Units: units/week

Change in F0 Price=
(Indicated FG Price-PG Price)/F G Price Adjustment Time
Units: dollars/unit/week

Change in FPR=

(Start Processing-Average FG Production Qty)/Time to average FPR
Units: units/week
Change in IT leve1=

(Target IT level-IT level)/IT Adjustment Time
Units: Dmnl

"Change in No. of Suppliers"=("Desired No. of Suppliers"-"No. of

144

(027)

(028)

(029)

(030)

(031)

(032)

(033)

(034)

(035)

(036)

(037)

(03 8)

(039)

Suppliers")/Supplier Adjustment Time

Units: Persons/unit/week

Change in RM Price=(lndicated RM Price-RM Price)/RM Price Adjustment Time
Units: dollars/unit/week

Change in RPR=(Start Producing-Average RM Production Qty)/Time to average
RPR

Units: units/week

Defect Rate=0
Units: Dmnl

Demand="AR: Current Dmd"
Units: units/week

Demand Backlog= INTEG (Demand-Fulfillment, Desired Demand Backlog)
Units: units

Demand Backlog Adjustment=(Demand Backlog-Desired Demand
Backlog)/Time to adjust Demand Backlog
Units: units/week

Desired Demand Backlog=Average Demand Rate*Normal FG Delivery Delay
Units: units

Desired F G Production=F G Inventory Adjustrnent+Average Demand Rate
Units: units

Desired FGWIP=Desired FG Production‘Mf Lead Time
Units: turits

"Desired No. of Suppliers" = WITH LOOKUP (Transaction Cost,
([(50,l)-(10000,20)],(50,20),(1449.69,19.25),(2575.54,l7.9167),(3579.66,
15.6667),(4370.79,l3.0833),(5000,10),(5527.06,7.16667),(6287.77,4.5),(7322.32
,2.75),(8630.73,1.75),(10000,1) ))

Units: Persons/week

Desired RM Backlog=Average RM Order Rate*Normal RM Delivery Delay
Units: units

Desired RM Production=RM Inventory Adjustrnent+Average RM Order Rate
Units: units

Desired RMWIP=Average RM Shipment Rate*Sp PD Lead Time
Units: units

145

(040)

(041)

(042)

(043)

(044)

(045)

(046)

(047)

(048)

(049)

Effect of AS] on TC = WITH LOOKUP (Investment on Asset Specificity,
([(0,0.5)-(3000,2)],(0,0.5),(403.67,0.552632),(752.294,0.625),(1027.52,0.717105
),(]302.75,0.848684),(l500,1),(1715.6,] .29167),(1935.78,1.56667),(2238.53,

1 .8),(26 l 4.68,] .925),(3 000,2) ))

Units: Dmnl

Effect of Bargaining POWCF Supplier's Bargaining Power/Buyer's Bargaining
Power
Units: Dmnl

Effect of BNCI on TC = WITH LOOKUP ("Buyer's Non-contractible
Investment",
([(0,0.5)-(12000,2)],(0,0.5),(9]7.431,0.565789),(l724.77,0.638158),(2275.23
,0.743421),(2715.6,0.861842),(3000,1),(3853.21,1 .35),(5834.86,1.64167),(8000
,1 .81667),(9944.95,1.925),(12000,2) ))

Units: Dmnl

Effect of Economics of Scales = WITH LOOKUP ("No. of Suppliers",
([(1,0.1)-(20,2)],(],2),(3.14985,].88333),(5.4159,1.75),(7.27523,1.55833
),(8.72783,1.3),(10,]),(10.8777,0.675),(12.3303,0.44]667),(]4.3058,0.283333
),(l7.0948,0.183333),(20,0.]) ))

Units: Dmnl

Effect of FG DS Balance=Relative FG Inventory Coverage
Units: Dmnl

Effect of IT on AS Investment = WITH LOOKUP (IT level,

([(0. ],0.])-(2,2)],(0.1,2),(0.314985,] .89167),(0.547401,l .79167),(0.739144

,1 .625),(0.878593,1.35833),(1,1),(1.13425,0.691667),(1.291 13,0.44]667),(1.5003
1,0.291667),(1.74434,0.183333),(2,0.1) ))

