A longitudinal approach to the effects of partner firm characteristics, the environment, and mutual trust on synergistic...


Material Information

A longitudinal approach to the effects of partner firm characteristics, the environment, and mutual trust on synergistic outcomes in long term, buyer-supplier relationships
Physical Description:
xiii, 245 leaves : ill. ; 29 cm.
Jap, Sandy D
Publication Date:


bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1995.
Includes bibliographical references (leaves 181-197).
Statement of Responsibility:
by Sandy D. Jap.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002057312
notis - AKP5335
oclc - 33825432
System ID:

Full Text







Copyright 1995


Sandy D. Jap

To my grandparents, with gratitude.


This dissertation was not the result of my efforts

alone. It was supported by financial grants from the

Marketing Science Institute (Clayton and grant #4-882), and

data support from the Institute for the Study of Business


My committee members and James Algina provided me with

critical comments during the initial stages. Their insights

strengthened the positioning and overall contribution of the

dissertation. Michael Browne, Claes Fornell, and Prasad Naik

gave helpful comments concerning the analysis. Joffre Swait

and Mark Schmit were particularly helpful in the final stages

of the analysis, getting me through some panic-stricken

moments. Kathy Brown and Margaret Jones have been an

invaluable help in the implementation of this study as well

as numerous other projects.

There are also many people in the department who have

impacted me throughout the doctoral program with their

advice, insight, and friendship. I am a richer person for

having known these individuals: Susan Fournier, Joffre

Swait, Alan Sawyer, Rich Lutz, John Lynch, Joe Alba, Steve

Shugan. It goes without saying that all of the doctoral

students have been exceptional cohorts--we are forever bound

by our "Florida experience."

The one person who has influenced me the most during the

program has undoubtedly been Bart Weitz. If I had to

characterize the advisor best suited for me, it would have to

be Bart, straight down the line. He has always given me the

freedom to handle situations and solve problems in my own

way, without abandoning me or allowing me to walk down roads

of futility. Working with him has taught me much about the

research process--from the conceptualization, to the

implementation, the analysis, and the selling. His advice

and insights, encouragement and support, have helped me to

rc.w personally and as a scholar. If I eventually become

even half the researcher, teacher, scholar, and mentor that

he is, I will have had a great career.

There are a number of people in my personal life who

have been with me through the ups and downs, the good and

bad, the pretty and ugly. Their friendship has been

invaluable to me: Chansone Durden, Susan Lewis, Stacey

Garnett, Tamar Riley, Suzanne Quintero, Chrissy Conkling,

David Allegood. It is my privilege to know them.

My parents and siblings have been a wonderful suEport

throughout the program, always praying for the best for me

and encouraging me to take hold of the good opportunities

that come my way. There are not enough words to describe how

much each of you mean to me. Without my grandparents, my

education would have been impossible--emotionally,

financially, practically. If a star fell out of the sky for

every good thing they've done for me, every kind word they

have said, every prayer that has been said for me, the sky

would be empty.

Finally and most importantly, I thank my God and

Savior, Jesus Christ. He has been more than good to me, more

than faithful. Apart from Him I can do no good thing.



AC:Kr], LEDGrIMErITS ............................................ iv

ABSTRACT ...................................... ............. xii


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

Strategic Advantage Through Buyer-Supplier
Relationships ........................................ 3
Needed Research on Strategic Advantage ................ 6
Purpose and Scope of the Dissertation ................. 7
Organization of the Dissertation ......................... 9

/2 REVIEW OF THE LITERATURE ................................. 12

Overview ............................................... 12
,Transaction Costs Analysis ............................. 12
S Relational Exchange ..................................... 17
Political Economy Paradigm .............................. 19
Key Contributions of the Frameworks .................. 22
Summary ................................................ 31

3 CONCEPTUAL FRAMEWORK .................................... 33

Overview ............................................... 33
Developing Strategic Advantages ....................... 33
Antecedents to Developing Strategic Advantaaes ........ 36
Consequences of Developing Strategic Advantages ...... 46
The Longitudinal Conceptual Model ..................... 47
Summary ............................................... 54

4 METHODOLOGY .......................................... 62

Overview ............................................... 62
The Longitudinal Design ................................. 62
Key Informant Methodology ............................ 66
Research Setting and Sample ............................. 69
Procedure ............................................. 74
Questionnaire and Scale Development ................... 76


Construct Validity at Time One ........................
Summary ...............................................

5 RESULTS ...............................................

Overv iew ..............................................
The Dyadic Measurement Model .......................
Measurement at Time Two ...............................
Longitudinal Measurement Model ........................
Structural Model Analysis .............................
Summary ...............................................

6 DISCUSSION ............................................

Overv iew ..............................................
Discussion of Results ..................... ...........
Contributions of the Dissertation .....................
Implications for Management ...........................
Directions for Future Research ........................

Summary ....

REFE .EIIr C ES .................................................


A BUYER SURVEY: TIME 1 .................................

B SUPPLIER SURVEY: TIME 1 ..............................

C BUYER SURVEY: TIME 2 .................................

D SUPPLIER SURVEY: TIME 2 ..............................

BIOGRAPHICAL SKETCH ........................................











. . . .


Table paGe

4-1 Participating Firms and Divisions .................. 87

4-2 Characteristics of the Exchange Relationship ....... 88

4-3 Scale Items, Factor Loadings, Standard Errors,
and Reliabilities at Time One ..................... 89

4-4 Measurement Models and Correlation Matrices,
Means, and Standard Deviations at Time One ......... 94

4-5 Tests of Construct Invariance Across Buyers and
Suppliers at Time One ................................ 96

5-1 Trust Correlated Uniqueness Model Estimates ....... 135

5-2 Assessment of Perceptual Agreement Among Informants 137

5-3 Scale Items, Factor Loadings, Standard Errors, and
Reliabilities at Time Two ......................... 138

5-4 Measurement Models and Correlation Matrices, Means,
and Standard Deviations at Time Two ................ 143

5-5 Longitudinal Measurement Model Parameter Estimates
for the Development of Strategic Advantages ....... 145

5-6 Tests of Longitudinal Measurement Model Invariance 146

5-7 Restricted Parameter Values for Factor Loadings and
Errors .......... ........ ........................... 149

5-8 Cross Sectional Structural Model Parameter
Estimates ........................................... 150

5-9 Cross Sectional Structural Model Parameter
Estimates Without Environmental Variables ......... 151

5-10 Cross Sectional Structural Model Summaries ......... 152

5-11 Cross Sectional Saturated Model Summaries .......... 154

Table pace

5-12 Longitudinal Structural Model Parameter Estimates .. 156

5-13 Longitudinal Saturated Model Summaries ............. 158


Figure Pa4e

3-1 Overview of the Strategic Advantage Process Model .. 56

3-2 Antecedents to the Development of Strategic
Advantage .......................................... 57

3-3 The Moderating Role of Trust .......................... 58

3-4 Consequences of Developing Strategic Advantages .... 59

3-5 Overview of Time 1 Model ............................. 60

3-6 The Longitudinal Conceptual Model .................. 61

5-1 The Correlated Uniqueness Model .................... 132

5-2 The Longitudinal Measurement Model ................. 133

5-3 Revised Longitudinal Structural Model .............. 134

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Sandy D. Jap

August, 1995

Chairman: Barton A. Weitz
Major Department: Marketing

This dissertation examines the antecedents and

consequences to the development of strategic advantage in

long term, buyer-supplier relationships. A conceptual

framework involving potential antecedents--partner firm

characteristics, characteristics of the environment, and

mutual trust--is hypothesized to lead to the development of

strategic advantages over competing dyads in the marketplace.

This activity expands the level of joint benefits available

to the dyad and enables it to achieve results of greater

consequence than either firm would have been able to

accomplish individually.

This process is examined over time via a longitudinal

survey methodology. Over 200 buyer-supplier dyads of four


Fortune 50 firms participated in the study; each individual

completed two sur'.'e's on the relationship, administered one

year apart. Data analysis was completed using latent

variables and structural equation modeling techniques.

The results point to the importance of goal

compatibility, complementary competencies and mutual trust

as key factors in the decision to develop strategic

advantages together. Goal compatibility provides assurance

of non-o:portunistic behavior early in the relationship and a

reason for relationship continuity afterwards. Complementary

competencies communicate powerful assurances of the payoffs

from developing strategic advantages, particularly for

buyers. Trust is also shown to play a dynamic role in

facilitating relationship outcomes and shaping member

perceptions throughout the process, despite the fact that it

becomes less important in the firms' decision to work closely

together over time. The process of developing strategic

advantages involves learning about the other member that goes

beyond a surface level knowledge. As a result, idiosyncratic

assets created in the process tend to decrease complementary

compeencies of the member firms. In general though, the

active development of strategic advantages is shown to lead

to synergistic results such as strategic advantages over

competing dyads, increased joint profits, and the creation of

idiosyncratic assets.



The development of competitive advantages--positions of

superior performance--is key to a firm's survival. Scholars

and corporate strategists emphasize that firms can attain

sustainable, competitive advantages through product

differentiation, brand loyalty, technological superiority,

and supplier relationship. However, we live in a time in

which these bases can be quickly eroded. The number of new

products introduced each year increases while the number of

true product innovations decreases; the result is rapid

product parity. The growing strength of private labels and

mismanagement of brand extensions has contributed to a

decline in national brand loyalty. Finally, the speed by

which faster and more powerful advances in technology are

developed enables firms to leap frog competitors, converting

a disadvantage into an advantage.

As a result, there is a growing interest among both

practitioners and academics in strategic advantage developed

through value chain relationships. A value chain describes

the path that products take from the raw material state to

the finished product state before they are ultimately sold to

final consumers. Each firm in the chain adds value to the

product before passing it downstream to the next member. For

example, a raw materials supplier might harvest or collect a

raw material and break it down into salable units for

manufacturers. The manufacturer then processes these units,

often changing the form of the product and bringing it to its

finished state. The product is then sold to final customers

or distributors who would further process the product (e.g.,

repackage the product in smaller, salable quantities, provide

services) and sell it to retailers or individual consumers.

Historically, each firm in the value chain would seek to

maximize its individual position, extracting the lowest costs

possible from its suppliers and maximizing its price to

customers. However, as industries downsize, competition

increases, and the players who remain are those who have

learned how to increase the efficiency and effectiveness of

the whole chain. Companies in successful value chains take a

more holistic view of the process of bringing a product to

market, demanding that every member in the chain strive to

coordinate its efforts and increase the efficiency of the

entire chain of firms.

Customers will demand low-cost products from

manufacturers, who in turn, demand low-cost products from

their suppliers. Firms work together to creatively devise

ways to accomplish their parts in the value chain more

efficiently and effectively than before. In this way, the

chain of companies is able to create long term, strategic

advantages over other value chains for competing products.

Such advantages are particularly long-lasting relative to

advantages created through product differentiation, brand

loyalty, or technological superiority, because they are

created in the context of a dyadic exchange relatior lip that

is inherently difficult for competitors to observe a i


In this chapter, the need for better understanding how

strategic advantages are created in the value chain for a

product is articulated. The chapter begins with an example

of a value chain that has managed to create strategic

advantages through its buyer-supplier relationships. The

discussion then turns to needed research on the development

of strategic advantages in value chain relationships, and

culminates in a statement of the purpose and scope of this

dissertation in advancing our understanding of this critical


Strategic Advantage Through Buver-Supplier Relationships

Consider the following example. Relationships between

the 3ose Corporation, a maker of high-fidelity systems, and

its suppliers are so close that their suppliers are

practically treated like Bose employees (Bleakley 1995).

Supplier representatives have desks near the factory floor.

