Content uploaded by Vincenzo Perrone
Author content
All content in this area was uploaded by Vincenzo Perrone
Content may be subject to copyright.
1047-7039/98/0902/0141/$05.00
Copyright q1998, Institute for Operations Research
and the Management Sciences O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 141
Does Trust Matter? Exploring the Effects
of Interorganizational and Interpersonal
Trust on Performance
Akbar Zaheer • Bill McEvily • Vincenzo Perrone
Carlson School of Management, University of Minnesota, 321 19th Avenue South, Minneapolis, Minnesota 55455
Graduate School of Industrial Administration, Carnegie Mellon University, Schenley Park, Pittsburgh,
Pennsylvania 15213-3890
Universita’ degli Studi di Cassino and SDA Bocconi University, Graduate School of Management, Italy
Interpersonal and inter-organizational trust have been widely cited as important components of eco-
nomics exchanges. However, rarely have these concepts been measured and their implications ex-
amined. This paper focuses on these difficult tasks. Jay B. Barney
Abstract
A conceptual challenge in exploring the role of trust in inter-
organizational exchange is translating an inherently individual-
level concept—trust—to the organizational-level outcome of
performance. We define interpersonal and interorganizational
trust as distinct constructs and draw on theories of interorga-
nizational relations to derive a model of exchange performance.
Specifically, we investigate the role of trust in interfirm ex-
change at two levels of analysis and assess its effects on ne-
gotiation costs, conflict, and ultimately performance. Proposi-
tions were tested with data from a sample of 107 buyer-supplier
interfirm relationships in the electrical equipment manufactur-
ing industry using a structural equation model. The results in-
dicate that interpersonal and interorganizational trust are related
but distinct constructs, and play different roles in affecting ne-
gotiation processes and exchange performance. Further, the hy-
potheses linking trust to performance receive some support, al-
though the precise nature of the link is somewhat different than
initially proposed. Overall, the results show that trust in inter-
organizational exchange relations clearly matters.
(Interorganizational Trust;Interfirm Relations;Rela-
tional Governance)
The sharp increase in various forms of interfirm cooper-
ation in the current economic environment has stimulated
a profusion of research on such interfirm ties, and more
specifically on the role of trust in facilitating the organi-
zation and coordination of economic activities between
firms. A special issue of The Academy of Management
Journal (1995) on the subject of collaborative interfirm
ties noted editorially that the role of trust in cooperative
relationships is of fundamental importance and that “the
study of trust and its impact on cooperative relationships
at all levels may be a particularly fruitful area of future
research” (Smith et al. 1995, p. 15). Despite the impor-
tance of understanding the nature of trust in cooperative
interfirm relationships, considerable ambiguity is evident
in the literature about the precise role of trust as it oper-
ates at different levels of analysis and its influence on
performance.
A fundamental challenge in conceptualizing the role of
trust in economic exchange is extending an inherently
individual-level phenomenon to the organizational level
of analysis. Not clearly specifying how trust translates
from the individual to the organization level leads to theo-
retical confusion about who is trusting whom because it
is individuals as members of organizations, rather than
the organizations themselves, who trust. The ambiguity
about the multilevel nature of trust is apparent in theo-
retical frameworks that address the role of trust in eco-
nomic exchange, notably transaction cost economics
(Williamson 1975, 1985). Specifically, transaction cost
theory implies that firms tend to behave opportunistically
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
142 O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998
Figure 1 Interorganizational and Interpersonal Trust
(Williamson 1975, 1985), thereby attributing (micro) in-
dividual motivations and behaviors to (macro) organiza-
tions and thus committing a “cross-level fallacy”
(Rousseau 1985). We maintain that theories of interfirm
exchange that simply view opportunism—or conversely,
trust—as a property of organizations without specifying
the link between micro and macro levels is inaccurate as
it tends to anthropomorphize the organization.
Some imprecision about the roles of trust at multiple
levels of analysis and how trust is related across levels is
also apparent in economic sociology, more specifically in
relational exchange (Dore 1983, Macneil 1980). A central
premise of relational exchange theory is that personal re-
lations generate trust and discourage opportunistic behav-
ior between firms. However, that perspective does not
address how interpersonal trust influences the economic
exchange between and among complex social systems,
such as organizations. Simply stated, the relational ex-
change perspective does not stipulate the mechanisms by
which individual-level action affects organizational-level
outcomes.
Further, though several theoretical traditions have rec-
ognized the importance of trust in economic exchange
(e.g., Arrow 1974, Granovetter 1985, Macauley 1963),
little research has been done to explain how trust, par-
ticularly when conceptualized as a multilevel phenome-
non, operates to affect the performance of interfirm ex-
change. In particular, the organizational and strategy
literature (e.g., Gulati 1995) has asserted that trust in in-
terfirm exchange is beneficial and can be a source of com-
petitive advantage (Barney and Hansen 1995). In the or-
ganizational economics literature, trust has been
theorized to reduce opportunistic behavior, and hence
transaction costs of exchange, ultimately resulting in
more efficient governance (Bromiley and Cummings
1995, John 1984). However, though the link between trust
and performance in economic exchange has been fre-
quently theorized in general terms, elucidating the precise
nature of the trust-performance relationship in a multi-
level context remains an important theoretical and em-
pirical challenge.
The primary objective of our research is to explain how
trust operates at both individual and organizational levels
of analysis, how trust at the two levels is related, and
particularly how the mechanisms by which this inherently
individual-level phenomenon translates into an
organizational-level outcome: performance. We argue
that in an actual exchange relationship, the role of indi-
vidual boundary spanners, acting on behalf of their or-
ganizations, has an important influence on interfirm ex-
change. In particular, we propose that institutionalized
practices and routines for dealing with a partner organi-
zation create a stable context within which interorgani-
zational and interpersonal trust develop. Further, we
maintain that there is a reciprocal relationship between
trust at the two levels of analysis, and neither alone is
sufficient for understanding relational exchange perfor-
mance. We conceptualize the trust-performance link as
being mediated by interfirm negotiation processes, and
we identify the influence of trust at each level on eventual
exchange performance.
We tested our propositions on a sample of 107 buyer-
supplier interfirm relationships in the electrical equipment
manufacturing industry using a structural equation model.
Our hypotheses linking trust to performance are partially
supported, but several of our findings were unexpected.
The results indicate that interpersonal and interorganiza-
tional trust are related but distinct constructs, and play
different roles in affecting exchange performance.