Units: Dmnl

Effect of RM DS Balance=Relative Sp Inventory Coverage
Units: Dmnl

Effect of Supplier's NC] = WITH LOOKUP ("Supplier's Non-contractible
Investment",
([(0,0.1)-(4000,2)],(0,2),(501.529,1.9),(917.431,1.76667),(1333.33,].575
),(1688.07,] .35833),(2000,1),(2250.76,0.775),(2532.1 1,0.491667),(2948.01,0.325
),(3412.84,0.191667),(4000,0.1) ))

Units: Dmnl

EOQ=sqrt( 2*Average Demand Rate*52*RM Price/item cost)
Units: units

FG Inventory= INTEG (+Finish Processing-PG Sales, Target FG Inventory)

146

(050)

(051)

(052)

(053)

(054)

(055)

(056)

(057)

(058)

(059)

(060)

(061)

(062)

(063)

Units: units

FG Inventory Adjustment= (Target FG Inventory-PG Inventory)/F G Inventory
Adjustment Time
Units: units/week

F G Inventory Adjustment Time=8
Units: weeks

FG Inventory Coverage=FG Sales/F G Inventory
Units: Dimensionless

F G Price= INTEG (Change in FG Price, 350)
Units: dollars/unit

F G Price Adjustment Time=4
Units: week

F G Sales=Demand Backlog/Normal FG Delivery Delay
Units: units/week

FG WIP Adjustment Time=6
Units: weeks

FGWIP Adjustment=(Desired FGWIP-Work In Process)/F G WIP Adjustment
Time
Units: units/week

FIC Perception Time=2
Units: weeks

FINAL TIME = 150
Units: week The final time for the simulation.

Finish Processing=Work In Process/Mf Lead Time
Units: units/week

Finish Producing=SpD Work In Process/Sp PD Lead Time
Units: units/week

Fraction of Buyer's Profit to NC Investrnent=0.0]
Units: Dmnl

Fraction of Profit to InvestmenF0.05
Units: Dmnl

147

(064)

(065)

(066)

(067)

(068)

(069)

(070)

(071)

(072)

(073)

(074)

(075)

(076)

(077)

Fraction of reflecting MSF to TRI=Lookup of IT Effect on Fraction(IT level)
Units: Dimensionless

Fraction of Supplier's Profit to NC Investment=0.0]
Units: Dmnl

Fulfillment=F G Sales
Units: units/week

Holding Cost= (FG Inventory+RM Inventory)*item cost
Units: dollars/week

Indicated FG Price=350*Effect of FG DS Balance
Units: dollars/unit

Indicated RM Price=]00*Effect of Bargaining Power“ Effect of RM DS
Balance/Effect of Economics of Scales
Units: dollars/unit

INITIAL TIME = 0
Units: week The initial time for the simulation.

Investment Budget for IT Development=IF THEN ELSE( Buyer's Cumulative
Profit<0, 0, Buyer's Cumulative Profit*Fraction of Profit to Investment)
Units: dollars

Investment on Asset Specificity=Normal Investment on Asset Specificity‘Effect
of IT on AS Investment
Units: dollars/week

IT Adjustment Time= 10
Units: weeks

IT level= INTEG (Change in IT level, 1)
Units: Dmnl

item cost=16.004
Units: dollars/unit

Logistics Cost=IF THEN ELSE(RM Shipping>0, Processing Cost+Transport
Cost, 0)
Units: dollars/week

Lookup of IT Effect on Fraction(
(0. 1 , 1 )-(2,2)],(0. l ,2),(0.3 55657,] .9473 7),(0.55321 1,1.89035),(0.756575,].79386

148

(078)

(079)

(080)

(081)

(082)

(083)

(084)

(085)

(086)

(087)

(088)

(089)

(090)

),(0.90]835,] .67105),(1,1.5),(].09939,1 .35965),(1.24465,1.2193),(1 .45963,] .109
65),(1.721 1,1 .03509),(2,1))
Units: Dmnl

Manufacturing CosFitem cost“ Start Processing*Production Change Ratio
Units: dollars/week

Mf Cap Factor=10
Units: Dmnl

Mf Lead Time=2
Units: weeks

Mf Order In=Periodic RM Order Qty
Units: units/week

Mf Order Out=RM Shipping
Units: units/week

Mf Production Qty=Average F G Production Qty+(Target FG Inventory-F G
Inventory)/F G Inventory Adjustment Time+Demand Backlog
Adjustrnent+(Average Sales Rate‘Mf Lead Time-Work In Process)/F G WIP
Adjustment Time