Their badges let them roam where they choose, attend

production-status meetings, stop by the research lab, and

click sales forecasts into computers. The;' write sales

orders and pass the bills to Bose. The payoffs for Bose

include reduced inventories, elimination of redundant

purchasing agents, and gains in cost-saving tips as a result

of their suppliers' familiarity with the company. In return,

the suppliers are able to maximize their present sales and

potential future sales from Bose. These suppliers are also

expected to treat their own vendors in a similar way. As a

result, the companies comprise a supply chain that cuts costs

all along the value-added chain for their products.

By fabricating their supplier relationships in this way,

Bose andL its suppliers are able to achieve cost, performance,

and learning advantages over competing buyer-supplier dyads.

Essentially, the firms have expanded the "size of the pie"--

tpe pie of potential profits available to them--by increasing

sales and/or decreasing costs and have achieved results of

greater consequence than each would have been able to earn

individually. The resulting competitive advantages are

rooted in the combination of firm-specific resources and

competencies that are brought to bear on the relationship in

conjunction with "causal ambiguity" (Reed & DeFillipi 1990).

Causal ambiguity refers to the "basic ambiguity concerning

the nature of the causal connections between actions and

results" (Lippman & Rumelt 1982). This ambiguity surrounding

the buyer-supplier relationship involves tacit, complex, and

specific information of the two firms. Since it is

unobservable by competing dyads, the resulting competitive

advantages accruing to the dyad are sustainable over a long

period of time.

It is important to note the long term nature of such

close supplier relationships. In order for firms to engage

in the strategic planning that leads to sustainable

competitive advantages, the members must have a history of

shared exchange. This also helps to reduce uncertainty about

the other's intentions and motives and provides opportunity

for learning between them. The members need to be able to

learn and react to successive actions (or lack of actions) of

the other member over time so that a specific history is

built within the dyad (Hinde 1979). This history provides

the members with information that is detailed, trustworthy,

rich, accurate, and cheap; it also asserts a strong influence

on relationship outcomes, individual perceptions,

evaluations, future expectations and present behavior (Anand

& Stern 1985; Granovetter 1985). Hence, the focus of this

study is on long term relationship dynamics between buyers

and suppliers in ongoing relationships.

Clearly, relationships like the Bose e:-:ample are not

always successful. The interdependencies formed between the

companies represent both the bases for the development of

strategic advantages and the opportunities to exploit each

other. And this does commonly happen (Bleakley 1995).

However, the research to date on interorganizational exchange

provides little insight on when such close relationships do

or do not work. One of the goals of this study is to provide

insights into the factors that make such relationships work.

Needed Research on Strategic Advantage

Despite the possibility of attaining strategic

advantages through buyer-supplier relationships there is

surprisingly little systematic work on how buyers and

suppliers develop and maintain strategic advantages together.

The primary contributors to date have been consulting


In marketing, much of the work on this topic has focused

on characteristics of satisfactory relationships (i.e., J. C.

Anderson & Narus 1990; E. Anderson & Weitz 1989; Crosby,

Evans, & Cowles 1990; Heide & John 1990), without explicitly

focusing on the attainment of strategic advantage. In the

past five years alone we have witnessed an outpouring of

studies on topics such as trust (Andaleeb 1992; Ganesan 1994;

Moorman, Zaltman, & Deshpande 1992, Moorman, Deshpande, &

Zaltman 1993), performance (Kumar, Stern, & Achrol 1992;

Noordewier, John, & Nevin 1990), commitment (E. Anderson &

Weitz 1992; Morgan & Hunt 1994), and communication/-nfluence

(Mohr & Nevin 1990; Boyle, Dwyer, Robicheaux, & Simpson 1992;

Ganesan 1993; Scheer & Stern 1992), as key aspects of strong

relationships. The understanding of how strategic advantage

is developed between independent firms in the value chain

represents an opportunity for marketing scholars to lead or

at least participate in the evolution of this new form of

interorganizational exchange.

In addition to understanding how strategic advantage is

developed between buyers and suppliers, there is a need to

understand interorganizational relationship dynamics over

time. Despite the constant calls for longitudinal studies

and the recognition that relationships have their own life-

cycles and phases (cf., Dwyer, Schurr, & Oh 1987), no one has

ever attempted to examine a sample of ongoing buyer-supplier

relationships at two points in time. As a result, there are

unanswered questions about the correct causal ordering of

variables. For example, a commonly debated issue regards the

relationship between trust and idiosyncratic assets. Is it

the case that the development of trust leads firms to make

irretrievable asset investments into the relationship or does

the presence of these assets create greater trust? Because

of the cross sectional nature of industrial buying research,

proponents of each view are able to point to empirical

studies that support both sides of the question.

The purpose of this dissertation is to address the gap

in our understanding of the development of strategic

advantage in long term, buyer-supplier relationships. As a

first step, a conceptual framework involving potential

antecedents and consequences of developing strategic

advantages is presented. Although clearly not exhaustive in

its consideration of all potential antecedents and

consequences, the framework highlights key constructs

identified from past work in industrial organization,

economics, marketing, law, and social psychology literatures.

This framework is then examined over a one year time period

among a sample of approximately 200 industrial buyer-supplier

dyads to gain insights into the long term dynamics of these


The unit of analysis throughout this study is the dyad,

comprised of a buyer and a supplier. The focus is on long

term, vertical relationships between financially independent

firms near the beginning of the value added chain for a

product (e.g., the raw materials supplier and the

manufacturing buyer). These firms do not rely on or create

unified financial systems or other forms of bureaucratic

control to oversee the relationship. Hence, joint ventures,

horizontal relationships, and vertically integrated

relationships are not considered in this study. Joint

ventures occur when two firms come together and create a

separate firm able to take advantage of an immediate (and

sometimes temporary) competitive advantage, such as

technology transfer, access to markets, etc. Such

relationships are usually terminated when either partner is

able to pursue the opportunity alone. Horizontal

relationships between competitors are likely to involve

different dynamics than those between vertically related

independence firms. On the other hand, vertically integrated

relationships rely on a common, bureaucratic mechanism to

oversee the relationship.

Organization of the Dissertation

The dissertation is organized into six chapters. A

brief description of each chapter follows.

Chapter 1 notes the importance of relationships in the

value added chain in providing sustainable competitive

advantages to members. As seen in this chapter, there is an

opportunity for marketers to lead the way in understanding

how strategic advantage is achieved in long term, vertical

relationships between independent firms in a value added

chain. There is also a need for understanding these

relationship dynamics over time and their implications for

the long-run future of the relationship.

Chapter 2 is a review of the research applicable to long

term, interorganizational relationships. Three major

paradigms commonly used in marketing are examined and their

contributions and shortcomings in understanding long term,

buyer-supplier relationship dynamics are outlined. Key

constructs from each of these paradigms are reviewed in

greater detail: idiosyncratic assets, trust, and

environmental uncertainty. This discussion provides the

backdrop upon which the conceptual framework of the study is


Chapter 3 describes the conceptual framework of the

dissertation. In this chapter, aspects of partner firm

compatibility, environmental uncertainty, and interpersonal

factors are defined and reviewed. Their relation as

antecedents to the firms' willingness to engage in strategic

advantage development activities is articulated. The

synergistic consequences examined in this study are also

highlighted. Finally, the longitudinal model is developed

and hypotheses are presented.

Chapter 4 describes the methodology used in the

dissertation. Issues concerning measurement of the phenomena

of interest underly the decision to use a multiple key

informant approach. The data collection procedure, sample

characteristics, and information on the research setting are

specified in detail. The questionnaires and measure

development process are described, and preliminary

measurement model results from the first wave of data

collection are presented in this chapter.

Chapter 5 presents the statistical analysis strategy and

results in detail. An incremental model building approach is

taken, in which dyadic, cross sectional, and longitudinal

measurement models are first estimated. This is followed by

estimation of the causal structural 7.ode1s, beginning at the

cross sectional level and then the longitudinal level.

Alternative, saturated models are also explored in

determining the adequacy of the hypothesized theoretical



Chapter 6 provides a discussion of the results and a

summary of the significance of the dissertation. In this

chapter, limitations of the study, recommendations for future

research, and implications for theory and management are



Research on long term, interfirm relationships in

marketing has predominantly drawn upon three theoretical

frameworks: transaction cost analysis (Coase 1937;

Williamson 1975, 1981, 1985), market/relational exchange

(Macauley 1963; Macneil 1980), and the political economy

paradigm (Benson 1975; Stern & Reve 1980). In this chapter

the basic tenets of each approach are reviewed and their

limitations in understanding long term, interorganizational

relationship dynamics are noted. A central construct from

each paradigm is highlighted in greater depth: idiosyncratic

assets, trust, and environmental uncertainty. These

constructs provide the fundamental building blocks for the

conceptual framework presented in the next chapter.

Transaction Costs Analysis

Theoretical Overview

Transaction cost analysis (TCA) traces its roots to the

writings of Commons(1934) and Coase (1937), with Williamson

(1975, 1981, 1985) leading the development of this stream of

thought within the past twenty years. Although there have

been a number of scholars who have delineated the

shortcomings of TCA (e.g., Bradach & Eccles 1989; Granovetter

1985; Oberschall & Leifer 1986; Perrow 1981, 1986; Powell

1990; Robins 1987; Williamson 1991), the framework has

endured over the years in the economics literature and has

provided the conceptual basis for a number of channel

research studies (i.e., E. Anderson 1985; E. Anderson & Weitz

1986; Dwyer & Oh 1988; Heide & John 1988, 1990; John 1984;

John & Weitz 1989).

Williamson draws on literature in economics,

organizational theory, and contract law to formulate the TCA

framework. TCA describes the organization of economic

activity as a decision between market exchange and vertical

integration. Market exchanges are one-time, spot

transactions based on the efficiencies of open, competitive

markets. Buyers use price as a primary criterion for their

purchases and multiple competing suppliers to insure that

goods and services are purchased at the lowest costs.

Vertical integration occurs when the firm internalizes the

market transaction and imposes a bureaucratic control

mechanism to oversee the exchange.

The decision to engage in market exchange or vertical

integration depends on the transaction costs associated with

each option. These costs vary as a function of environmental

uncertainty, asset specificity, and transaction frequency.

When any of these dimensions are high, it becomes more

efficient for a firm to expand its boundaries and internalize

the transactions and resource flows that previously occurred

in the marketplace. Market transactions are appropriate when

the environment is stable and the products are standardized

(not specialized through transaction specific investments for

a specific firm).

Li i c-- n s -- f TCA

The TCA logic is well suited for the explanation of some

forms of interorganizational exchange strategy. However, it

is very limited in its ability to explain long term forms of

intercrganizational exchange for a number of different

reasons. First, TCA provides an explanation for the

occurrence of competitive market exchange and vertical

integration only and does not address alternative forms of

interorganizational exchange governance modes that may blend

aspects of markets and bureaucratic control mechanisms

(Bradach & Eccles 1989; Powell 1990; Williamson 1991; Zajac &

Olsen 1993).

In today's business environment, firms have developed a

wide variety of governance options that are neither markets

nor bureaucratic control forms. Consolidation and downsizing

are the dominant trends in a number of industries, resulting

in fewer and larger product suppliers. The increased need

for specialized products and the reduction in the number of

alternative suppliers make it increasingly attractive for

buyers to engage in noncompetitive market exchanges.

There is also a growing disenchantment with vertical

integration. Firms have moved away from the acquisition and

management of unrelated firms and refocused their attentions

on their core business, recognizing that greater financial

returns can be gained by exploiting unique sources of

competitive advantage. Collectively, these trends point to a

movement away from a TCA-based view of exchange to

alternative exchange forms between independent firms governed

by contracts (i.e., franchising) and/or relational norms

(i.e., trust) (Weitz & Jap 1995).