Theory and Hypotheses
Interpersonal and Interorganizational Trust
We use the term interpersonal trust to refer to the extent
of a boundary-spanning agent’s trust in her counterpart
in the partner organization. In other words, interpersonal
trust is the trust placed by the individual boundary span-
ner in her individual opposite member. The term inter-
organizational trust is defined as the extent of trust placed
in the partner organization by the members of a focal
organization (see Figure 1).
Although considerable research in psychology and so-
ciology has focused on trust in individuals (e.g., Rotter
1971) and in social groups (e.g., Lewis and Weigert
1985a, Zucker 1986), in the organizational and interor-
ganizational context the role of trust has only recently
attracted interest, notably in the marketing channels lit-
erature (e.g., Anderson and Narus 1990), and in strategy
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 143
and organizational research (e.g., Bradach and Eccles
1987, Ring and Van de Ven 1992, Zaheer and
Venkatraman 1995). As a result of both the range of dis-
ciplinary lenses used to study trust and the inherent am-
biguity of the trust construct, there is currently a confus-
ing assortment of conceptual perspectives on trust. For
our purposes, two principal issues are involved in devel-
oping a working definition of the trust construct: (1) the
intrinsically complex and multifaceted nature of trust
(Corazzini 1977) and (2) the variety of units and levels
of analysis to which trust has been applied.
In addressing the first issue, we focus on the exchange
dyad to develop a definition of trust that is inherently
relational because we are specifically interested in study-
ing the role of trust in economic exchange and, more gen-
erally, in interorganizational relationships. The literature
on interorganizational relations provides two general def-
initions of trust: confidence or predictability in one’s ex-
pectations about another’s behavior, and confidence in
another’s goodwill (Ring and Van de Ven 1992). We use
and extend both conceptualizations and define trust as the
expectation that an actor (1) can be relied on to fulfill
obligations (Anderson and Weitz 1989), (2) will behave
in a predictable manner, and (3) will act and negotiate
fairly when the possibility for opportunism is present
(Anderson and Narus 1990, Bromiley and Cummings
1995). Our conceptualization of trust as an expectation
rather than a conviction reflects an uncertain anticipation
of the referent’s future behavior. Our definition explicitly
allows for the possibility of betrayal, which we view as
an inherent feature of trust. This probabilistic element of
trust is analogous to making a “leap of faith” by placing
confidence in a referent without knowing with absolute
certainty that the referent’s future actions will not produce
unpleasant surprises. Our definition therefore character-
izes interorganizational trust as being based on three com-
ponents—reliability, predictability,
1
and fairness—that
capture some of the complexity of trust (Lewis and
Weigert 1985b, Rempel and Holmes 1986, Rempel et al.
1985).
Our conceptualization of trust also explicitly recog-
nizes the distinction between relational and dispositional
trust. Whereas dispositional trust is an individual trait re-
flecting expectancies about the trustworthiness of others
in general (Rotter 1971), relational forms of trust pertain
specifically to the counterpart in the dyad. That distinc-
tion is important because high relational trust does not
imply naively trusting all exchange partners. Rather, re-
lational trust is likely to be based on experience and in-
teraction with a particular exchange partner (Ring and
Van de Ven 1992).
The second issue involved in developing a working
definition of the trust construct concerns the level of anal-
ysis involved. The Oxford English Dictionary defines
trust as “confidence in or reliance on some quality or at-
tribute of a person or thing” (emphasis added), highlight-
ing the fact that the referent of trust may vary. In other
words, it is conceptually consistent to view trust as being
placed in another individual or in a group of individuals
such as the partner organization. However, the same can-
not be said for the origin of trust. In our view, trust has
its basis in individuals, although individuals in an orga-
nization may share an orientation toward another orga-
nization. From this perspective, interorganizational trust
describes the extent to which organizational members
have a collectively-held trust orientation toward the
partner firm, which is quite different from saying that
organizations trust each other. In contrast, we view inter-
personal trust as also made up of the three elements—
reliability, predictability, and fairness—but with an in-
dividual as both the referent and origin of trust.
We note that trust as represented by the three compo-
nents can be expressed in different forms—cognitive, be-
havioral, and emotional—at both the interpersonal and
the interorganizational levels. However, we recognize
that to do full justice to the complexity of trust would
require considerably more elaborate schemes (e.g.,
Cummings and Bromiley 1996).
Micro-Macro Links in Interorganizational Relations
Organizations are far from monolithic. Rather, as Pfeffer
(1982, p. 64) notes, they are “. . . pluralistic, divided into
interests, sub-units, and subcultures.” Similarly, rather
than the relationship between organizations being face-
less and monolithic, it is actively handled and managed
by individual boundary spanners (Katz and Kahn 1978)
whose orientations and motivations may well be different
from those of the organization as a whole. Boundary
spanners are more closely involved in the interorganiza-
tional relationship than other members of the organiza-
tion, and tend to interact with their counterparts to a
greater extent (Friedman and Podolny 1992). Hence,
when examining the characteristics of an interorganiza-
tional relationship, we need to study the individual and
organizational levels simultaneously (Rousseau 1985). In
the eyes of some scholars, theory-driven research on mul-
tilevel phenomena is what “sets [this] field apart from its
parent disciplines in that most of what we study in and
about organizations are phenomena that are intrinsically
mixed-level” (Rousseau 1985, p. 2). In particular, the out-
comes of trust may be different when the boundary-
spanning individual is considered separately from the rest
of the organization, with important consequences for per-
formance.
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
144 O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998
We argue that the connection between interpersonal
and interorganizational trust is based on institutionalizing
processes. Over time, repeated ties between two firms
evolve into deeper and more stable cooperative arrange-
ments (Gulati 1995). Although individual boundary span-
ners come and go, role definitions are stable and enduring
(Ring and Van de Ven 1994). Institutionalizing processes
codify informal commitments made by individual bound-
ary spanners, which over time become established and
taken-for-granted organizational structures and routines
(Zucker 1977). Thus, as new individual boundary span-
ners enter the interfirm exchange relationship, they be-
come socialized into the norms of the mini-society it
represents (Macneil 1980). Norms from the interorgani-
zational relationship are internalized and recreated in
boundary spanners’ interpersonal trust orientations to-
ward each other in the process of conducting exchange.
At the same time, interpersonal trust becomes reinstitu-
tionalized and boundary spanners’ trust orientation in turn
influences the orientation of other organizational mem-
bers toward the partner organization. The creation and
recreation of trust structures and action at the interper-
sonal and interorganizational levels is akin to Giddens’
(1979) concept of structuration. In sum, the foregoing ar-
guments imply that interorganizational trust and interper-
sonal trust exert a positive influence on each other.