Units: units/week

Mf Review Period=2
Units: weeks

Mf Review Period2=(EOQ/(Average Demand Rate*52))*52
Units: weeks

Mfg Sales Forecast=SMOOTH( Demand, Smth Time)
Units: writs/week

"No. of Suppliers"= INTEG ("Change in No. of Suppliers", 10)
Units: Persons/week

Normal FG Delivery Delay=2
Units: weeks

Normal Investment on Asset Specificity=1500
Units: dollars/week

Normal RM Delivery Delay: 2
Units: weeks

149

(091)

(092)

(093)

(094)

(095)

(096)

(097)

(098)

(099)

(100)

(101)

(102)

(103)

(104)

Perceived FIC=SMOOTH( FG Inventory Coverage, FIC Perception Time )
Units: Dmnl

Perceived SIC=SMOOTH( Sp Inventory Coverage, SIC Perception Time )
Units: Dmnl

Periodic RM Order Qty=Mf Production Qty
Units: units/week

Processing Cost=20
Units: dollars/week

Product Customization Level=1 -.
Units: Dmnl I”

Production Change Ratio = WITH LOOKUP (Product Customization Level,
([(0.1,0.625)-(2,1.255)],(0.1,0.625),(0.367278,0.667873),(0.599694,0.723355
),(0.779817,0.799013),(0.890214,0.892325),(1,1),(l.14006,1.08368),(1.34924
,1 .14724),(1.55841,1.19145),(1.77339,1.22737),(2,1.255) ))

Units: Dmnl

Profit=Buyer's Revenue-Total Operating Cost
Units: dollars/week

Purchasing Cost=RM Price*RM Shipping
Units: dollars/week

Reference FG Inventory Coverage=0. 158447
Units: Dimensionless

Reference Sp Inventory Coverage=0.155574
Units: Dimensionless

Relative FG Inventory Coverage=Perceived F IC/Reference FG Inventory
Coverage I
Units: Dmnl

Relative Sp Inventory Coverage=Perceived SIC/Reference Sp Inventory
Coverage
Units: Dmnl

 

 

 

 

RM Arriving to Mf=RM in Transit/Transit Lead Time
Units: units/week

RM Backlog= INTEG (Mf Order In-Mf Order Out, Desired RM Backlog)
Units: units

150

(105)

(106)

(107)

(108)

(109)

(110)

(111)

(112)

(113)

(114)

(115)

(116)

(117)

(118)

RM Backlog Adjustment=(RM Backlog-Desired RM Backlog)/Time to adjust
RM Backlog
Units: units/week

RM in Transit= INTEG (+RM Shipping-RM Arriving to Mf, 400)
Units: units

RM Inventory= INTEG (RM Arriving to Mf-Start Processing, 400)
Units: units

RM Inventory Adj ustrnent=(Target RM Inventory-Sp Inventory)/ Sp Inventory
Adjustment Time
Units: units/week

RM Price= INTEG (Change in RM Price, 100)
Units: dollars/unit

RM Price Adjustment Time= 4
Units: week

RM Shipping=RM Backlog/Normal RM Delivery Delay
Units: units/week

RMWIP Adjustment= (Desired RMWIP-SpD Work In Process)/Sp WIP
Adjustment Time
Units: units/week

SAVEPER = TIME STEP
Units: week [0,?] The frequency with which output is stored.

"SF: Mf'=1.28*"AR: StDev"
Units: Dmnl

"SF:Sp"=1.28*"AR: StDev"
Units: Dmnl

SIC Perception Time=2
Units: weeks

Smth Time=4
Units: weeks

Sp Cap Factor=10
Units: Dmnl

151

(119)

(120)

(121)

(122)

(123)

(124)

(125)

(126)

(127)

(128)

(129)

(130)

(131)

Sp Inventory= INTEG (Finish Producing-RM Shipping, Target RM Inventory)
Units: units

Sp Inventory Adjustment Time=8
Units: weeks

Sp Inventory Coverage=RM Shipping/Sp Inventory
Units: Dimensionless

Sp PD Lead Time=2
Units: weeks

Sp Production Qty=Average RM Production Qty+(Target RM Inventory-Sp
Inventory)/ Sp Inventory Adjustment Time+RM Backlog Adj ustment+(Average
RM Shipment Rate*Sp PD Lead Time-SpD Work In Process)/Sp WIP
Adjustment Time