Second, TCA emphasizes a single-party, cost-minimization

view of exchange and neglects the interests of the other firm

(Zajac & Olsen 1993). Many interfirm exchanges involve

multiple, voluntary transactions between two firms over a

long time frame. These firms are seeking to create and

sustain an exchange relationship that is mutually beneficial

to both firms while simultaneously seeking to satisfy their

own individual interests. Although this does not necessarily

mean that firms will act altruistically toward their e:xchanae

partners, the implication is that the firms do not want the

relationship to be terminated prematurely due to one

partner's dissatisfaction with the relationship. Hence, each

firm's actions are not strictly a function of its transaction

costs alone.

Third, TCA emphasizes structure while neglecting

processes in interorganizational exchange (Robins 1987; Zajac

& Olsen 1993). More specifically, the notion of asset

specificity, or highly specialized assets that lose

productive value outside the relationship, is analogous to

the notion of exit barriers (cf., Caves, 1987) in the

industry structure literature. In the latter case, exit

barriers are a market structure that influences market

conduct and performance. With TCA, specific assets form the

transaction's structure, which in turns influences the firm's

conduct and performance in the exchange relationship. In

both cases, these structure characteristics create

transformation rcrcesses that fundamentally change ultimate

performance. The developmental processes that bring about

changes in interorganizational exchange performance are never

articulated by TCA.

C.A also ignores the potential influence of social

relationships in interorganizational exchange (Bradach &

Eccles 1989; Granovetter 1985; Maitland, Bryson, & Van de Ven

1985; Perrow 1986). TCA assumes that humans have limited

information processing capability and limited ability to

formulate and solve complex problems. They are motivated to

behave opportunistically. Hence, vertical integration is

suggested as a mean by which to guard against human

limitations, because the bureaucratic governance structure

provides greater control over employee actions, motivation,

and incentives. However, the establishment of bureaucratic

control does not necessarily produce trust. Personal

relations, social norms, and the resulting obligations

inherent in them can often discourage opportunism on a

consistent basis, apart from institutional arrangements.

Information from such relationships is a major determinant of

behavior because such information is often cheap, richly

detailed, trustworthy, and overlaid with strong social

expectations of trustworthiness, future interactions, and

abstention from opportunism.

Relational Exchange

Theoretical Overview

Macauley's (1963) seminal article on the use of legal

contracts in business relationships points to the importance

of social and relational norms as well as the impact of trust

and commitment as means of governing exchange. Building upon

this notion, Macneil (1978, 1980) draws on modern contract

law relationships to suggest that the nature of the

interpersonal relationship surrounding a contract is

paramount. This relationship provides important social norms

that aid in the governance of contractual behavior, provides

a reference for dispute resolution, and a setting that

fosters a desire for relationship continuance. This is

somewhat analogous to Ouchi's (1979) concept of a clan

mechanism. In a clan, a normative process occurs in which

members adopt the norms of the larger system through

socialization efforts. Hence, deviance or opportunism is

dealt with through self-control based on internalized values.

Macneil distinguishes between discrete and relational

transactions. Discrete transactions are market exchanges,

characterized by limited communications and narrow

informational content focused on the nature of the -xchange.

Discrete relationships account for only a limited number of

relationships involving specialized functions within an

economy. Relational transactions, the predominant

relationship type, occur over time and are viewed in terms of

a history and anticipated future. Macneil highlights

differences between the two types of exchange in terms of the

timing of the exchange, the number of parties involved,

obligations, personal relations, contractual solidarity,

transferability, cooperation, measurement and specificity,

power, and division of benefits and burdens.

Although Macneil did not provide operational definitions

for these dimensions, researchers have made their own

interpretations. In marketing, there are a number of

empirical studies that have applied Macneil's relational

contracting concepts to an interorganizational context (i.e.,

Dwyer, Schurr, & Oh 1987; Frazier, Spekman, & O'Neal 1988;

Kaufmann & Dant 1992; Kaufmann & Stern 1988; Noordewier,

John, & Nevin 1990; Palay 1984). Relational exchanges have

been compared to bilateral power systems (Bonoma 1976), in

that individual goals are reached through joint

accomplishments, and concern for the long-run benefit of the

system restrains individual tendencies to pursue self-

interests opportunistically., Relational exchanges offer

advantages associated with both a market exchange and a

vertically integrated arrangement: scale economies obtained

from dealing with independent specialists and cross-

functional coordination.

Limitations of Relational Exchange

Relational contracting theory is useful in accounting

for alternative forms of interorganizational exchange apart

from market transactions and vertical integration. However,

the relational paradigm, in contrast to TCA, fails to specify

the conditions under which these alternative forms may occur

or when they are appropriate, despite the fact that it is

implicitly based on a recognized need to adapt relationships

to changing circumstances (Heide 1994). Macneil (1980) fails

to make explicit predictions in this respect, instead making

only the general assumption that bilateral elements are

required in order for parties to "project their exchange into

the future."

Foliti:7l Econom',' 7aradirim i

Theoretical Overview

The political economy paradigm (Benson 1975; Stern &

Reve 1980; Wamsley & Zald 1973, 1976; Zald 1970a; Zald 1970b)

was introduced during the 17th century to denote what is now

known as economics. Over time, the paradigm has taken on

different meanings, resulting in the development of

literature streams such as radical political economy, the

Chicago school of economics, etc., and has been shown to be

useful in the analysis of individual organizations (Zald

1970b) and interorganizational networks (Benson 1975). The

paradigm integrates concepts from social exchange, the

behavioral theory of the firm, and transaction cost


Stern and Reve (1980) were the first marketing

researchers to extend the political economy framework into

the context of channel dyads. The social system is viewed as

the resulting interaction of internal and external political

economies and is thought to affect behavior and performance

of firms in the system. The dyad's internal political

economy consists of an economic structure that links the

members, processes by which terms of trade are established, a

sociopolitical structure that establishes a pattern of power-

dependence relations, and processes that determine dominant

member sentiments (e.g., cooperation, conflict). The

external political economy includes factors such as the

current economy (i.e., the nature of the industry's vertical

and horizontal markets, the prevailing and prospective

environment), and the external sociopolitical system.

The heart of the political economy approach rests in

simultaneous analysis of the power and control system of the

channel and productive aspects that transform inputs to

outputs with an emphasis on interdependencies. Centralized

planning processes increase efficiency and effectiveness, but

also create interdependencies that prevent the dyad from

reacting quickly to changes in the environment. When power

relations between the members are balanced, cooperative

behavior will dominate (cf., Kaplan 1957); the more the

relationship is characterized by cooperative behavior, the

greater the profits attainable to the firms as a whole.

Subsequent investigations of the political economy

framework in marketing channels have focused on the impact of

environmental variables such as environment type (Achrol,

Reve, & Stern 1983), uncertainty and dependence constraints

(Dwyer & Welsh 1985), and determinants of decision-making

uncertainty (Achrol & Stern 1988) on interorganizational

responses. Mohr and Nevin (1990) provide a theoretical model

of communication in which an individual member's influence

strategy is moderated by the impact of channel conditions

(e.g., structure, climate, and power) on channel outcomes

(e.g., coordination, satisfaction, commitment, and


Limitations of the Political Economy Paradigm

Approaches such as the political economy paradigm and

Frazier's (1983) interorganizational exchange framework are

commendable in their attempts at completeness and their focus

on the impact of environmental factors and processes.

Additional strengths of the paradigm include its attention

toward authority and control patterns, conflict and conflict

management procedures, and internal and external determinants

of institutional change (Arndt 1983). On the downside, the

political economy paradigm is overly general and basic. It

serves primarily as a method for classifying constructs

rather than developing theories for the relationships between

them. Additionally, several of the constructs and

relationships are difficult to tap with standard measurement

tools such as cross sectional surveys. Arndt (1983) notes

that the approach lacks an emphasis on channel performance or

goal achievement with respect to efficiency and

effectiveness; this is also evidenced by Reve's (1980)

failure to find a significant relationship between channel

structure and performance.

Key Contributions of the Frameworks

TCA and the relational exchange paradigms are valuable

in that they elucidate potential governance modes for

interorganizational exchange. The political economy approach

draws attention to the need to understand the whole channel

social system when investigating relationship dynamics.

However, all of the frameworks leave unanswered the question

of how the interfirm relationship can be used to "expand the

size of the pie" of potential benefits available to the

members and gain strategic advantages over competing dyads.

As a first step in answering this question, the conceptual

framework for this study uses three key constructs

highlighted by TCA, relational, and political economy

frameworks respectively: idiosyncratic assets, trust, and

environmental uncertainty.

Idiosyncratic Assets

Asset specificity is a key construct in the TCA

framework, increasing the potential cost of transacting

between firms. Although it appears detrimental in the TCA

framework, idiosyncratic assets can be used to promote two

important aspects of interorganizational exchange: a basis

for strategic advantage and relationship stability.

Strategic advantage. The development of idiosyncratic

assets can enable the parties in a relationship to develop

sustainable strategic advantages over competing

relationships. These idiosyncratic assets are unique and

cannot be easily observed or duplicated by competitors. For

example, when Intel designs and produces a unique

microprocessor that optimizes the performance of a new Ford

automobile, the Ford automobile is able to provide superior

value compared to a General Motors (GM) car using a standard

microprocessor available to all automobile manufacturers.

Both Ford and Intel benefit from the investment in this

idiosyncratic microprocessor. Ford sells more cars and Intel

sells the microprocessors used in the cars. To match the

automobile performance achieved through the Ford-Intel

relationship, General Motors must build a similar

relationship with Intel or another microprocessor supplier to

develop a microprocessor that optimizes its car design.

Building such a relationship takes considerable time and

effort. Thus, the competitive advantage arising from the

Ford-Intel relationship is not easily duplicated by

competitive dyads.

Relationship stability. Idiosyncratic assets promote

relationship stability because they provide meaningful

signals to the other party that go well beyond verbal

promises to work together over the long run. By creating

idiosyncratic assets, the parties in the relationship will

suffer adverse economic consequences in the event that the

relationship ends. Thus, they act as credible signs of

commitment to the relationship. E. Anderson and Weitz

(1992) provide evidence that idiosyncratic assets are

strongly associated with manufacturer and distributor

commitment, as well as perceptions of other-party commitment.

Idiosyncratic assets also increase cooperative

activities because they provide incentives to maintain the

relationship (Williamson 1985). Since the assets lose value

outside the relationship, firms are unlikely to engage in

opportunistic behavior that might risk relationship

dissolution. Idiosyncratic assets also facilitate

expectations of relationship continuity and increase

willingness to engage in joint activities (Heide & John

1990). Thus, these assets help to align the members'

incentive structures and bring about relationship-stabilizing



Much of the work on inter.rarnizational exchange

dynamics points to the importance of trust in developing and

sustaining long term relationships (i.e., Andaleeb 1992;

Bradach & Eccles 1989; Moorman, Zaltman, & Deshpande 1992,

1993; Morgan & Hunt 1994). There is a vast literature on the

role of trust in interpersonal relationships (i.e., Boon

1994; Boon & Holmes 1991; Holmes 1991; Kelley & Thibaut 1978;

Lewis & Weigert 1985a, 1985b; Luhman 1988; Silver 1989).

Trust is widely acknowledged as a key social norm in

governing and coordinating relational :-:*changes. It is

particularly important in the development of strategic

advantage in interfirm relationships because developing

strategic advantages is a risky endeavor. Often, the parties

are unable to specify in advance what the payoffs will be

from working closely together. Reve (1980) reports that the

development of joint programs is positively related to close

interpersonal relationships. This result is echoed in the

interpersonal relationship literature, where studies have

shown that as a relationship develops, feelings of mutual

responsibility for the other party's outcome increase, and

parties explore mutually beneficial activities (Altman &

Taylor 19"3; Levinger 1980; Levinger s Snoek 1972; Taylor

1979). Hence, in strategic relationships, trust provides a

powerful assurance to the individual members that they will

receive their fair share of any resulting benefits.