H
YPOTHESIS
1. There is a positive relationship be-
tween the extent of interpersonal trust and interorgani-
zational trust in an interfirm dyad.
Trust as Reducing Costs and Improving
Performance
Trust as Reducing Costs of Negotiation. Consistent with
the organization economics literature, we define negoti-
ating costs as those involved in reaching mutually ac-
ceptable agreements. By this definition, negotiating costs
include the time and effort required to determine “effi-
cient courses of action, and to settle on divisions of costs
and benefits” (Milgrom and Roberts 1992, p. 147). In
particular, bounded rationality, uncertainty, and infor-
mation asymmetries stemming from imperfect commu-
nication, private information, and observation and veri-
fication difficulties, all contribute to increasing costs of
negotiation.
Negotiations are less costly under conditions of high
interorganizational trust because agreements are reached
more quickly and easily as parties are more readily able
to arrive at a “meeting of the minds.” Interorganizational
trust mitigates the information asymmetries inherent in
interfirm exchange by allowing more open and honest
sharing of information. As a result, partners’ expectations
about how changes in the external environment will affect
their exchange relationship tend to converge (Malmgren
1961). Specifically, when unforeseen contingencies arise,
such as costs not explicitly covered by the terms of a
contract, high levels of trust facilitate the development of
a common understanding about the contingencies and
how they might be resolved. When interorganizational
trust is high, negotiating positions are based on similar
underlying assumptions, and agreements are likely to be
reached more quickly. In a related vein, we suggest that
trust promotes negotiating efficiency by enabling each ex-
change partner to be more flexible in granting conces-
sions because of the expectation that the other exchange
partner will reciprocate in the future (Dore 1983).
2
From a transaction cost perspective, the most compel-
ling argument for the superior efficiency of interorgani-
zational relationships that involve trust is simply that trust
reduces the inclination to guard against opportunistic be-
havior (i.e., deliberate misrepresentation) on the part of
the exchange partner (Bromiley and Cummings 1995,
Dore 1983, John 1984). Under conditions of low trust,
lengthy and difficult negotiations over unforeseen contin-
gencies are likely to take place between exchange part-
ners because of the possibility of both ex ante and ex post
opportunism (Williamson 1975). In addition, contractual
and structural safeguards are put in place to protect in-
vestments in the relationship (Williamson 1985). Under
high trust conditions, in contrast, firms are less inclined
to rely on elaborate safeguards for specifying, monitor-
ing, and enforcing agreements. Relations high in trust
therefore imply more efficient exchange governance in
the form of eased negotiations.
H
YPOTHESIS
2. There is a negative relationship be-
tween interorganizational trust and costs of negotiation
between the partners to the exchange.
At the level of individual boundary spanners, the ne-
gotiation literature suggests that a trust-based relationship
is conducive to the discovery of mutually beneficial, in-
tegrative solutions (Walton and McKersie 1965). Trust-
ing relations between boundary spanners also have been
found to be associated with easier negotiations (Currall
and Judge 1995). Negotiations may be easier because of
boundary spanners’ willingness to share sensitive infor-
mation and their confidence that information provided by
the counterpart is not misrepresented. Consequently, we
expect mutually beneficial agreements to be reached more
quickly when boundary spanners trust each other. We ar-
gue that the relationship between interpersonal trust and
costs of negotiation is likely to be negative, similar to that
between interorganizational trust and negotiating costs.
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 145
H
YPOTHESIS
3. There is a negative relationship be-
tween interpersonal trust and costs of negotiation be-
tween the partners to the exchange.
Trust as Reducing Conflict. Because firms’ objectives
in exchange relationships are rarely identical, divergent
goals and unforeseen contingencies in their day-to-day
relationships are likely to result in disagreements. Rela-
tional exchange ties that are imbued with trust are char-
acterized by the internal harmonization of conflict and an
array of norms and social processes that work to preserve
the relationship (Macneil 1980). Partners in relationalex-
change that have forged a high level of interorganiza-
tional trust are more likely to give each other the benefit
of the doubt and greater leeway in mutual dealings. Such
leeway will tend to reduce the scope, intensity, and fre-
quency of dysfunctional conflict.
Alternatively, rather than reducing the incidence of
conflict, high levels of interorganizational trust may ac-
tually promote conflict. In particular, the “mixed-motive”
nature (Sako 1992) of buyer-supplier relationships creates
tension between cooperation and competition. Interorga-
nizational trust gives exchange partners the confidence to
be open with each other, knowing that information shared
will not be used against them. However, though such
openness may facilitate the discovery of solutions that
increase the size of the collective economic pie, it could
also lead to disagreement over the pie’s distribution. As
additional opportunities for profit are recognized, the in-
centives for exchange partners to attempt to capture a
larger proportion for themselves are raised (Helper and
Levine 1992). The divergence in their goals (e.g., to ap-
propriate as much of the pie as possible for themselves)
may become more salient. As a result, efforts by ex-
change partners to appropriate the gains from the rela-
tionship for themselves are likely to be manifested as con-
flict.
One way of resolving the apparent inconsistency be-
tween arguments for positive and negative effects of trust
on conflict is to view those alternatives in the context of
extended exchange relationships. From that perspective,
higher trust may temporarily be associated with higher
conflict, but over time interfirm disagreements will tend
to lead to either termination or reaffirmation of the rela-
tionship. In the latter case, trust is likely to reduce con-
flict. Based on this line of reasoning, we hypothesize,
H
YPOTHESIS
4. There is a negative relationship be-
tween interorganizational trust and the level of conflict
in the exchange relationship.
Because the boundary spanners are active participants
in the process of managing the exchange relationship,
their interpersonal relationship is likely to affect the level
of conflict (Currall and Judge 1995). Where interpersonal
trust between boundary spanners is high, the parties are
likely to develop solutions that are focused on the prob-
lem at hand rather than on the personalities involved
(Fisher and Ury 1991). Moreover, when interpersonal
trust is high, a boundary spanner is likely to give the
counterpart the benefit of doubt rather than jumping to
conclusions about the other’s motives and intentions. Fur-
ther, as conflict often derives from unexpected actions by
one party, the predictability inherent in high levels of in-
terpersonal trust is likely to be associated with lower lev-
els of conflict.