Units: units/week

Sp WIP Adjustment Time=6
Units: weeks

SpD Work In Process= INTEG (+Start Producing-Finish Producing, Average RM
Shipment Rate*Sp PD Lead Time)
Units: units

Start Processing=MIN( Mf Production Qty/(l-Defect Rate), "AR: Mean"*Mf Cap
Factor )
Units: writs/week

Start Producing=MIN( Sp Production Qty, "AR: Mean"*Sp Cap Factor)
Units: units/week

Supplier Adjustment Time=8
Units: weeks

Supplier's Bargaining Power = WITH LOOKUP ("No. of Suppliers",
([(1,0.1)-(20,2)],(],2),(3.324]6,1.925),(5.5321l,1.81667),(7.85627,1.59]67
),(9. 13456,] .30833),(10,1),(1 1.2844,0.725),(13.1437,0.533333),(15.2355,0.35
),(]7.5596,0.208333),(20,0.1) ))

Units: Dmnl

Supplier's Cumulative Profit= INTEG (Supplier's Profit, 200000)
Units: dollars/week

Supplier's inventory cost=13. ] 51*Sp Inventory
Units: dollars/unit

152

 

(132)

(133)

(134)

(135)

(136)

(137)

(138)

(139)

(140)

(141)

(142)

(143)

Supplier's manufacturing cost=13.]51*Start Producing
Units: dollars/unit

"Supplier's Non-contractible Investrnent"=IF THEN ELSE( Supplier's Cumulative
Profit<0 , 0 , Supplier's Cumulative Profit*Fraction of Supplier's Profit to NC
Investment )

Units: dollars/week

Supplier's Profit=Supplier's Revenue-Supplier's total operating cost
Units: dollars/week/week

Supplier's Revenue=RM Price‘RM Shipping
Units: dollars/week

Supplier's total operating cost=Supplier's manufacturing cost+Supplier's inventory
cost+"Supplier's Non-contractible Investment"
Units: dollars/week

Target F G Inventory= Mfg Sales Forecast“(Transit Lead Time+Mf Lead
Time+Mf Review Period)+"SF: Mf'*sqrt(Transit Lead Time+Mf Lead Time+Mf
Review Period)

Units: units

Target IT level = WITH LOOKUP (Investment Budget for IT Development,
([(0,0.1)-(30000,2)],(0,0. 1),(3853.21,0.183333),(7339.45,0.3),(10550.5,0.483333
),( l 321 1,0.716667),(]5000,1),(16880.7,1 .3),(l9266.1,1 .575),(22293.6,1 .79167
),(25779.8,l .90833),(30000,2) ))

Units: Dmnl

Target RM Inventory=Fraction of reflecting MSF to TRI*(Mfg Sales
Forecast*(Sp PD Lead Time+Mf Review Period)+"SF:Sp"*sqrt(Sp PD Lead

Time+Mf Review Period))
Units: units
TIME STEP = 1

Units: week [0,?] The time step for the simulation.

Time to adjust Demand Backlog=8
Units: weeks

Time to adjust RM Backlog=8
Units: weeks

Time to Average Demand Rate=4
Units: weeks

153

 

 

(144)

(145)

(146)

(147)

(148)

(149)

(150)

(151)

(152)

(153)

(154)

(155)

(156)

Time to average F PR=4
Units: weeks

Time to Average Order Rate=4
Units: weeks

Time to average RM shipment=4
Units: weeks

Time to average RPR=4
Units: weeks

Time to average sales=4
Units: weeks

Time to Update Demand=2
Units: weeks

Total Operating Cost=Manufacturing Cost + Holding Cost + Logistics Cost +
Purchasing Cost+1nvestment Budget for IT Development+Transaction
Cost+1nvestment on Asset Specificity+"Buyer's Non-contractible Investment"
Units: dollars/week

Transaction Cost=5000*Effect of A81 on TC*Effect of BNCI on TC
Units: dollars/week

Transit Lead Time=2
Units: weeks

Transport Cost=Transport Rate(RM Shipping/U nit Order)
Units: dollars/week

Transport Rate((0,0)-

(1000,1000)],(0,0),(1 1 0.092,61 .4035),(200, 123. 1 54),(32] . 101 ,2] 0.526
),(437.309,355.263),(500,500),(57l865,653.509),(666.667,771.93),(770.642,
868.421),(877.676,942.982),(]000,1000))

Units: dollars/week

Unit Order=1
Units: unit/week

Work In Process= INTEG (+Start Processing-Finish Processing, Average Sales
Rate*Mf Lead Time)

Units: units

154

 

 

BIBLIOGRAPHY

Anderson, E., and D. Schmittlein, 1984, Integration of the Sales Force: An Empirical
Examination, Rand Journal of Economics 15(Autumn), 385-395.