Trust is a complex construct that affects the

relationship in a variety of ways, full discussion of which

is be':.ond the scope of this review. The literature on trust

is vast and arguments over its precise definition are never

-.din:. In light of this, suffice it to say that there are

essentially three dominant views of trust: trust as reliable

behavior (cf., Crosby, E.'ans, & Cowles 1990; Mohr & Spekman

1991), trust as motivation (cf., Gabarro 1987; Golembiewski &

McConkie 1975), and trust as both reliable behavior and

motivation (Andaleeb 1992; J. C. Anderson & Narus 1990;

Deutsch 1958; Ganesan 1993; Lindskold 1978; Zucker 1986). In

this study, trust is defined as the ability to reliably

predict the actions of the other party in the relationship

and the belief that the other partner will not act

opportunistically if given the chance to do so.

This particular view of trust was chosen because prior

field interviews with buyers in the marketplace indicated

that their views of trust involve both aspects.

Conceptualizations that view trust as solely past behavioral

reliability or perceived motivation governing future behavior

are mpoo'.erished in a buyer-supplier setting. These buyers

contend that trust involves both the ability to accurately

predict the other party's future behavior in a variety of

contexts based on explicit past behaviors and an implicit

understanding of the other party's motivations for future

behavior. Hence, trust is viewed as an interpersonal, rather

than an interorganizational construct, and "the other party"

refers to the focal boundary-spanning employee working for

the other firm in the relationship.

Trust as reliable behavior. One critical aspect of

trust is the belief that the partner's word is reliable and

that he/she will fulfill an exchange obligation (Giffin 1967;

Schurr & Ozanne 1985). In order to forecast party A's future

actions effectively, party B must rely on a knowledge of

party A's consistency of responses in the past. Party B's

willingness to engage in trusting responses or actions will

be made on the basis of these estimates; as long as the

relationship's history remains positive, there exists a

preference to interact with current, predictable partners

rather than new or unfamiliar ones (Kelley 1983).

Trust as motivation. The second critical aspect of

trust is the belief that a partner will act non-

opportunistically if given the chance to do so (Bradach &

Eccles 1989; Larzalere & Huston 1980; Rotter 1967). while e

the first aspect of trust refers to dependability in

fulfilling explicit promises, this second aspect relates to

the anticipated actions of the other party when novel

situations arise.

As the dyad deals with problems, members infer each

other's motives by their actions and responses as the

relationship progresses through various stages. As

attitudes, backgrounds, and motivations increase in

similarity, the relationship becomes more satisfying, and

expectations of the other par--t facilitating one's goals are

increased (Ross & Ferris 1981; Weiss 1978; '.exiey, Alexander,

Jreenawalt, & Couch 1980). Thus, the parties anticipate that

they will consider each other's best interest, even when

explicit promises are not made.

Trust is a key aspect in relationships centered around

the attainment of strategic advantage. Jap and Weitz (1994)

show that such relationships have significantly higher levels

of trust, idiosyncratic assets, commitment, dependence, time

orientation, performance, goal compatibility, strategic

development activities, confidential information exchange,

and risk-taking compared to alternative relationship types.

Environmental Uncertainty

One important characteristic of the political economy

paradigm is the focus on the external environment surrounding

the dyad. This is likely to be an important factor in

deciding whether to work closely with a buyer or supplier,

because when the environment is uncertain, decision-makina

can be particularly difficult and complex. The work on

environmental effects on channel dyads is sparse when

compared to the organizational behavior literature. In that

literature, organizations will react to environmental

conditions in an effort to protect themselves from external

forces and cope with environmental demand (Thompson 1967;

Zajac & Shortell 1989).

Although there has been a long debate on how to best

measure environmental forces (Boyd, Dess, & Rasheed 1993;

Downey, Hellriegel, & Slocum 1975; Lawrence & Lorsch 1967;

Snyder & Glueck 1982; Tosi, Aldag, & Storey 1973), there is

some consensus on the three central dimensions of

environmental uncertainty (Dess & Beard 1984): munificence,

dynamism, and complexity.

Munificence. Munificence is the extent to which the

environment can support sustained growth in terms of

absorbing and providing the dyad with necessary resource

opportunities (Starbuck 1976). In the channels literature,

this has come to refer to demand uncertainty, or the extent

to which the demand for channel products is unstable and

difficult to predict over time (Achrol, Reve, & Stern 1983;

Etgar 1976; Guiltinan 1974; Lawrence & Lorsch 1967; Weick


There have been a number of authors who point to the

importance of demand uncertainty on organizational behavior

(e.g., Aldrich & Pfeffer 1976; Burns & Stalker 1961; Emery &

Trist 1965; Lawrence & Lorsch 1967; Pfeffer & Salancik 1978;

Terreberry 1968; Thompson 1967). The general consensus of

these studies is that as demand becomes more uncertain and

difficult to predict, organizational systems become

decentralized, with a focus on lateral interactions and

communication networks. "hen the environment is munificent--

there exists high potential for economic growth, strong

consumer demand, etc.--then there is less environmental

uncertainty, and hence, less incentive for vertical

coordination (Achrol & Stern 1988; cf., .chrol, Reve, & Stern

1983; :.wyer & Welsh 1985).

Dynamism. Dynamism refers to change that occurs

frequently and is difficult to predict. When the external

environment is extremely dynamic (i.e., marketing practices,

competitor strategies, and consumer preferences are

constantly changing), then it is difficult for member firms

to Tmaage lrng-range business planning, coordination,

product-mix, and inventory decisions. For example, when the

partnering dyad faces a number of *competitors that vary in

size and resources, there exists a great deal of uncertainty

in predicting behavior of competing dyads.

Complexity. Complexity involves both the heterogeneity

of external forces affecting resources and their levels of

interconnectedness and concentration in the environment. For

example, a heterogeneous environment might occur if consumers

are dissimilar in terms of social background, needs, and

preferences. In this case, the formulation of strategic

programs and marketing responses becomes more difficult

(Dwyer & Welsh 1985). Alternatively, a competitive field

that is extremely interconnected or concentrated (i.e., a

high level of competitor domination) increases environmental

complexity and the need for strategic activities. When

hetercgeneity is high, member firms tend to increase

cooperative behavior and develop closer linkages with their

partners (Achrol et al. 1983). Additionally, Etgar (1976)

has found that as the number of competitive dyads that carry

similar or competing brands increases, channel dyads are more

likely to work closely together as a unit.

In the face of environmental uncertainty, or in other

words, low munificence, high dynamism, and high complexity,

dyads that mutually work together (by being flexible, willing

to adapt, and share information) are in a better position to

cope with environmental uncertainty than dyads in which both

members individually attempt to deal with the environment.1

Such partnerships enable the partnering firms to adjust their

strategies quickly, thereby improving dyadic performance and

reducing opportunism (Achrol et al. 1983; Heide & John 1990).


In this chapter, three paradigms for understanding

interorganizational exchange have been presented.

Transaction cost analysis was developed as an explanation for

a firm's decision to engage in market e-::chnr.ge or vertical

integration. Relational exchange is useful in understanding

intermediate forms of long term exchange between autonomous

companies. Both of these frameworks emphasize governance

modes: competitive markets, bureaucracies, and social norms.

1 Although one might make the argument that close
relationships result in decreased flexibility and should thus
be avoided in the face of environmental uncertainty, the
literature has consistently shown the opposite to be true.
An explanatory moderating variable for predicting when firms
should and should not work closely is presented in the next

In contrast to these, the political economy paradigm calls

for understanding of the entire channel social system

surrounding the dyad.

All three frameworks fail to address the issue of how

interorganizational relationships can be used as a basis for

the attainment of strategic advantages over competing dyads.

However, each provides key constructs that are useful

starting points: idiosyncratic assets, trust, and

environmental uncertainty. Idiosyncratic assets enable the

attainment of strategic advantage and facilitate relationship

stability. Trust is key in providing assurance of fair

payoffs from working together. Finally, environmental

uncertainty leads member firms to work closely together in

order to cope with environmental demands.



In this chapter, the conceptual framework for the

dissertation is advanced. Characteristics of the partner

firm, environmental uncertainty, and interpersonal

characteristics are hypothesized to lead firms to work

closely together to create strategic advantages that "expand

the size of the pie" between them. The development of

strategic advantages results in synergistic outcomes--the

attainment of sustainable advantages over competing dyads and

subsequent increased joint profits--and leads to the creation

of idiosyncratic assets. An overview of the basic strategic

advantage process is depicted in Figure 3-1. The chapter

concludes with an elaboration of this basic model over time.

Developing Strategic Advantages

The central goals of this study are to identify the

factors that lead independent firms in long term, value chain

relationships to create strategic advantages together and the

payoffs from doing so. Hence, the central construct in the

framework is the active development of strategic advantages.

This behavior is formally defined as, "the extent to which

the 5'ad engages in activities aimed at achieving strategic

advantages" (Jap & Weitz 1994). Another way to think of this

construct is to essentially view it as efforts to expand the

pie, via the creation of strategic ad-.anrages over competing

dyads in the marketplace.

Because of the explicit focus on the attainment of

strategic adcr.anrages, this construct is not equivalent to

co:e> ration or joint action, concepts that have appeared in

past work on buyer-supplier relationships. Both of these

constructs merely describe the nature of the relationship,

not the objective to develop strategic advantage.

Cooperation is basic to any successful exchange, the

development of strategic advantages is not. Even in one-

time, spot transactions, the buyer and seller must cooperate

in terms of providing the goods and services as specified and

rendering fair payment. This willingness to work together is

a necessary but not sufficient condition for the attainment

strategic advantages.

Joint action refers to the degree of interpenetration of

organizational boundaries (Guetzkow 1966; Heide & John 1990;

Laumann, Galaskiewicz, & Marsden 1978). This occurs when the

buyer and supplier become involved in activities that

typically are the other member's responsibility. Mere

observance of this behavior also ioes not necessarily mean

that the dyad is focused on the attainment of strategic

advantage. Hence, the strategic advantage development

construct in this framework differs from other similar

constructs in the literature in that the behavior occurs with

the motivation of jointly achieving strategic advantages.

Developing strategic advantages together creates

interdependence between the parties involved. This

interdependence can be beneficial, in that both parties may

be able to achieve synergistic outcomes that would not have

been achievable in the absence of the partner. Consider the

case of two firms with complementary skills. These firms are

considering the possibility of working together to bring a

new product to market. A number of competitive firms in

their industry are also working on the new product. Everyone

realizes that there are substantial benefits that will accrue

to the firm who is the first to introduce this new product.

Suppose these two firms realize that by developing a new

manufacturing process together they can dramatically reduce

their lead time to market. The development of this new

process will require considerable time, effort, and energy.

The firms will make the necessary investments that bind their

fates together if they perceive that the joint value to be

gained from the relationship will outweigh the benefits that

each firm would have been able to earn independently.

However, the preceding situation also poses

opportunities for greater losses as the stakes involved are

greater than they would be in the absence of the partner.