A contrasting view of the link between interpersonal
trust and conflict is based on the premise that interper-
sonal trust may actually allow conflict to surface without
disruptive consequences. We argue that individuals in-
volved in exchange relationships characterized by trust
may be more likely to confront and resolve disagreements
rather than smoothing them over. This notion is akin to
the idea that boundary spanners are more likely to con-
front trusted counterparts with “harsh truths” than those
they do not trust. In such cases, the parties are able to
accept periodic disagreements without fear of exploita-
tion because the risk that conflict will permanently dam-
age the interpersonal relationship is lower in the presence
of interpersonal trust (Walton and McKersie 1965). Para-
doxically, we expect such confrontations ultimately to re-
duce conflict among boundary spanners by enabling the
individuals to resolve issues in a comprehensive way that
gets at the underlying cause of problems rather than treat-
ing them superficially. As a result, the sum total of con-
flict is lower in a high trust than in a low trust relationship.
The process of resolving core differences may be con-
flictual, but in the end it promotes smoother and more
harmonious interactions in the future.
H
YPOTHESIS
5. There is a negative relationship be-
tween interpersonal trust and the level of conflict in the
exchange relationship.
Costs of Negotiation and Performance. A central
premise of organization economics is that costs are as-
sociated with conducting exchange (Coase 1937). Such
transaction costs stem from the inability of exchange part-
ners to specify contingencies fully in a contract
(Williamson 1985). The efficiency and performance of an
exchange relationship are consequently greatly influ-
enced by the parties’ ability to limit the costs associated
with contracting, in particular the costs of negotiation and
renegotiation. As mentioned earlier, we conceptualize ne-
gotiating costs as the time and effort expended by the
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
146 O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998
exchange partners to arrive at agreements about the dis-
tribution of costs and benefits (Milgrom and Roberts
1992).
We maintain that exchange performance
3
is lowered
when negotiation costs are high because “. . . bargaining
position[s] carry direct costs that the bargainers may wish
they could all avoid. In this regard, the time and energy
so often spent haggling, posturing, and delaying agree-
ments in attempts to influence the terms of the deal are
related wastes” (Milgrom and Roberts 1992, p. 149). The
resources expended in such costly haggling and negoti-
ation take their toll on performance because the exchange
partners are distracted from their business. More specif-
ically, when contingencies arise, such as unexpected costs
not explicitly covered by the contract or unanticipated
design changes, high levels of interorganizational trust
enable the partners to address these contingencies without
resorting to legalistic remedies. For example, rather than
concentrating on assigning blame and debating respon-
sibility for bearing the costs of changes, the parties will
tend to direct their efforts toward determining how best
to reach mutually beneficial solutions.
H
YPOTHESIS
6. There is a negative relationship be-
tween supplier performance and costs of negotiation be-
tween the partners to the exchange.
Conflict and Performance. Disagreements are a nat-
ural aspect of interorganizational exchange and may re-
sult in damaging conflict or may be worked out amicably
with consequences that are largely positive (Anderson
and Naurus 1990, Dwyer et al. 1987). As one industry
participant we interviewed put it:
. . . to maximize return we can only do it through a managed—
not adversarial—relationship. Although some conflict is there,
we respect each others’ business goals.
If conflict is frequent and pervasive, however, the out-
comes may be dysfunctional because of the increased
time and effort spent resolving the conflict, the involve-
ment of other, more senior, members of the organization,
the spill-over of the conflict into other areas, and the like.
In such cases the performance of the supply relationship
tends to decline as the parties’ efforts are directed toward
non-value-enhancing activities.
H
YPOTHESIS
7. There is a negative relationship be-
tween supplier performance and the level of conflict in
the exchange relationship.
Control Variables
Other elements of the transaction context play a role in
the structuring of the dyadic relationship and in the per-
formance of the relationship. In particular, asset specific-
ity, or the extent to which a party’s assets are dedicated
to the transaction, influences dyadic governance structure
by requiring safeguards against appropriation of the as-
sociated rents. Uncertainty in the transaction environment
increases the number of contingencies, creating greater
potential for opportunistic renegotiation of the terms of
the contract. Together, asset specificity and uncertainty
have been theorized to influence the governance struc-
ture, which is the framework within which firms negoti-
ate, execute, and monitor exchange agreements
(Williamson 1985). Prior research has demonstrated that
relationship performance is influenced by a governance
structure that safeguards the firm’s specific assets in the
presence of uncertainty (e.g., Heide and John 1990,
Walker and Weber 1984). Given the extensive history of
research in this area, we control for asset specificity, en-
vironmental uncertainty, and governance form, and do
not hypothesize specific relationships between these con-
structs.
Methods
Data
Data on dyadic exchange relationships of electrical equip-
ment manufacturers and their component suppliers were
gathered in four phases. In phase 1 an original group of
manufacturing firms in the “Electronic and Other Elec-
trical Equipment and Components” industry (SIC code
3600) was constructed. We selected this industry because
it contains a wide range of purchasing arrangements and
provides a sampling frame of adequate size. A list of pur-
chasing managers in the firms who are members of the
National Association of Purchasing Managers (NAPM)
was made available to us by the association. Extensive
semistructured interviews, each 45 minutes to one hour
long, were held with 20 purchasing managers from dif-
ferent firms in the industry.
In phase 2 we identified 1,050 NAPM members who
were eligible to participate in our study on the basis of
the following criteria: (1) their firms purchase compo-
nents, (2) they deal directly with supplier firms, and (3)
they have purchasing relationships with at least six sup-
pliers. These NAPM members received a letter requesting
their participation in our study of buyer-supplier relation-
ships. We also asked purchasing managers to identify a
specific supplier firm that provides their firm with some
key component used in the final product.
To control for potentially confounding effects on re-
lational governance caused by the importance of a sup-
plier and by the amount of purchases made from it, we
used a randomizing procedure (Anderson and Narus
1990). We asked purchasing managers to select their
fourth largest supplier and to identify a specific individual
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 147
(i.e., the individual counterpart) with whom they have
been dealing personally. The procedure of selecting the
fourth largest rather than the largest supplier also miti-
gates social desirability bias, which is sometimes present
in questionnaire research. A total of 153 purchasing man-
agers agreed to participate and sent the requested infor-
mation about a specific supplier.
Testing for Nonresponse Bias. The participation rate
of approximately 15% (i.e., 153/1,050) was somewhat
low and suggested the potential for nonresponse bias. We
therefore conducted a telephone survey of 100 randomly
selected nonparticipants to determine whether there were
any systematic differences between our sample and the
rest of the population. We carried out t-tests for differ-
ences in the means of participating and nonparticipating
firms on certain key variables. No significant differences
in the means were found for the size of firms (t41.10),
length of the business relationship with the supplier com-
pany (t410.81), or satisfaction with the buyer-supplier
relationship (t410.56). Hence, we believe the threat
to internal validity of the results is limited.