Anderson, E., and H. Gatignon, 1986, Modes of Foreign Entry: A Transaction Cost
Analysis and Propositions, Journal of International Business Studies 17, 1-26.

Anderson, J .C., H. Hakansson, and J. Johansson. 1994, Dyadic Business Relationships
within a Business Network Context, Journal of Marketing 58, October, 1-15.

Arora, S., 2000, Reengineering: A focus on enterprise integration, Interfaces 30(5), 54-
71.

Bakos, J.Y., 1987, Interorganizational Information Systems: Strategic Opportunities for
Competition and Cooperation. Ph. D. Dissertation, Sloan School of Management,
Massachusetts Institute of Technology, Cambridge, MA.

Bakos, J .Y., 1991, A Strategic Analysis of Electronic Marketplaces. MIS Quarterly 15(3),
295-3 10.

Bakos, J.Y., and E. Brynjolfsson. 1993, From Vendors to Partners: Information
technology and incomplete contracts in buyer-supplier relationships. Journal of
Organizational Computing 3(3), 301-328.

Bardi, E.J., T.S. Raghunathan, and PK. Bagchi, 1994, Logistics Information
SystemszThe Strategic Role of Top Management. Journal of Business Logistics
15(1), 71-85.

Barrett, 8., 1986, Strategic Alternatives and Inter-organizational System
Implementations: An Overview, Journal of Management Information Systems 3(3),
5-16.

Bensaou, B.M., 1999, Portfolios of Buyer-Supplier Relationships, Sloan Management
Review 40(4), 35-44.

Bensaou, B.M. and N. Venkatraman, 1996, Not by Partnership Alone: Managing a
Portfolio of Relationships, INSEAD Working Paper.

Berry, D., D.R. Towill, and N. Wadsley, 1994, Supply Chain Management in the
Electronics Products Industry, International Journal of Physical Distribution and
Logistics Management 24(10), 20-32.

155

Bourland, K.E., S.G. Powell, and DR Pyke, 1996, Exploiting Timely Demand
Information to Reduce Inventories, European Journal of Operational Research 92(2),
239-253.

Bowersox, D.J., 1989, Logistics In The Integrated Enterprise, the Annual Conference of
the Council of Logistics Management(St. Louis, MO).

Bowersox, D.J., and DJ. Closs, 1996, Logistical Management: The Integrated Supply
Chain Process, The McGraw-Hill Companies.

Bowersox, D.J., and PJ. Daugherty, 1995, Logistics Paradigms: The Impact of
Information Technology. Journal of Business Logistics 16(1), 65-80.

Brynjolfsson, E., T. Malone, V. Gurbaxani, and A. Kambil, 1991, Does Information
Technology Lead to Smaller Firms? Technical Report 123. Massachusetts Institutue
of Technology, Center Coordination Science, Cambridge, MA.

Buxbaum, P.A., 2000, Web of Collaboration. Executive Edge 2(3), 1-5.
Byme, RM. and WJ. Markham, 1991, Improving Quality and Productivity in the
Logistics Processes: Achieving Customer Satisfaction Breakthroughs, Oak Brook,

IL:Council of Logistics Management.

Cachon, GP. and M. Fisher, 2000, Supply Chain Inventory Management and the Value
of Shared Information, Management Science 46(8), 1032-1048.

Cannon, JP and C. Homburg, 2001, Buyers-supplier relationships and customer firm
costs, Journal of Marketing 65(1), 29-43.

Carter, J .R. and R. Narasimhan. 1995, Purchasing and Supply Management : Future
Directions and Trends. Tempe, AZ: Center for Advanced Purchasing Studies.

Chen, F ., 1998, Echelon Reorder Points, Installation Reorder Points, and the Value of
Centralized Demand Information. Management Science 44(125), 221-234.

Chiu, H.N., 1995, The integrated logistics management system: A framework and case
study, International Journal of Physical Distribution & Logistics Management 25(6),
4-22.

Chopra, S. and P. Meindl, 2001, Supply Chain Management, Prentice-Hall, Inc., NJ.