The process of developing strategic advantages creates an

interdependence that results in a loss of individual autonomy

and raises new uncertainties because of the necessary

reliance on the other party. As such, the decision to become

interdependent so as to attain strategic advantages is likely

to be driven by the percei-.-d rewards and costs, or

expectations of rewards and costs from engaging in such


..tr.ecedent= -j -.'.'e.1:=inc 3trat- qic .j'.'.Jnraq es

The decision to develop strategic advantages is often

shrouded in uncertainty. Firms may experience doubts as to

whether the other firm is the best firm for their needs and

whether the payoffs from becoming interdependent are

attainable. Hence, there are advantages and disadvantages

inherent in their choice of partner and it is not always

clear what the payoff will be. In interpersonal

relationships, coping with uncertainty is a critical

challenge in relationships and a major motivational force in

shaping the individuals' mental representations of their

partners (cf., Brehm 1988; Brickman 1987; Holmes & Rempel

1989; Murray & Holmes 1993).

The same is true for the value chain members who are

contemplating developing strategic advantages together.

These firms can gain insight into the specific nature of the

costs and benefits associated with the decision to become

interdependent by evaluating various antecedent

characteristics of the relationship such as technological

capabilities, organizational changes in systems and skills,

commitment, financial strength, consistent performance, close

communication, etc. (Jap 1992).

In this study, three classes of antecedent variables are

examined, based on past work in the channels, organizational

behavior, economics, and social psychology literatures and

interviews with channel members: partner firm

characteristics, environmental characteristics, and

interpersonal characteristics. These antecedents are

illustrated in Figure 3-2. The groups were chosen primarily

because they represent different aspects of the channel

relationship--the interfirm aspect, the influences of the

external environment, and the social aspect--that are likely

to impinge upon the dyad's decision to develop strategic

advantages together. Although clearly not exhaustive, the

specific antecedents selected for this study represent a

first step in understanding how strategic relationships can

result in synergistic outcomes for the W".Tad.

Partner Firm Characteristics

As channel members consider whether to develop strategic

advantages together, one of their main concerns is whether

they will realize a fair return from becc.rTi-. interdependent.

Two aspects of this uncertainty involve the synergistic

benefits and returns to the dyad (the degree to which the pie

will be expanded) and the returns to the individual firms

(the fairness in splitting the pie). As such, members will

look for key characteristics of the partnering firm that will

help to reduce their uncertainty about their returns from the


There is a paucity of empirical and conceptual work on

selecting an appropriate partner to develop strategic

advantages with. However, past research on supplier

selection and interviews with purchasing agents indicate that

there are two important factors that need to be taken into

consideration when deciding whether to work closely to

develop strategic advantages with a supplier: the degree to

which the partnering firm has compatible strategic goals and

its ability to provide complementary competencies or

resources that will enhance the likelihood of goal


Goal compatibility. Goal compatibility refers to the

extent to which firms perceive the possibility of

simultaneous goal accomplishment (Eliashberg & Michie 1984;

John & Reve 1982; Schmidt & Kochan 1977). Firms are more

likely to engage in the development of strategic advantages

when they have similar goals because similar goals will

insure that they will pursue similar directions and increase

the probability of successful pie expansion activities. In

the earlier example of two firms working to de.elop a

manufacturing process, the firms have a common goal of

reducing their lead time to market. This goal also provides

an assurance that the other firm is not likely to pursue

goals which are advantageous to its competitive position but

not the competitive position of its partner.

As goals between member firms become increasingly

aligned, there is a strong incentive to form a close

relationship so that members can exploit joint potential and

effectively safeguard past investments (Heide & John 1988).

Research in social psychology indicates that shared goals

necessitating mutual effort increases cooperative behavior

and has a unifying effect on individual members (Blake &

Mouton 1979; Sherif 1966).

Complementary competencies. Partner selection might

also occur on the basis of distinct capabilities, knowledge,

and resources that provide for the possibility of a

complementary relationship. To "complement" another

individual means to supply another's lack or "to fill out or

complete" another's performance (Webster's Dictionary).

Hence, complementary competencies refer to the degree to

which the firms are able to 'fill out or complete' each

other's performance by supplying distinct capabilities,

knowledge, and resources that enhance the likelihood of goal

achievement (J. C. Anderson & Narus 1990). By working

closely with partners possessing complementary abilities and

resources, both firms are able to reach objectives that would

have been unattainable if they had worked alone and hence,

expand the pie of benefits available to the dyad.

Industry leaders are natural first candidates for

potential strategic partnerships because they have

demonstrated competency, and economic and performance

reliability over long periods of time. However, industry

leadership alone is neither a necessary or sufficient

condition for strategic partner selection. The critical

factor is whether the potential partner can supply missing

key competencies necessary for goal achievement, not firm

size or market power. In a strategic relationship,

complementarity in resources, assets, abilities and

interpersonal roles are likely to facilitate

interdependencies that can provide the basis for synergistic


E.'.ircnmentb Ch aracteristics

Along with concern about the magnitude of the positive

increase from working together, the members also need to

understand the potential for exploiting the unique

combination of capabilities, assets, and knowledge present

between them. The environment plays an important role in

impacting the potential rewards from activities directed

toward gaining strategic advantage. In this study, three

aspects of the environment considered are demand munificence,

dynamism. and complexity.

Demand munificence. Munificence is the extent to which

the environment can support sustained growth in terms of

absorbing and providing the dyad with necessary resource

opportunities (Starbuck 1976). In the channels literature,

munificence has typically been examined with respect to

demand (Achrol, Reve, & Stern 1983; Etgar 1976; Guiltinan

1974). When the demand for the dyad's products is high,

there is an incentive to work closely together to develop

better ways of meeting the demand over competing dyads.

Hence, demand munificence raises the potential benefits to be

gained from working closely to develop strategic advantages.

Dynamism. Dynamism describes changes in product and

competitor strategies that occurs frequently and is difficult

to predict (Achrol & Stern 1988). When the environment is

constantly changing, decision-making and coordination across

the dyad becomes increasingly difficult for members to manage

individually. Dynamism raises environmental uncertainty for

the dyad.- As such, a dynamic environment provides an

incentive to develop closer relations in order to cope with

the environment and meet the dyad's decision-making and

coordination needs.

Complexity. Complexity involves both the heterogeneity

of external forces affecting resources and their levels of

concentration in the -environment. When heterogeneity in the

environment surrounding the dyad is high, member firms tend

to increase cooperative behavior and devel-c closer linkages

with their partners (Achrol et al. 1983). Like dynamism,

complexity increases environmental uncertainty for the dyad

and provides an incentive to work closely together in order

to better cope with and manage the environment.

Partner firm and environmental characteristics are two

types of factors that in and of themselves, provide

incentives to the dyad to work closely to develop strategic

advantages together. These factors are useful to the members

in determining the potential rewards of working closely

together; they impact the potential size of the pie that can

be realized from engaging in strategic adv-antage development

activities. Essentially, these factors operate as main

effects on the likelihood that buyers and suppliers will

develop strategic advantages together. In the next section,

an important moderator of these effects is hypothesized and

the nature of the interaction is explained in greater detail.

Interpersonal Characteristics

Along with the concern over whether the pie of benefits

will expand and how the environment will affect its size,

there is also some concern over how the expanded pie will be

divided between the partners. Since the members are

typically unable to specify in advance what the e:-:panaed pie

size will be and since some benefits may only be realized in

retrospect, assurances of fair pie division can be extremely

important in determining whether the development of strategic

advantages is worthwhile. In such an ambiguous situation,

members are not able to rely on legal contracts to specify

and account for all possible contingencies. As a result, the

social or interpersonal relationship between the boundary-

spanning members of the two firms can become important in

determining whether the firms will develop strategic


Recently, there has been a growing interest in the role

of personal relationships between boundary-spanning members

in the conventional channel. Personal relationships have

been found to shape economic outcomes in a number of studies

on interorganizational exchange in a number of contexts such

as: the publishing industry (Coser, Kadushin, & Powell

1982), international joint ventures (Doz 1988; Hakansson &

Johanson 1988; Walker 1988), and small to mid-sized textile

firms in Italy (Lorenzoni & Ornati 1988). Similarly, one can

imagine that if the interpersonal relationship between

boundary-spanning members is characterized by conflict and

distrust, this could act as an impediment to potential

interorganizational outcomes.

In an inductive field study of dyadic relationships in

high-growth entrepreneurial firms, Larson (1992) found that

personal relationships shaped the context for new exchanges

between firms by reducing risks and uncertainty about the

motives and intentions of the other member. She also found

that individual and firm reputations were an important

consideration in deciding whether to develop the

interorganizational exchange relationship. Companies and

individuals saw themselves as members of an inner circle or

network within a broader industry circle. As a result,

credibility and a positive reputation--for business and

performance--were important attributes for coordinating

exchange between firms. Hence, social factors such as

personal relationships and reputations (personal trust),

coupled with a knowledge of the firm's skills and

capabilities (economic trust) were prime considerations in

interorganizational exchange.

Mutual trust. In light of this, trust is hypothesized

to play a key moderating role in determining whether firms

will develop strategic advantages. Trust is defined as the

ability to reliably predict the actions of the other party in

the relationship and the belief that the other partner will

not act opportunistically if given the chance to do so

(Andaleeb 1992; J. C. Anderson & Narus 1990; Deutsch 1958;

Ganesan 1993; Lindskold 1978; Zucker 1986).

Trust is key to the engagement in activities directed

toward the development of strategic advantage. Individuals

who trust each other are more willing to share relevant ideas

and comprehensive information (Bialeszewski & Giallourakis

1985; Moorman et al. 1992; Zand 1972), clarify goals and

problems (Zand 1972) and tend to approach the relationship

with a problem-solving orientation (Golembiewski & McConkie

1975; Zand 1972). As trust is developed between the parties,

the individuals' perceptions and expectations of relationship

continuity will increase (E. Anderson & Weitz 1989), and

members are able to communicate more efficiently than dyads

in which trust is low (Aldrich & Pfeffer 1976; E. Anderson &

Weitz 1989; Knapp 1978; Ouchi 1980). In the interpersonal

relationship literature, studies have shown that as a

relationship develops, feelings of mutual responsibility for

the other party's outcome increases, and parties explore

mutually beneficial activities (Altman & Taylor 1973;

Levinger 1980; Levinger & Snoek 1972; Taylor 1979).

Trust as a moderator of strategic advantage development.

Trust has been examined recently in channel relationship

research, and as a critical aspect of interorganizational

exchange (cf., Morgan & Hunt 1994). Academics and

practitioners collectively agree that trust plays a key role

in close, partnering relationships. This increasing interest

tends to foster an implicit belief that trust is a necessary

component of strategic relationships. In this study, trust

is not viewed as a necessary or sufficient condition for the

development of strategic advantage. Instead, it is seen as

an important facilitator of strategic advantage in that it

moderates the effects of partner firm and environmental


This is to say that interpersonal relationships between

boundary-spanning individuals can enhance or impede the

emphasis placed on developing strategic advantage. For

example, suppose a firm wants to devise new ways to take cost

out of its distribution system to another firm. The firms

perceive significant potential benefits accruing to both

firms. Once it identifies potential partners who have

similar goals and complementary competencies, the firm will

have a preference to work with companies that it has

developed a positive history of experience with, or in other

words, those companies with which the firm has trusting

relationships with because it has greater chances of gaining

a fair return for the strategic advantage development

activities. If interpersonal relationships are marked by a

great deal of suspicion, distrust, and latent conflict, the

firms will be less likely to benefit from working together,

even when the potential interorganizational benefits are

great. The nature of the interactions expected can be seen

in Figure 3-3.

HI: Trust moderates the effects of goal
compatibility on the development of strategic

H2: Trust moderates the effects of complementary
competencies on the development of strategic

H3: Trust moderates the effects of demand
munificence on the development of strategic

H4: Trust moderates the effects of environmental
dynamism on the development of strategic

H5: Trust moderates the effects of environmental
complexity on the development of strategic

Consequences of Developing Strategic Advantages

The consequences of engaging in activities to develop a

strategic relationship are the realization of strategic

advantages, higher joint profits, and the development of

idiosyncratic assets (see Figure 3-4). Strategic advantages

(i.e., superior access to resources, key technological

knowledge and development, etc.) are advantages that are

impossible to acquire alone and puts the dyad in a better

position to carry out its strategies.