Questionnaires. To increase the reliability and valid-
ity of our measures we developed two separate question-
naires on the basis of the semistructured interviews and
previous research: one for the purchasing managers and
another for a second respondent in the purchasing orga-
nization (Bagozzi et al. 1991). The questionnaire for the
second respondent was identical to that for the purchasing
managers with the exception that all items corresponding
to the interpersonal relationship between the purchaser’s
and the supplier’s representatives were eliminated. Ques-
tionnaires were pretested among local purchasing man-
agers in the same industry, who were subsequently ex-
cluded from the sample.
In phase 3 of the data collection, we mailed the ques-
tionnaires to the 153 purchasing managers and to the sec-
ond respondents. In total, questionnaires were mailed to
306 individuals in the buyer organization. In the final
phase of our data collection we implemented Dilman’s
(1978) techniques for maximizing the response rate with
follow-up correspondence, questionnaires, and telephone
calls.
We received a total of 205 responses for a final re-
sponse rate of 67% of individuals eligible and willing to
participate (205/306). Of the 205 completed question-
naires received, 120 were from purchasing managers and
85 were from the second respondent in the purchasing
organization.
Assessing Multiple Informant Competence. We col-
lected data from multiple informants with the aim of ad-
dressing concerns about single-informant bias in research
on interorganizational relations (Kumar et al. 1993, p.
1634). However, that approach poses a further problem
of identifying additional individuals competent to report
on a particular dyadic relationship. We evaluated the
competency of the second informant in the purchasing
organization by including an item designed to measure
specific knowledge of the firm’s business relationship
with the focal supplier. Of the 85 second informants sur-
veyed, 77 (91%) responded to the item “I am familiar
with most aspects of our business relationship with Sup-
plier X” by circling four or higher on a scale from one
(strongly disagree) to seven (strongly agree). Any second
informants responding three or lower were removed from
our dataset and not included in any analyses. Of the 77
qualified second informants retained, 68 had matching
questionnaires from purchasing managers.
Measurement
Scales. Table 1 reports the details of the measurement
items and scales used to operationalize our theoretical
constructs. Where available, we used measurement in-
struments from the literature to develop constructs. Some
items were modified to reflect the specific context of the
study. The Cronbach alpha reliability value for each con-
struct is also reported in Table 1. The reliability values
of the measurement scales all exceed the recommended
value of 0.70 (Nunnally 1978) with the exception of that
for the joint action construct, which is marginal at 0.64.
Details of the development of the trust constructs follow.
Trust: To develop measures of trust at the two levels
of analysis we relied primarily on a measurement instru-
ment created and validated by Rempel et al. (1985) as
modified by Rempel and Holmes (1986). We chose to
base our trust scales on that instrument because it was
designed specifically to tap trust in close, personal rela-
tionships rather than a more general trusting orientation.
However, not all of the Rempel and Holmes items were
applicable to our research context. In particular, we ex-
cluded items measuring a faith element and replaced them
with items designed to measure the fairness component
of trust based on our conceptualization of trust in an ex-
change context. We also eliminated items that did not
apply to both the interorganizational and interpersonal
levels. We further adapted the instrument for use in an
interfirm context by altering the referent of trust (which
was either the organization or the individual counterpart,
referred to as the “contact person”). Last, we ensured that
items corresponding to all three forms of trust—cogni-
tive, behavioral, and emotional—were represented in the
scales, consistent with our conceptualization.
We created parallel instruments to measure trust at both
levels of analysis. The initial item pool for interorgani-
zational and interpersonal trust scales contained the same
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
148 O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998
Table 1 Measurement Instruments
Measures and Items
Internal
Consistency
Reliability (
a
) Source
Interorganizational Trust 0.7664 Adapted from Rempel
and Holmes (1986)
1. Supplier X has always been evenhanded in its negotiations with us.
2. Supplier X may use opportunities that arise to profit at our expense.
a
3. Based on past experience, we cannot with complete confidence rely on Supplier X to keep
promises made to us.
a
4. We are hesitant to transact with Supplier X when the specifications are vague.
a
5. Supplier X is trustworthy.
(1 4strongly disagree, 4 4neither agree nor disagree, 7 4strongly agree)
Interpersonal Trust 0.8799 Adapted from Rempel
and Holmes (1986)
1. My contact person has always been evenhanded in negotiations with me.
2. I know how my contact person is going to act. S/he can always be counted on to act as I
expect.
3. My contact person is trustworthy.
4. I have faith in my contact person to look out for my interests even when it is costly to do
so.
5. I would feel a sense of betrayal if my contact person’s performance was below my
expectations.
(1 4strongly disagree, 4 4neither agree nor disagree, 7 4strongly agree)
Negotiation 0.8824 New Items
1. How easy are negotiations between your business unit and Supplier X over sharing the
burden of costs (not explicitly covered by the contract) when
a. Your business unit requests engineering changes?
b. Supplier X’s raw material costs increase?
(1 4not at all, 4 4somewhat, 7 very much)
2. How quick are negotiations between your business unit and Supplier X over sharing the
burden of costs (not explicitly covered by the contract) when
a. Your business unit requests engineering changes?
b. Supplier X’s raw material costs increase?
(1 4very easy, 4 4somewhat difficult, 7 4very difficult)
Conflict 0.8104 Adapted from Van de
Ven and Ferry (1980)
1. During the past year how often were there significant disagreements between
a. Your business unit and Supplier X?
b. You and the contact person?
(1 4never, 4 4sometimes, 7 4very often)
Supplier Performance 0.7502 New Items
1. Please rate Suppliers X’s performance on fulfilling each of the following goals:
a. Competitive price
b. Timeliness of delivery
c. High quality supply
(1 4very poor, 4 4fair, 7 4excellent)
Asset Specificity 0.7069 Adapted from Heide and
John (1990)
1. Our production system has been tailored to meet the requirements of dealing with Supplier
X.
2. Gearing up to deal with Supplier X requires highly specialized tools and equipment.
3. Our production system has been tailored to using the particular items bought from Supplier
X.
4. We have made significant investments in tools and equipment dedicated to our relationship
with Supplier X.
(1 4strongly disagree, 4 4neither agree nor disagree, 7 4strongly agree)
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 149
Table 1 Continued
Measures and Items
Internal
Consistency
Reliability (
a
) Source
Uncertainty 0.7700 Adapted from
Noordewier, John, and
Nevin (1990)
1. How would you describe the supply of the component purchased from Supplier X
compared to other similar products (not only provided by Supplier X)?
a. Stable availability
a
b. Easy to monitor technological trends
a
c. Stable industry volume
a
d. Accurate sales forecasts
a
(1 4not at all, 4 4somewhat, 7 4very much)
Joint Action 0.6412 New Items
1. Supplier X provides our business unit with cost structure information for the component we
purchase from them.