Christopher, M., 1992, Logistics and Supply Chain Management, British.

Christopher, M., 1994, Logistics and Supply Chain Management, Burr Ridge, IL:
Financial Times.

156

Clemons, E.K., and M. Row, 1989, Information Technology and Economic
Reorganization. Proceedings of the 10’” International Conference on Information
Systems, Boston, MA.

Clemons, E.K., S.P. Reddi, and M. Row. 1993, The Impact of Information Technology
on the Organization of Economic Activity: The “Move to the Middle” Hypothesis.
Journal of Management Information Systems 10(2), 9-36.

Closs, D.J., A.S. Roath, T.J. Goldsby, J .A. Eckert, and SM. Swartz, 1998, An empirical
comparison of anticipatory and response-based supply chain strategies.
International Journal of Logistics Management 9(2), 21-34.

Cohen, MA. and S. Mallik, 1997, Global Supply Chains: Research and Applications.
Production and Operations Management 6(3 ), 193-210.

Cooper, R. and R. Slagrnulder, 1999, Supply Chain Management for Lean Enterprises:
Interorganizational Cost Management, Strategic Finance 80 (April), 15-16.

Cooper, MC. and L.M. Ellram, 1993, Characteristics of Supply Chain Management and
the Implications for Purchasing and Logistics Strategy, The International Journal of
Logistics Management 4(2), 13-22.

Coyle, R.G., 1977, Management System Dynamics, London; New York: Wiley.
Cusurnano, MA. and A. Takeishi, 1991, Supplier Relations and Management: A Survey

of Japanese, Japanese-Transplant, and US. Auto Plants, Strategic Management
Journal 12, 563-588.

Daugherty, P.J. 1994. Strategic Information Linkage. The Logistics Handbook. The Free
Press. 757-769.

Disney, S.M., M.M. Naim, D.R. Towill, 1997, Dynamic simulation modelling for lean
logistics, International Journal of Physical Distribution & Logistics Management
27(3/4), 174-196.

Dowlatshahi, S., 1999, Bargaining power in buyer-supplier relationships, Production and
Inventory Management Journal 40(1), 27-35.

Dyer, J .H., 1996, Specialized Supplier Networks as a Competitive Advantage: Evidence
from the Auto Industry, Strategic Management Journal 17, 271-291.

Earl, M., 2000, Opinion: How to be a CEO for the information age, Sloan Management
Review 41(2), 11-24.

Eisenhardt, K.M., 1985, Control Organizational and Economic Approaches, Management
Science 31, February, 134-149.

157

 

 

Fisher, ML, 1997, What is the Right Supply Chain for Your Product?, Harvard Business
Review 75(2), March-April, 105-117.

Forrester, J .W., 1961, Industrial Dynamics, Cambridge, MA: The MIT Press.
Galbraith, J.R., 1973, Designing Complex Organizations, MAzAddison-Wesley.

Garvineni, S., R. Kapuscinski, and S. Tayur, 1996, Inventories in Supply Chains under
Cooperation, Proceedings of INFORMS, Washington DC.

Gossman, J., 1997, Presentation to Supply Chain Management Council Meeting, Kellog
Center, Michigan State University.

Govindarajan, V., 1985, Decentralization, Strategy, and Effectiveness of Strategic
Business Units in Multibusiness Organizations, Academy of Management Review 11,
844-856.

Gourley, C., 1998, What's driving the automotive supply chain?". Warehousing
Management 5(10), 44-48.

Grahovac, J. and A. Chakravarty, 2001, Sharing and Lateral Transshipment of Inventory
in a Supply Chain with Expensive Low-Demand Items. Management Science 47(4),
579-594.

Gregory, A., 1995, Make supply chain technology work for you, Works Management
48(7), 37-39.

Grossman, SJ. and OD. Hart, 1986, The Costs and Benefits of Ownership: A Theory of
Vertical and Lateral Integration, The Journal of Political Economy 94(4), 691-719.

Harnbrick, DO, 1983, An Empirical Typology of Mature Industrial-Product
Environments, Academy of Management Journal 26, 213-230.

Hammant, J ., 1995, Information technology trends in logistics, Logistics Information
Management 8(6), 32-37.

Helper, S., 1991a, How Much Has Really Changed between US. Automakers and Their
Suppliers? Sloan Management Review, Summer, 15-27.