The process of developing strategic advantages should

lead to the creation of idiosyncratic assets to support the

relationship. These assets may be tangible, such as

manufacturing facilities dedicated to support the

relationship or intangible, such as a newly developed

capability, skill, or technology that enables the achievement

of strategic advantages. These assets are what sustains and

supports the created synergy between the firms and provides

the dyad with a joint value that could otherwise not be

created by either firm independently.

H6: The process of developing strategic advantages
leads to the achievement of strategic
advantages over competing dyads.

H7: The process of developing strategic advantages
leads to higher joint profits.

H8: The process of developing strategic advantages
leads to idiosyncratic assets.

The Longitudinal Conceptual Model

One of the central goals of this study is to understand

the development of strategic advantages over time. It could

be that the relationship between variables posited thus far

are likely to change over the course of the relationship,

because as time passes, learning occurs, perceptions of the

environment changes, and member attitudes, goals, and

opinions are typically updated as a function of these

changes. Hence, what was appropriate at the beginning of the

relationship may become less appropriate as members'

knowledge of the other increases, and similarly, what may was

important earlier in the relationship may become less

important over time and vice-versa. If the roles of the

variables change over time, then it means that the firms must

change their approach co managing these variables over time

in order to insure the maximum joint benefit possible from

the relationship.

The main interest in the longitudinal theoretical model

is still centered around the development of strategic

advantages. The basic process model of Figure 3-5 is

replicated over time, creating the full longitudinal model of

Figure 3-6.

Intermediary Effects Between Time One and Two

As the dyad works closely together to develop strategic

advantages, they learn more about each member's needs, goals,

future intentions, and capabilities. All of this information

is used to update their perceptions, attitudes, and

expectations of the other member.

Interpersonal theorists in the ethological/attachment

theory tradition maintain that as the relationship between

individuals is developed, the goal of the attachment system

is to sustain a sense of security. Hence, individuals are

motivated to develop "working models," or cognitive

representations of the partner that maintains a sense of

confidence in their psychological availability (Bowlby 1977).

Effects on coal compatibility. Between time one and

two, the fruits of success--strategic advantages and joint

profits--are likely to influence perceptions of goal

compatibility and the environment. When strategic advantages

are gained, confidence between the members is built and

familiarity increases. Gulati (1995) shows that when this

happens, firms tend to move away from cautious contracting to

looser practices in their relationship. As familiarity

increases, members might realize new and potential areas for

working together in the future, hence, increased goal


H9: Strategic advantages gained and the
achievement of joint profits at time one
increases perceptions of goal compatibility at
time two.

Effect on complementary competencies. Complementary

competencies at time two are hypothesized to be affected

primarily by the presence of idiosyncratic assets at time

one; however, the direction of the effect is unclear.

Idiosyncratic assets may affect complementary competencies by

increasing or decreasing them. Increases can occur if the

assets help the firms to improve the unique competencies that

already exist in the relationship. However, this is rather

unlikely in the current context. Returning to the GM-Intel

example of the previous chapter, this would mean that GM

becomes better at manufacturing cars and Intel becomes better

at making chips as a result of working together to provide

the marketplace with a superior value car.

Instead, it's more likely that the process of ..'orkingr

closely together requires the two firms to share details of

their competencies that would not have ordinarily been

shared. .s firms learn more about each other's competencies,

the uniqueness of the abilities each firm brings to the

relationship decreases and a new asset is created. Hence,

the presence of relationship-specific assets in time one

decreases the level of unique, complementary competencies

brought to bear on the interorganizational relationship in

time two.

H10: The presence of idiosyncratic assets at time
one decreases complementary competencies at
time two.

Effects on environmental perceptions. The consequences

of working closely together at time one are likely to affect

the members' tercertions of the environment. Recall that

firms work closely together in response to their environment;

by working together, they help each other to better cope with

a dynamic and complex environment. As a result, when the

dyad achieves success together in the form of strategic

advantages and joint profits, both parties' perceptions of

the environments should also change.

When the dyad works closely together at time one,

familiarity and a better understanding of each other's needs,

directions, and capabilities occurs. In light of this new

knowledge, the dyad may realize additional potential

opportunities to work together and mutually benefit. This

would increase their perceptions of the demand munificence of

the environment. When GM and Intel provide a superior-value

car to the marketplace, positive consumer response to the car

may increase the firms' perceptions of potential

opportunities for future success together.

Hll: Strategic advantages gained and the
achievement of joint profits at time one
increases perceptions of environmental
munificence at time two.

Strategic advantages, joint profits, and the presence of

idiosyncratic assets are also likely to make the environment

appear less dynamic and uncertain. Interpersonal theorists

associated with an interdependence theory tradition note that

the benefits of uncertainty reduction in the relationship are

increased prediction and control of one's future outcomes.

As relationship-specific investments build in the

relationship and commitments crystallize, uncertainty becomes

increasingly dissonant with the members behavior orientation.

Hence, they are motivated to achieve confident, trusting

attitudes through psychological means (Holmes & Rempel 1989;

Johnson & Rusbult 1989).

As the buyer-supplier dyad achieves success and

idiosyncratic assets are created, there is motivation to

change their perceptions of the other firm's goals and values

in a manner consistent with the consequences of working

closely together. Similarly, their perceptions of the

environment should change in a manner that is also

consistent. The certainty of attaining strategic advantages,

the increased joint profits earned, and the other members'

commitment to the relationship via idiosyncratic assets are

likely to make the dyad feel more confident and in control of

their position in the marketplace. Hence, the dyad may

perceive its environment to be less dynamic and complex.

H12: Strategic advantages gained, the achievement
of joint profits, and the presence of
idiosyncratic assets at time one decreases
perceptions of environmental dynamism at time

H13: Strategic ad'.antages gained, the achievement
of joint profits, and the presence of
idiosyncratic assets at time one decreases
perceptions of environmental complexity at
time two.

Str-tei: Ad'.'antaqe Development at Time Two

Two of the three classes of antecedents to developing

strategic advantages at time one--partner firm

characteristics and environmental characteristics--are likely

to still be key factors in the decision over time. However,

the moderating effects of trust are likely to be reducci over

time. This is because trust is a powerful assurance when

decision-raking is made under high uncertainty and risk.

When the situation is highly uncertain, individuals are

motivated to learn from and incorporate any new or relevant

information that may help reduce uncertainty. However, when

uncertainty decreases, people become more comfortable in the

situation and will avoid situations that might confront them

with new or potentially inconsistent information (Sorrentino,

Holmes, Hanna, & Sharp 1995).

When firms have limited experience in developing

strategic advantages together, trust acts as a very powerful

pledge--an implicit contract of good intentions--such that

members are willing to infer an optimistic outlook regarding

the consequences of depending on each other and are likely to

have a greater sense of security in working together. Over

time, trust may become a less powerful predictor of

developing strategic advantages, since members have concrete

interactional information on each other. Their shared

experiences, the idiosyncratic assets that are created, and

the increase in perceived environmental based benefits bind

the relationship and make less tangible commitments or

assurances such as trust less important.

All of this is not to say that the role of the firm or

environmental characteristics also become less important. In

fact, their effects on the likelihood of developing strategic

advantages should not change. In and of themselves, they

provide incentives for the development of strategic

advantages. Hence, goal compatibility and complementary

competencies are still expected to be positively related to

the likelihood that the firms will work closely together,

however, their effects at time two are hypothesized to no

longer be moderated by trust.

H14: Firm (goal compatibility, complementary
competencies) and environment (demand
munificence, dynamism, and complexity)
antecedents to developing strategic advantages
at time two are not moderated by trust.

The consequences of developing strategic d'.'antages at

time two are hypothesized to lead to synergistic outcomes and

idiosyncratic assets in a similar manner as in time one.

Hence, working closely together to develop strategic

advantages should result in advantages over competing dyads,

increased joint profits, and the creation of idiosyncratic


H15: The development of strategic advantages at
time two leads to strategic advantages, joint
profits, and idiosyncratic assets.


In this chapter, the conceptual model for the

dissertation was presented. Characteristics of the partner

firm (i.e., goal compatibility and complementary

competencies), and the environment (i.e., demand munificence,

d-.namism, and complexity) are thought to be positively

associated with the degree to which firms will develop

strategic advantages together. However, the effects of these

interorganizational and environmental factors are likely to

be moderated initially by the level of trust between

boundary-spanning individuals. The development of strategic

advantages should lead to synergistic outcomes (i.e., higher

joint profits, strategic advantages over competing dyads) for

the dyad and the creation of idiosyncratic assets.

Over time, the consequences of working together

differentially affect the antecedents to working together in

the future--goal compatibility, complementary competencies,


demand munificence, dynamism, and c:miplexity. These

antecedents lead to the development of strategic advantages

as in time one, however, the moderating role of trust is no

longer present. Finally, the process of working closely

together to develop strategic advantages at time two should

lead to synergistic results--strategic advantages, joint

profits, and idiosyncratic assets./



Development of




Figure 3-1: Overview of the Strategic Advantage Process Model

Partner Firm




-E r'-.VIOR





Develcprment of
Strategic Advantages



Figure 3-2: Antecedents to the Development
of Strategic Advantage.

High -


High Trust

Low "-"

Low Trust




iNJTECEDEIT (e.g., goal compatibility,complementary
competencies, demand munificence,dynamism, complexity)

Figure 3-3: The Moderating Role of Trust


Strategic Advantages


H7 Higher Joint Profits
Development of --
Strategic Advantages


Idiosyncratic Assets

Figure 3-4: Consequences of Developing Strategic Advantages



E 'E-. =TIC




SDevelopment of
Strategic Ad'.'anr, es


Ad-.'ar,n aee

Higher Joint





Figure 3-5: Overview of Time 1 Model

S-H \ U
00 %4 U
H 0J0


4j 00K4

0 (0 0 G 0)

O -H 0

S ) o co1\
\- 4/ 0g /

0: 0, 4 U \
4 U) H




a 1- 4y c E \

o 0



This chapter explicates the survey methodology,

procedures, and initial measurement issues. The first

section discusses the chosen design and methodology. In the

second section, the research setting and sample are

described. The sampling frame and respective industry

characteristics are outlined. The data collection procedure

is delineated in the third section. The last section of this

chapter presents details concerning questionnaire and scale

development, such as pretest results, scale reliabilities and

evidence of the constructs' initial convergent and

discriminant validity.

The Longitudinal Design

The best way to examine potential dynamics in the

development of strategic advantages between firms is via a

longitudinal design. Longitudinal designs have been used

extensively in the organizational behavior literature. One

of the first longitudinal studies examined the reciprocal

causal relationships between employee performance and

expectancy model constructs using a panel design (Lawler

1968). Since then, over 27 longitudinal studies have been

published that examine a number of organizational phenomena

including the relationship between work and nonwork

satisfaction, employee satisfaction and performance, stress

and coping strategies, and leader-subordinate relationships

(see Williams & Podsakoff 1989 for a review).

Despite the difficulty of obtaining data, longitudinal

designs are recommended over cross sectional designs for a

number of reasons. First, longitudinal studies are more

powerful than cross sectional studies in accounting for

longitudinal variation (Kessler & Greenberg 1981). Although

this doesn't mean that causation is proved in an absolute

sense, longitudinal designs allow for explicit testing of the

effect between a construct measured at two points in time,

which is a more suggestive result concerning the true

strength and direction of causation between variables than is

possible with a cross sectional design (Kerr & Schriescheim


Not only do longitudinal designs allow one to examine

the directionality of the relationship between variables, but

they also allow assessment of the degree of mutual dependence

between variables (Bohrnstedt 1975; Heise 1970; Liker,

Augustyniak, & Duncan 1985). Hence, such designs are ideally

suited for examining reciprocal relationships between

attitudes, behaviors, and other organizational phenomena.