2. We share with Supplier X our long-term plans for our products.
3. Our business unit helps out Supplier X in whatever ways they ask.
4. Our business unit has been actively pursuing JIT (Just-In-Time) practices with Supplier X.
5. Our business unit relies heavily upon TQM (Total Quality Management) policies to manage
its relationship with Supplier X.
(1 4strongly disagree, 4 4neither agree nor disagree, 7 4strongly agree)
a
Reverse coded.
set of items. However, factor analysis of the two trust
measures revealed that a reduced final set of items pro-
vided measurement properties superior to those of our
earlier conceptualization of trust as containing the three
components of predictability, reliability, and fairness. In
particular, we found that the emotional form of trust
emerged in the interpersonal trust construct but not in the
interorganizational one. Further, the predictability items
dropped out of the interorganizational scale and reliability
did not emerge in the interpersonal scale. The final item
pool contained five items each for interorganizational
trust and interpersonal trust, of which three correspond to
the original items in the Rempel and Holmes (1986) scale.
Of the five items for the construct interorganizational
trust, two items reflected the fairness component of trust,
one item directly assessed interorganizational trust, and
the other two tap the reliability aspect of trust. We dis-
carded items measuring predictability after assessing the
construct for unidimensionality, as mentioned previously.
The five-item scale for interpersonal trust consists of
one item related to predictability, three items related to
fairness, and one item that directly assesses interpersonal
trust. Items measuring the reliability element of interper-
sonal trust were eliminated as a result of the assessment
for unidimensionality.
Negotiation: We operationalized the costs of negotia-
tion with a four-item scale. Two items captured the ease
and speed of negotiations pertaining to engineering
changes and two items measured the ease and speed of
negotiations pertaining to increases in raw materials
costs.
Conflict: Two items measured the degree of conflict
between the two organizations and the level of conflict
between the two boundary-spanning individuals. The
items were adapted from Van de Ven and Ferry (1980).
Performance: We operationalized the performance
construct with a four-item scale reflecting the degree to
which the supplier organization fulfilled the goals of com-
petitive price, timeliness of delivery, high quality supply,
and supplier flexibility (Heide and Stump 1995, Walker
1994).
4
The operationalization is based on the reasoning
that “the less cooperative the supplier is in meeting the
buyers’ needs, the higher the transaction costs the buyer
incurs in trying to achieve its goals in the supply rela-
tionship” (Walker 1994, p. 583). This operationalization
is also consistent with the marketing channels literature
(Heide and Stump 1995).
Asset specificity: Our asset specificity construct was
based on a four-item scale measuring the extent to which
the production system is customized for the exchange
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
150 O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998
partner in terms of investments in specialized tools and
equipment (Heide and John 1988).
Uncertainty: Four items, adapted from Noordewier et
al. (1990), measured the degree of environmental uncer-
tainty in terms of industry volume, technological trends
of the market, stable availability, and accuracy of fore-
casts.
Joint Action: The joint action construct was measured
on a five-item scale reflecting the degree of inter-
penetration of organizational boundaries and the extent
to which the exchange activities are carried out in a co-
operative and coordinated manner (Heide and John
1990).
Modeling Approach. We used the estimation proce-
dure of LISREL 7 (Jo¨reskog and So¨rbom 1989) to con-
struct a structural equation model to test our hypotheses.
The LISREL technique has the advantage over standard
regression analysis of explicitly considering the measure-
ment error in the indicators and simultaneously estimat-
ing a system of structural equations. However, research-
ers have noted the difficulty of fitting LISREL models
with a large number of items per latent variable (Williams
and Hazer 1986). Specifically, as the number of items and
parameters increases, the model “can be unwieldy be-
cause of likely high levels of random error in typical
items and the many parameters that must be estimated”
(Bagozzi and Heatherton 1994, p. 43). Because of the
complexity of our conceptual model and the relatively
large number of manifest variables, particularly for the
trust constructs, we used the partial aggregation model
described by Bagozzi and Heatherton (1994). The partial
aggregation approach addresses this modeling problem
by consolidating the manifest items of a latent variable
into a smaller number of composite indicators.
To construct the composite indicators, we first evalu-
ated each construct for unidimensionality using factor
analysis (Williams and James 1994). We then ranked
each construct’s items according to their loadings and as-
signed them to one of two indicators. The mean of the
items assigned to each indicator was used as the value for
the indicator. The indicators were then loaded onto the
respective latent variables in the measurement model. All
constructs were modeled with the partial aggregation
modeling approach, with the exception of conflict which
contained only two items.
Means, standard deviations, and zero-order correla-
tions among composite indicators are reported in Table
2, and the loadings of the composite indicators on latent
variables are reported in Table 3. We handled possible
non-normality by using the maximum likelihood esti-
mation procedures of LISREL 7, which “have been found
to be robust against non-normality” (Jo¨reskog and
So¨rbom 1989, p. 21) even with sample sizes as low as 50
(Anderson and Gerbing 1988).
Convergent and Discriminant Validity. Convergent
validity is “. . . the degree to which multiple attempts to
measure the same concept by different methods are in
agreement” (Phillips 1981, p. 399; Campbell and Fiske
1959). In organizational research, “different methods” of-
ten is taken to mean obtaining reports from independent,
qualified respondents
5
(Bagozzi and Phillips 1982). We
assessed convergent validity in our study by examining
the correlation between interorganizational trust as re-
ported by the purchasing manager and by the second re-
spondent in the purchasing organization. As shown in
Figure 2 (
f
13
), the correlation between interorganiza-
tional trust measured by the two respondents (methods)
is 0.706 (t44.27, p,0.01). The positive and highly
significant correlation provides strong evidence for the
convergent validity of our interorganizational trust con-
struct.
Drawing on Campbell and Fiske’s (1959) discussion of
construct validity, Bagozzi (1993, p. 54) defines discrim-
inant validity as “the degree to which measures of differ-
ent concepts are distinct. The notion is that if two or more
concepts are unique, then valid measures of each should
not correlate too highly.” We tested for discriminant va-
lidity by comparing a model with the correlation between
the two constructs constrained to equal one with an un-
constrained model. A significantly lower chi-square value
for the model with the unconstrained correlation provides
support for discriminant validity (Jo¨reskog 1971).
We conceptualized interpersonal and interorganiza-
tional trust as theoretically distinct but related constructs.
Consequently, we were primarily concerned with assess-
ing the discriminant validity of those constructs.