Helper, S., 1991b, Supplier Relations and Investment in Automation: Results of Survey

Research in the US. Auto Industry. Working Paper, Dept. of Economics, Case
Western Reserve University, Cleveland, OH, September.

158

 

Henderson, J. and N. Venkatraman, 1990, Strategic Alignment: A Model for
Organizational Transformation via Information Technology. CISR/Sloan Working
Paper, Massachusetts Institute of Technology, Cambridge, MA. November.

Hewitt, F., 1994, Supply Chain Redesign. The International Journal of Logistics
Management 5(2), 1-8.

Hoekstra, S. and J. Rome, 1991, Integral Logistics Structures, Industrial Press Inc.

J anszen, F., 2000, The age of innovation .' making business creativity a competence, not a
coincidence, London : Financial Times Prentice Hall.

Johnston, R. and P. Lawrence, 1988, Beyond Vertical Integration--the Rise of the Value-
Adding Partnership. Harvard Business Review, July-August, 94-101.

Jones, D.T., P. Hines, and N. Rich, 1997, Lean Logistics, International Journal of
Physical Distribution and Logistics Management 27(3/4), 153-173.

Kalakota, R. and AB. Whinston, 1996, Frontiers of Electronic Commerce, Addison-
Wesley Publishing Company, Inc., Massachusetts.

Kalakota, R. and AB. Whinston, 1997, Electronic Commerce: A Manager’s Guide,
Addison-Wesley Longrnan, Inc., Massachusetts.

Karoway, C., 1997, Superior Supply Chains Pack Plenty of Byte, Purchasing Today
8(11), 32-35.

Keen, P., 1993, Information Technology and the Management Difference: A Fusion Map,
IBM Systems Journal 32, 17-39.

Kim, I.H., 1998, Information Sharing In a Supply Chain, Working Project in Purdue
University.

Kim, S.C., 1999, Strategic Choices in the Development of a Business-To-Business
Electronic Commerce System, Communications of the ACM 42, 36-42.

Kraljic, P., 1983, Purchasing Must Become Supply Management, Harvard Business
Review 61(5), 109-117.

Krajewski, L., and L. Ritzman, 2000, Operations Management: Strategy and Analysis, 5th
eds., Addison-Wesley.

Lambert, D.M., M.C. Cooper, and JD. Pagh, 1998, Supply chain management:
Implementation issues and research opportunities, International Journal of Logistics
Management 9(2), 1-19.

159

 

Lassar, W.M. and J .L. Kerr, 1996, Strategy and Control in Supplier-Distributor
Relationships: An Agency Perspective, Strategic Management Journal 17, 613-632.

Lawrence, A., 1997, Customer power forces supply chain integration, Works
Management 50(4), 43-45.

Lee, H.L., K.C. So, and CS. Tang, 2000, The Value of Information Sharing in a Two-
Level Supply Chain. Management Science 46(5), 626-643.

Lee, H.L., V. Padmanabhan and S. Whang, 1997a, The Bullpr Effect in Supply
Chains, Sloan Management Review, Spring, 93-102.

Lee, H.L., V. Padmanabhan, and S. Whang, 1997b, Information Distortion in a Supply
Chain: The Bullpr Effect, Management Science 43(4), 546-558.

Levine, R. and H. Fitzgerald, 1992, Analysis of Dynamic Psychological Systems (2 vols.),
New York: Plenum Press.

Macbeth, D., and N. Ferguson, 1994, Partnership Sourcing: An Integrated Supply Chain
Approach, Pitrrran Publishing, London.

Magee, J .F., W.C. Copacino, and DB. Rosenfield, 1985, Modern Logistics Management,
John Wiley & Sons: New York.

Magretta, J ., 1998, The power of virtual integration: An interview with Dell Computer's
Michael Dell, Harvard Business Review 76(2), 72-84.

Malone, T.W., 1985, Organizational Structure and Information Technology: Elements of
Formal Theory, Working Paper I 30, Center for Information System Research, Sloan
School of Management, Massachusetts Institute of Technology, Cambridge, MA,
August.

Malone, T.W., J. Yates, and RI. Benjamin, 1987, Electronic Markets and Electronic
Hierarchy, Communications of the ACM 30(6), 484-497.

McMillan, J., 1990, Managing Suppliers: Incentive Systems in Japanese and US.
Industry. California Management Review 32(4), 38-55.