Finally, longitudinal research conducted in the

informants' natural setting with the individuals most closely

related to the phenomena of interest are also generally less

susceptible to the criticisms associated with a lack of

external validity and provides an opportunity for determining

the effect of one variable on another in a setting of

greatest importance to managers (Sashkin & Garland 1979).

Time Laq Choice

However, the use of a longitudinal design raises the

issue of determining what the appr:orrlate time lag should be.

In the 27 studies reviewed in Williams and Podsakoff (1989),

there was very little consistency. Ten studies used time

lags ranging from one week to 25 years. Typically, the lag

is chosen out of convenience, and not theory. At this point,

more systematic work is needed in the area that -dentifies

more effectively the appropriate time lag.

In this study, a one year lag is used, partly out of

convenience, but also because pretest interviews with buyers

in the marketplace suggested that within a one year time

period, there appears to be a lot of variation in possible

outcomes. Some informants told of relationships

disintegrating within a one-year period of time, while others

reported little change.

There is also some anecdotal evidence that suggests that

a one-year time lag may not be an unreasonable choice.

Recent research involving interviews with ETC'3 of strategic

alliances indicates that firms need to work together at least

four years before feeling comfortable or secure enough about

their partner to consider working closely to develop

strategic advantages together (Spek-an, Isabella, MacAvoy, &

Forbes 1994). If this is even generally true, then a one year

time lag should be long enough for changes to occur in the

relationships between variables, while still not too long of

a time frame such that changes are no longer detectable.

Unmeasured Variables

As with all designs, longitudinal designs are

susceptible to criticisms concerning potential omitted

variables (Cliff 1983; Duncan 1969). Omitted variable

effects are summarized in the error terms of a system of

equations. However, in a longitudinal design, the adverse

effects of omitted variables are particularly problematic,

since the error terms of similar variables can occur and

potentially influence relationships between variables at

numerous and repeated points in the system of equations.

The only way to reduce the possible adverse effects of

omitted variables is to move beyond some of the simple two-

wave, two-variable models examined to date and try to

carefully select those predictors of a variable that are most

likely to exert significant effects on it (Williams &

Podsakoff 1989). In this study, antecedents to developing

strategic advantages are selected from two independent

sources--those specific to firm characteristics and those

specific to the external environment surrounding the dyad.

Hence, maximally different types of variables are chosen in

attempting to explain the phenomena of interest.

Additionally, this study moves beyond a simple, two variable

relationships to a mediated, multi-variable process. The

full, longitudinal theoretical model is depicted in Figure 3-


Key Informant a ethdol.1"--

With the exception of trust and environmental factors,

the conceptual model is primarily concerned with

interorganizational constructs--characteristics of the member

firms involved. These constructs are unobservable,

theoretical, shared constructions describing the member firms

or what they do. As such, key informants are asked to

provide the researcher with reports on observed measures in

order to infer the nature of these interorganizational

constructs. The key informant technique was originally

associated with participant observation studies in

ethnographic research (Lofland 1971), although nowadays it is

often used in survey contexts to obtain quantifiable


., Key informants differ from survey respondents in that

informants are asked to generalize about patterns of behavior

after summarizing either observed (actual) or expected

(prescribed) interorganizational relations. In contrast,

survey respondents provide descriptions of personal feelings,

opinions, and behaviors. Hence, key informants are asked to

explain and predict the behavior of organizations rather than

individuals (Seidler 1974). Since their task is more

complex, key informants are chosen on a non-random basis;

they are typically chosen because they have special

qualifications such as a particular status, specialized

knowledge, or even accessibility to the researcher.

Three issues about the use of this methodology should be

noted. First, since individuals are asked to make

generalizations about organizations that have characteristics

distinct from the individuals themselves, systematic biases

may affect the informant's responses. Over or under reporting

can occur due to the informant's position, job satisfaction,

or other characteristics. Fortunately, measures have been

developed to evaluate informant characteristics and help

assess the degree to which this problem exists (Kumar, Stern,

& J. C. Anderson 1993). A more detailed discussion of this

will be presented in the section on informant


Finally, despite the potential for systematic or

perceptual biases, researchers still recommend the use of

multiple key informants because data from only a single

informant precludes a rigorous assessment of the measures'

validity and reliability, as well as the convergent and

discriminant validity of informant reports (Campbell 1959;

Phillips 1980). A number of researchers have developed

quantitative techniques based on multi-trait, multi-method

approaches that allow decomposition of the error term into

variance due to the trait, the informant, and random error

(E. Anderson 1985; Bagozzi & Phillips 1982; Bagozzi & Yi

1994; John & Reve 1982; J6reskog 1971; Kenny & Kashy 1992;

Kumar & Dillon 1990; Marsh 1989; Phillips 1981). These

techrniues will be used in the present analysis.

Second, the use of multiple informants leads to another

issue--perceptual agreement--the degree of similarity in

reports from multiple informants. Social psychologists have

long asked whether people's perceptions of others are

reflections of reality or merely "the eye of the beholder."

If perceptions are the former--in other words, if they are

data driven--then researchers should observe similar reports

from multiple informants (Higgins & Bargh 1987). If

perceptions are theory driven, then individual reports need

not converge. Historically, research in the channels

literature utilizing multiple informants has failed to show

str--. evidence of perceptual agreement (J. C. Anderson &

Narus 1990; John & Reve 1982; Kumar et al. 1993; Phillips

1981), with the exception of Bagozzi and Phillips (1982) and

J. C. Anderson (1987). The causes behind the divergence in

reports and the appropriate methods to handle discrepant

reports are still unresolved issues.

Research Setting and Sample

The Firms

The vertical dyads examined in this study are

manufacturing companies and their suppliers. Industrial

buying relationships are ideal for examining strategic

relationship dynamics compared to buying relationships

between manufacturers and distributors suth as wholesaling

companies and retailers. Relationships between manufacturers

and distributors differ from industrial buying relationships

in that distributors must procure a variety of products from

competing suppliers. As a result, it is difficult for them

to develop a close, strategic relationship with any one

supplier.1 Strategic relationships inherently possess an

element of exclusivity, in that the firms are trying to

create a unique, strategic advantage that directly competes

with other buyer-supplier dyads in the marketplace. If

retailers and their suppliers did this, it would likely

alienate other important sources of supply or key customers.

Hence, the distributor's need to carry wide assortments makes

a manufacturer-retailing setting unsuitable for this study.

1 Some might argue that the development of efficient
consumer response (ECR) systems between retailers and
manufacturers are examples of strategic relationships.
However, these computer-linked inventory management systems
are becoming so common in the retailing industry that they no
longer provide channel dyads with a unique, sustainable
advantage over competing dyads. These systems will soon
become an industry standard--a cost of doing business in a
particular industry.

Access to the participating firms was provided by the

Marketing Science Institute (MSI) anr the Institute for the

itudy of Business Markets (ISBM) at Pennsylvania State

University. Each firm was offered an executive summary,

presentation of results, and customized analyses in return

for their participation. The procurement divisions of four

Fortune 5 manufacturing companies agreed to participate--a

computer manufacturer, a photography equipment manufacturer,

a chemical manufacturer, and a brewery.2 Their participating

divisions are listed in Table 4-1. Each of these companies

were as.ed to identify corresponding suppliers as potential

participants in the study.

The Transactirona1 .eaIr:ns-niDs

The sur'-.eyed i'ads represented significant purchasing

arrangements, as shown in Table 4-2. The firms had worked

with each other 13.7 years on average and annually purchased

over $560 million in materials and services. This

represented ,2pr::.iKately 27% of the supplier's total annual

sales in the category. Nearly 60% of the purchases typically

made were a mixture of first-time, routine, and modified

routine purchases. Over 50% of the relationships surveyed

described themselves as being in the mature phase of their

relationship life cycle. This means they felt that they had

2 All four companies have requested that their identities be
kept anonymous.

an ongoing, long term relationship in which both members were

receiving acceptable levels of satisfaction and benefits from

the relationship.

Unit of Analysis

Since the goal of this study is to understand strategic

relationship dynamics and a relationship requires at least

two parties, the unit of analysis is the channel dyad. As a

result, the measures used in this study were designed to tap

aspects of the mutual relationship between the firm, not the

individual perceptions of the two firms, since the focus of

the conceptual model is on joint, mutually shared activities

and outcomes, not potential asymmetries between the two


This means that the surveys were designed with the

intent of using individual buyer and supplier representatives

as independent informants of the constructs in the model,

i.e., they completed identical items tapping the nature of

the mutual relationship between the firms. The use of

multiple informants in this manner allows for assessment of

the construct validity of the organizational properties

measured and partitioning of error variance into trait,

informant, and random error variances.

Informant Characteristics

Informants were manufacturing buyers and supplier

representatives. In order to maximize the sample size and

minimize potential adverse effects of informant attrition at

time two, buyers were asked to report on two supply

relationships. These relationships did not necessarily need

to be strategic, nor did they have to involve their largest

suppliers. However, the buyers were asked to select

relationships that were maximally different (e.g., a positive

and a negative relationship) if possible, so that range

restriction on the measures would not inhibit model


Within the supplier firms, buyers were asked to identify

the supplier representative that they interacted with most

frequently on a regular basis, with the one stipulation that

the representative be a person with whom they have had

frequent interaction with for at least one year. Preliminary

interviews with buyers during the survey pretest indicated

that they typically required at least one year of interaction

with a supplier representative to feel comfortable about

making reports on the level of interpersonal trust present in

the relationship. This one year interaction requirement also

helped insure that the sampling frame consisted of long term

relationships instead of new, developing relationships. The

average relationship length between buyers and supplier

representatives in this study was 3.7 (s.d. 3.5) years,

which, according to anecdotal evidence (Spekman et al. 1994),

seems to be a reasonable length of time for members to learn

enough about each other's goals, intentions, and

trustworthiness to feel comfortable about the possibility of

working more closely together.

Historically, researchers have relied on global and

specific measures to assess informant competency. Examples

of global measures of informant competency include length of

an informant's tenure with a firm (Phillips 1981), the

informant's knowledge of or involvement with other firms

engaged in relationship with the focal firm (Heide & Miner

1992), or the length of time the informant has observed or

interacted with other firms (Phillips 1982).

In this study, the buyers had 11.2 (s.d. 8.2) years of

experience in purchasing on average, and had been with their

present companies 20.9 (s.d. 7.4) years on average, both are

positive indicators of informant competency. The supplier

representatives averaged 15.1 (s.d. 9.7) years of experience

and had been with their companies 14.2 (s.d. 9.4) years on

average. It could be that turnover is more pervasive among

supplier representatives than buyers. However, these

indicators do attest to the representatives' competency as


Researchers have also used specific measures of

competency to assess the level of an informant's knowledge of

each major issue included in a study (e.g., Cusumano &

Takeishi 1991). In this study, at the end of the

questionnaire, the informants were asked questions such as,

"How knowledgeable are you about the level of trust in your

firm's relationship with this firm?" Their responses were

marked on 7-point Likert scales (l=Not Very Knowledgeable,

7=Very Knowledgeable). The average responses to these scales

for buyers and suppliers was 5.6.