6
Tests
of a model with a correlation between constructs con-
strained to unity yielded v
2
(7 df) 417.70, p,0.013,
and a model with unconstrained correlation yielded the
following: v
2
(6 df) 48.05, p,0.234. The difference in
chi-square (1 df) of 9.65 is statistically significant at p,
0.01, providing strong support for the discriminant valid-
ity of interpersonal and interorganizational trust.
Analysis
To test our hypotheses we specified two models using the
maximum likelihood estimation procedure in LISREL: a
trust correlation model and a structural model. The trust
correlation model (see Figure 2) tested H1, predicting a
positive link between trust at the two levels of analysis,
with data from both the primary and secondary respon-
dents of the purchasing organization. The significance of
the correlation coefficients estimated by the LISREL pro-
cedure, and reflected in the
f
parameter, provides a test
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 151
Table 2 Descriptive Statistics and Zero-Order CorrelationsConstructs
Mean S.D. 123456789101112131415
1 Interorganizational trust 1 5.888 1.189
2 Interorganizational trust 2 5.513 1.013 0.578
3 Interpersonal trust 1 5.628 1.305 0.392 0.166
4 Interpersonal trust 2 5.136 1.344 0.402 0.177 0.792
5 Negotiation 1 2.923 1.343 10.517 10.424 10.083 0.029
6 Negotiation 2 2.560 1.171 10.440 10.376 10.019 10.014 0.741
7 Conflict 1 2.442 1.327 10.439 10.413 10.245 10.214 0.413 0.422
8 Conflict 2 1.845 1.035 10.306 10.375 10.125 10.141 0.315 0.432 0.655
9 Performance 1 5.621 0.987 0.573 0.495 0.263 0.251 10.291 10.296 10.352 10.159
10 Performance 2 5.746 0.972 0.575 0.523 0.376 0.387 10.305 10.316 10.371 10.164 0.757
11 Asset specificity 1 3.080 1.530 10.193 10.172 10.038 10.052 0.149 0.241 0.121 0.030 10.180 10.165
12 Asset specificity 2 3.170 1.421 10.209 10.165 0.013 0.091 0.117 0.192 0.109 0.039 10.184 10.149 0.676
13 Uncertainty 1 2.813 1.159 10.203 10.223 10.102 10.086 0.128 0.081 0.113 10.009 10.230 10.265 0.064 0.135
14 Uncertainty 2 2.894 1.202 10.235 10.173 10.143 10.142 0.079 0.024 0.214 10.028 10.205 10.195 10.139 10.034 0.654
15 Joint action 1 5.038 1.289 0.049 0.026 0.079 0.118 10.040 10.046 10.088 0.076 0.171 0.256 0.184 0.364 10.183 10.215
16 Joint action 2 4.742 1.289 0.102 0.174 0.077 0.162 10.101 10.148 10.144 10.026 0.239 0.319 0.194 0.292 10.144 10.264 0.611
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
152 O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998
Table 3 Parameter Estimates for Structural Model
(Figure 3)
Parameter
Standardized
Solution t-value
kx
1
0.844 (fixed parameter)
kx
2
0.686 7.430**
kx
3
0.815 (fixed parameter)
kx
4
0.972 7.111**
kx
5
0.603 (fixed parameter)
kx
6
1.119 3.707**
kx
7
0.822 (fixed parameter)
kx
8
0.793 4.260**
ky
1
0.902 (fixed parameter)
ky
2
0.822 8.297**
ky
3
0.909 (fixed parameter)
ky
4
0.720 5.734**
ky
5
0.839 (fixed parameter)
ky
6
0.737 5.350**
ky
7
0.823 (fixed parameter)
ky
8
0.898 9.438**
b
1
0.264 1.388*
b
2
0.034 0.264*
b
3
0.261 2.829**
c
1
10.810 15.865**
c
2
10.593 14.465**
c
3
0.359 3.231**
c
4
0.017 0.148
c
5
0.395 3.840**
c
6
10.345 12.829**
c
7
1.011 3.716**
c
8
10.066 10.477**
**p ,0.01.
*p ,0.05.
of H1. The structural model (see Figure 3) tested H2
through H7, which link trust with performance, mediated
by costs of negotiation and level of conflict. Data from
the purchasing manager respondents were used to esti-
mate the path coefficients that are reflected in the respec-
tive cand bvalues. The statistical significance of the path
coefficients provides tests of the hypotheses.
Results
Trust Correlation Model. The model consists of three
latent variables: interpersonal and interorganizational
trust as reported by the purchasing manager and interor-
ganizational trust as reported by the second respondent in
the purchasing organization. The degree of overall fit be-
tween the actual and predicted covariances among vari-
ables of a model is reflected in a series of goodness-of-
fit measures. This model had a chi-square value (6 df) of
8.05 (p40.234). Standard fit indices, as reported in Fig-
ure 2, indicate a satisfactory fit of the model with the
data.
7
The predicted relationship (H1) is tested by the strength
of the correlation between interpersonal trust reported by
the purchasing manager and interorganizational trust re-
ported by the second respondent (see Figure 2,
f
23
). In-
terpersonal trust and interorganizational trust are highly
correlated (
f
23
40.442; t42.74; p,0.01), providing
support for H1. As a second test of H1, we specified a
correlation path between interpersonal and interorgani-
zational trust both as reported by the purchasing manager
(see Figure 2,
f
12
). Once again, the correlation is positive
and statistically significant (
f
12
40.546; t43.18; p,
0.01), adding further support for H1.
Structural Model. The structural model (see Figure 3)
specifies the relationships between trust and performance,
mediated by conflict and negotiating costs, with controls
for transaction cost variables. Following Venkatraman
(1989), we tested for the mediated relationships by spec-
ifying an additional direct path along with the mediation
paths. In other words, to test the prediction that the in-
terorganizational trust-performance link is mediated by
negotiation costs, we specified three paths: a link between
interorganizational trust and negotiation, a further path
between negotiation costs and performance, and a third
direct path between interorganizational trust and perfor-
mance. The same pattern of structural paths was specified
for each of the remaining mediated relationships hypoth-
esized. This model had a chi-square value (86 df) of 94.82
(p40.242). Standard fit indices, as reported in Figure
3, suggest a good fit of the model with the data.