Milgrom, P. and J. Roberts, 1990, The Economics of Modern Manufacturing:

Technology, Strategy, and Organization, The American Economic Review 80(3),
511-528.

Miller, D., 1987, The Structural and Environmental Correlates of Business Strategy,
Strategic Management Journal 8, 55-76.

160

 

Miller, D., 1988, Relating Porter’s Business Strategies to Environment and Structure:

Analysis and Performance Implications, Academy of Management Journal 31, 280-
308.

Miller, D., and RH. Friesen, 1986, Porter’s(1980) Generic Strategies: An Empirical
Examination with American Data, Organization Studies 7, 37-55.

Moinzadeh, K., 2002, A Multi-Echelon Inventory System with Information Exchange.
Management Science 48(3), 414-426.

Morash, EA. and SR. Clinton, 1998, Supply Chain Integration:Customer Value Through
Collaborative Closeness Versus Operational Excellence, Journal of Marketing
Theory and Practice 6(4), 104-120.

Morgan, J .P., 2000, New solutions to making EDI really work in purchasing, Purchasing
129(6), 42.

Narasimhan, R., 1979, Analyzing the Value of Information using System Dynamics, MIS
Quarterly 3, April.

Pedler, M., 1994, Information technology in the supply chain, Purchasing & Supply
Management, (September), 15-16.

Pine, B.J., B. Victor, and AC. Boynton, 1993, Making Mass Customization Work,
Harvard Business Review (Sept-Oct), 108-1 19.

Plouffe, C.R., M. Vandenbosch, and J. Hulland, 2001, Intermediating Technologies and
Multi-group Adoption: A Comparison of Consumer and Merchant Adoption
Intentions toward a New Electronic Payment System, The Journal of Product
Innovation Management 18, 65-81.

Porter, M.E., 1980, Competitive Strategy. New York, NY: The Free Press.

Radstaak, B.G., and M.H. Ketelaar, 1998, Worldwide Logistics: The Future of Supply
Chain Services, Holland International Distribution Council, Hague, The Netherlands.

Richardson, GR, 1991, Feedback Thought in Social Science and Systems Theory,
University of Pennsylvania Press, Philadelphia.

Shore, B., 200], Information sharing in global supply chain systems.
Journal of Global Information Technology Management 4(3), 27-50.

Stalk, G.H., and TM. Hout, 1990, Competing against Time; How Time-based
Competition is Reshaping Global Markets, New York: Free Press.

161

 

 

Stevens, G., 1989, Integrating the Supply Chain, International Journal of Physical
Distribution and Materials Management 19(8), 3-8.

Sterman, J .D., 2000, Business Dynamics: Systems Thinking and Modeling for a Complex
World, Boston: Irwin McGraw Hill.

Towill, D.R., M.M. Naim, and J. Wikner, 1992, Industrial Dynamics Simulation Models
in the Design of Supply Chains, International Journal of Physical Distribution and
Logistics Management 22(5), 3-13.

Vollmann, T.E., C. Cordon, and J. Heikkila, 2000, Teaching Supply Chain Management
to Business Executives, Production and Operations Management 9(1), 81-90.

Ward, P.T., J. Deborah, G. Bickford, and K. Leong, 1996, Configurations of
Manufacturing Strategy, Business Strategy, Environment and Structure, Journal of
Management 22(4), 597-626.

Wikner, J., D.R. Towill, and M. Naim, 1991, Smoothing Supply Chain Dynamics,
International Journal of Production Economics 22(3), 231-248.

Williamson, 0. E., 1975, Markets and Hierarchies: Analysis and Antitrust Implications,
New York: Basic Books.

Williamson, 0. E., 1979, Transaction cost economics: The governance of contractual
relations, Journal of Law and Economics 22, 3 - 6].

Williamson, CE, 1981, The Modern Corporation: Origins, Evolution, Attributes,
Journal of Economic Literature 19(4), 1537-1569.

Wiiliamson, DE, 1985, Asset Specificity and Economic Organization, International
Journal of Industrial Organization 3(4), 365-379.

Winter, R.A., 1993, Vertical Control and Price versus Nonprice Competition, Quarterly
Journal of Economics 108, 61-76.

Zheng, J., T.E. Johnsen, C.M. Harland, and RC. Larnming, 2001, A Taxonomy of
Supply Networks, Proceedings of the 10'" International Annual IPSERA Conference,
Jonkoping, Sweden.

162