Time One

Two hundred buyers from the four firms were surveyed in

February 1994, creating an initial sampling frame of 400

dyadic relationships. These questionnaires were mailed to

the buyers along with a pre-addressed, postage-paid envelope,

a cover letter from the researcher and a memorandum from

corporate executives that explained the purpose of the study,

stressed the need for participation, and assured

confidentiality of all responses. A copy of the cover

letters and survey is reproduced in appendix A. Each survey

required approximately fifteen minutes to complete.

In the survey, buyers were asked to z-cply the names and

addresses of supplier companies and remcresentatives who met

the aforementioned criteria--maximally different relationships

that have been ongoing for at least one year. The buyers

then returned the questionnaires directly to the researcher

and similar questionnaires were sent to the named supplier

alng with a self-addressed, postage-paid envelope, and cover

letters from the researcher and corporate executives. A copy

of the cover letters and survey is presented in Appendix B.

The supplier survey required ten minutes to complete. The

difference in time required is accounted for by the fact that

the buyer had to supply the supplier's name, the

representative's name and address.

Thirty-one of the buyers were eliminated from the sample

for various reasons: some had been recently reassigned or

didn't work with suppliers on a long term basis, or simply

refused to complete the surveys. Two hundred seventy-five

buyer surveys were returned (a 75% response rate) and 220

corresponding supplier surveys were completed (an 80%

response rate). Non-respondent differences were assessed by

testing for differences between early and late respondents

(Armstrong & Overton 1977). Early responses were defined as

the first 75% of the questionnaires returned, while the

remaining 25% were defined as late responses. Univariate

tests between the observed responses used in the analysis

were conducted. The results were essentially not

significantly different.

Time Two

In February 1995, the buyers and suppliers who returned

surveys in the initial data collection were again sent cover

letters, memorandums, and surveys in a similar fashion to

time one. This survey was essentially identical to the time

one survey. A copy of the buyer and supplier surveys and

cover letters are displayed in appendix C and D,


Of the 275 buyers and 220 suppliers surveyed, 42 buyers

and 12 suppliers changed representatives and were no longer

working with the same buyer or supplier representative from

time one, two supplier sur'.eys were returned to the sender

and no forwarding addresses could be found, two buyers and

four suppliers refused completion of the survey, and ten

purchasing relationships were terminated over the one year

period. Collectively, this represents a 20% attrition rate

for buyers (54 informants), and a 13% rate for suppliers (28

informants). 167 buyer surveys and 154 supplier surveys were

used in the analysis. In total, 80% of the buyers and 83% of

the suppliers at time one responded to the second survey.

As before, non-respondent differences were assessed by

conducting univariate tests between the observed responses of

the first 75% of the surveys returned and the last 25%.

There were essentially no significant differences.

Questionnaire and Scale Development


In January 1994, depth interviews and protests of the

survey were conducted with buyers in the computer

manufacturing and chemical manufacturing firms. In the first

pretest, six buyers in the personal computer division were

given the proposed survey and cover letters. Their comments

and suggestions for improvement were used to revise the

survey. Three weeks later, six buyers from several

purchasing divisions in the chemical manufacturing firm were

asked to complete the revised survey.

In both pretest sessions, the amount of time required

for completion was recorded along with problematic issues or

ambiguities, and buyers were probed to insure that the survey

would be understandable to informants. Additional in-depth

interviews were conducted with buyers at both firms to

provide the researcher with a better understanding of the

buyers' environments, internal and external demands, and the

nature of their task and supplier relationships. Interviews

were also conducted with manufacturing buyers at the other

participating companies as well as non-participating

companies that the researcher had access to through MSI and



Measure development was based on the procedure

recommended by Nunnally (1978). All constructs were measured

with multiple-item, 7-point Likert scales in simple terms

using the language commonly employed by the informants.

Scales for constructs with no precedent in the literature

(e.g., the development of strategic advantages and

complementary competencies) were created for the purpose of

this study. Scale items and measures for constructs that

have appeared before in the channels literature were adapted

and used whenever possible. All the scales for the constructs

used in this study (with the exception of complementary

competencies and environmental variables) were used

previously in a nationwide study of purchasing agents (Jap &

'Jeitz 1994), where they demonstrated acceptable levels of

reliability and construct validity. The items were

identically worded for both the buyer and supplier.

C.:nstruct "'alidit*',. a~ Time 'ne

Construct validity is the extent to which an observation

measures the concept it is intended to measure. The goal of

measure validation is to demonstrate the internal

consistency, reliability, and unidimensionality of the


Selecting a Method

Historically, researchers have advocated the calculation
of item-total correlations as a first step in the measurement

validation process (Nunnally 1978; Churchill 1979). However,

an item-total analysis does not account for external

consistency, and hence, may fail to discriminate between sets

of indicators that represent different, though correlated

factors (Gerbirna & J. C. Anderson 1988).

Principal components analysis and common factor analysis

have also been used to ascertain construct validity.

However, these have significant shortcomings in the

measurement of organizational constructs because they often

assume uncorrelated traits or factors and their use on data

exhibiting correlated factors can produce distorted factor

loadings and incorrect conclusions about the number of

factors (Bagozzi 1983). Moreover, both of these approaches

assume the absence of measurement error (Lawley & Maxwell

1971). This is particularly problematic in the measurement

of organizational constructs, since covariance among

organizational properties and error in measurement is

expected to occur.

As a result, confirmatory factor analysis was used in

the present analysis because it allows for explicit

representation of the degree of correspondence between

observed measures and concepts. A multiple-indicator

measurement model of the constructs of interest allows for

unambiguous assignment of meaning to estimated constructs. A

first-order, single-factor model for each set of congeneric

items per construct was estimated using full-information

maximum-likelihood (MLE) techniques in LISREL 8.03 (Jbreskog

& S6rbom 1993). All of the measurement model estimation in

this chapter and the next was conducted using this method and

statistical package. MLE estimation was chosen because it

provides consistent estimates that are also asymptotically

normal, unbiased, consistent, and efficient (Kmenta 1971);

these parameters are particularly useful in the derivation of

asymptotically valid tests, and are not available with

ordinary least squares, or two- or three-stage least squares.

Items with low or nonsignificant loadings or items that

loaded on multiple factors were eliminated from the


3 When this did occur, only one or at the most two items
were eliminated per construct.

The measurement models were initially estimated for

buyers and suppliers separately. Each buyer survey was

treated as an independent response thr':qughut the analysis in

this study in order to provide the large sample size

necessary for stable parameter estimation. Since the

statistical properties of MLE estimators are asymptotic;

hence, they require information from higher order moments and

are only true for large samples. The complete data analysis

was also conducted with truly independent buyer surveys--only

one reported relationship per buyer (n=129), and the results

were not significantly different, so the multiple responses

from buyers were retained.

Problems with En'vironmental C.mpleyxit, :cale

Estimating a unidimensional factor model for

environmental complexity was difficult. Several items loaded

on more than one construct and Heywood cases (negative error

variances) were observed. This could be due to the fact that

the construct itself is too broadly defined--encompassirng

competitor, product, and consumer forces--and impossible to

measure as a unidimensional phenomenon. Unfortunately,

because of space constraints in the survey, it was not

possible to include more than five indicators for this

construct. Additional indicators might have provided the

stability needed to measure environmental complexity more

completely using multiple dimensions.

It is also likely that the informants had difficulty in

establishing a reference point for their reports. The

measures were intended to measure the dyadic external

environment surrounding the informants within the product

category being purchased. Pretests indicated that buyers and

suppliers were only able to report on their immediate

environment with confidence. Often, the informants were not

able to comment on environmental forces that shaped their

firms' decisions beyond their SBU.

If the informants did not read the directions carefully,

they might have completed the measures using their own

industries as referents, which could account for the

problematic fit. Several alternative models, including

higher-order factor models (Gerbing & J. C. Anderson 1984)

were estimated, in an attempt to recover as much usable data

as possible. Because of these problems, it was difficult to

unambiguously assign meaning to the estimated constructs.

The end result was that the measures for environmental

complexity could not produce an adequate measurement model,

and had to be eliminated from the study altogether.

Measurement Model Evaluation

Table 4-3 contains the scale items, factor loadings,

standard errors, and reliabilities for the latent constructs

for buyers and suppliers separately. Reliability was

estimated using the reliability formula of J6reskog (1971).

This expression is superior to coefficient alpha in that it

does not assume equal item reliabilities within a

confirmatory factor analysis context and will not

underestimate the reliabilities when the scale is comprised

of a small number of items.

Inspection of the factor loadings and standard errors at

this point is informative in assessing the internal fit of

these single-factor, measurement models. As a rule, the

loadings should be approximately .6 and higher in value, and

the standard errors should be small, approximately .4 or less

(Bagozzi & Yi 1988). All parameters should be significant

and the scale reliabilities should be approximately .6 or

higher. Convergent validity refers to the degree to which

multiple attempts to measure the same concept are in

agreement. Factor loadings with t-values of two or more

demonstrate evidence of convergent validity.

Table 4-4 contains the summary information on the

measurement models, completely standardized correlation

matrix between constructs, and means for each construct for

buyers and suppliers, respectively. The overall chi-square

is a likelihood ratio statistic testing a hypothesized model

against the alternative that the covariance matrix is

unconstrained. Although this is useful in gaining

information about a statistically false model, the chi-square

test is well known for its sensitivity to sample size. As

the sample size increases the chances of rejection also

increases, regardless of whether the model is true or false.

Hence, the significant chi-squares listed in this table for

the measurement models should not be taken too seriously as

an indicator of overall model fit. The chi-square will be

more useful when nested, alternative models are tested in the

next chapter.

Bentler's (1990) comparative fit index (CFI) is one of

many goodness of fit indices (GFI) available to assess the

fit of a model to data. These indices range in possible

values between zero and one, indicating a lack of fit and

perfect fit between the theoretical model's covariance matrix

to the observed covariance matrix, respectively. The CFI is

chosen over Bentler and Bonnet's (1980) normed fit index

(NFI) and non-normed fit index (NNFI). The NFI has been

shown to significantly underestimate fit when sample sizes

are less than 200 (Marsh, Balla, & McDonald 1988) and will

vary considerably in value for the same model and data

depending on whether MLE or generalized least squares is used

(Tanaka 1987). The fit indices provided by LISREL do not

have the same underestimation bias in samples of less than

200, but do demonstrate a notable sample size effect. This

is to say that as the sample size increases, the fit indices

tend to increase as well (Marsh et al. 1988). The CFI is an

index that is able to avoid extreme under- or over-

estimation that occurs with other GFIs.

Discriminant validity is the degree to which measures of

different concepts are distinct. This is typically assessed

by constraining the correlation parameter between two

constructs to unity and performing a chi-square difference

test on the values obtained for the constrained and

unconstrained models (J6reskog 1971). If the chi-square for

the unconstrained model is less than the chi-square for the

constrained model, this indicates that the traits are not

perfectly correlated and that discriminant validity is

achieved (Bagozzi & Phillips 1982).

A more stringent test of discriminant validity was

developed by Fornell and Larcker (1981). They recommend

examining the amount of variance captured by the construct in

relation to the amount of variance due to measurement error.

Unlike the previous approach, this test recognized that

measurement error can vary in magnitude across a set of

construct indicators. The correlations that did not meet

this requirement are underlined in the table. The chi-square

difference test proposed by Joreskog (1971) was then

conducted on these pairs and in every case, the constructs

were shown to be significantly different.

The currpoe of this section was to demonstrate that the

observed measures appear to be corresponding to the

constructs that they are intended to measure. Although it

may appear upon close inspection that an occasional parameter

loading or scale reliability fell below "acceptable" levels

of model criteria,4 this is typical when estimating a large

measurement model with multiple indicators on a large data

set. On balance, the constructs do appear to be

4 Researchers continue to debate on what the exact levels of
acceptability are on some evaluative criteria.