Hypothesis 2, predicting a negative relationship be-
tween interorganizational trust and costs of negotiation,
is strongly supported (c
1
410.810; t415.87; p,
0.01). Hypothesis 3 predicting a negative relationship be-
tween interpersonal trust and costs of negotiation and is
not supported. Rather, the relationship is both positive
and statistically significant (c
3
40.359; t43.23; p,
0.10). As predicted in H4, a negative relationship was
observed between interorganizational trust and conflict
(c
2
410.593; t414.47; p,0.01). Contrary to H5,
a negative relationship was not observed between inter-
personal trust and conflict. The relationship is positive,
although not statistically significant (c
4
40.017; t4
0.148; n.s.). Hypothesis 6, predicting a negative relation-
ship between supplier performance and costs of negoti-
ation, is not supported (b
1
40.264; t41.39; n.s.), nor
is H7, predicting a negative relationship between supplier
performance and conflict (b
2
40.034; t40.264; n.s.).
The direct path that we specified from interorganizational
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 153
Figure 2 Correlational Model of Trust
trust to performance is positive and statistically signifi-
cant (c
7
41.011; t43.716; p,0.01), but the direct
path from interpersonal trust to performance is not sig-
nificant (c
8
410.066; t410.477; n.s.).
Discussion
In this paper, we examined trust-based relational ex-
change to explicate a mechanism for the connection be-
tween trust and performance, and to distinguish between
the consequences of interorganizational and interpersonal
trust. Our results broadly support the thesis that trust in
relational exchange influences negotiation processes and
performance, although the precise nature of the link is
somewhat different from what we initially proposed.
Eased negotiation and reduced conflict are indeed out-
comes of trust, yet the pattern of findings differs sharply
across levels. Interorganizational trust emerges as the
overriding driver of exchange performance, negotiation,
and conflict, whereas interpersonal trust exerts little direct
influence on those outcomes. Nevertheless, interpersonal
trust may also matter through its institutionalizing effects
on interorganizational trust.
Key Findings
Trust at Two Levels of Analysis. As hypothesized, we
found interorganizational trust and interpersonal trust to
be related, although empirically and theoretically distinct.
Moreover, that finding is not merely an artifact of single-
source bias given that we used data from two different
respondents in the buying organization to test the rela-
tionship. From the perspective of the boundary-spanning
individual in the buying organization, this means that the
more one trusts the supplier representative with whom
one deals, the more one’s organization trusts the supplier
organization. We see this relationship as operating in the
opposite direction as well, suggesting mutually reinforc-
ing effects of trust at the two levels. The implication of
these reciprocal effects is that, although interorganiza-
tional trust may appear to be the more important influence
in relational exchange, interpersonal trust must also be
considered for its effects on interorganizational trust. At
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
154 O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998
Figure 3 Structural Model of Relational Exchange
the same time, interorganizational research examining
trust solely at the individual level may be missing im-
portant institutionalized effects of interorganizational
trust. Simply aggregating interpersonal trust as a proxy
for interorganizational trust ignores the influence of social
context in the form of individuals’ interactions (Coleman
1990) and organizational rules (Sitkin and Roth 1993)
that constrain and orient its members.
Perhaps our most striking finding is that interpersonal
and interorganizational trust operate quite differently
within relational exchange. Interorganizational trust is as-
sociated strongly with lowered costs of negotiation and
conflict, but interpersonal trust is not related to conflict
and showed a seemingly anomalous positive association
with negotiation costs. These findings imply that the ef-
fects of trust in the interorganizational context are distinct
at the individual and organizational levels of analysis.
Interorganizational Trust and Performance. Our hy-
potheses about the links between trust and performance
at both levels of analysis are partially supported. Al-
though interorganizational trust is related strongly to
costs of negotiation and the level of conflict in the dyad,
the relationship between trust and performance is not me-
diated by negotiation costs and conflict, contrary to what
we had hypothesized. Further, although we did not hy-
pothesize a direct effect of trust on performance, our em-
pirical results in fact reveal a direct link between inter-
organizational trust and performance, though not
between interpersonal trust and performance.
Given their strong theoretical rationale in the organi-
zational economics literature, the lack of support for the
positive performance effects of eased negotiations and
reduced conflict is somewhat surprising. Nonetheless, be-
fore concluding that the trust-performance relationship is
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 155
Figure 4 Surface Plot of Negotiation Costs with Interorga-
nizational and Interpersonal Trust
a direct rather than a mediated one, we note that eased
negotiations and reduced conflict resulting from interor-
ganizational trust may in fact be enabling conditions al-
lowing exchange partners to pursue a variety of bilateral
governance mechanisms—such as exchange of personnel
and shared decision making—that lead to improved per-
formance.
The preceding point notwithstanding, we believe the
strong direct relationship between interorganizational
trust and performance is a noteworthy and meaningful
finding. It indicates that the performance of interfirm ex-
change is in fact associated with the level of interorga-
nizational trust. Such a link has often been suggested in
previous work on interfirm exchange, but relatively little
empirical support for this relationship exists. A key im-
plication of this finding is that firms in exchange rela-
tionships may derive competitive advantage from rela-
tionships imbued with high levels of interorganizational
trust. Interestingly, the basis for the performance en-
hancement does not appear to be based on efficiencies
gained from eased negotiation processes. Rather, we
speculate that the enhancement of transaction value
(Zajac and Olsen 1993)—such as cooperation in the ex-
ploration of new information and coordination technolo-
gies, new market opportunities, and product and process
innovation—may account for the link between interor-
ganizational trust and exchange performance. As one of
the industry participants we interviewed put it:
It is necessary to get away from a “remedy-oriented” contract
which includes all contingencies. We need to focus on what can
go right instead of what can go wrong.
Our findings rule out negotiation processes as a mediating
link, but future research should ascertain the role of these
and other value-enhancing exchange processes.
Although the entire chain of proposed theoretical re-
lationships is not supported, our hypotheses about the link
between interpersonal and interorganizational trust and
the links between trust and the costs of negotiation and
conflict are partially supported. These findings are con-
sistent with transaction-cost-based explanations for the
role of trust in exchange (Bromiley and Cummings 1995),
while also emphasizing the positive outcomes from in-
terfirm exchange (Zajac and Olsen 1993). Similarly, the
negative relationship between interorganizational trust
and the level of conflict in the dyad can be attributed to
the willingness of the parties to grant one another greater
leeway when differences arise. In low trust situations,
suspicion of the partner’s motives may result in dysfunc-
tional conflict and impede effective exchange processes.
Taken together, the findings support the conclusion that
interorganizational trust acts to smooth negotiation pro-
cesses and thereby reduce the transaction costs of inter-
firm exchange.
The Role of Interpersonal Trust. We found no support
for our hypotheses relating interpersonal trust to the costs
of negotiation or to the level of conflict in the exchange
relationship. In fact, the finding that interpersonal trust
related positively to the costs of negotiation is contrary
to our initial hypothesis. These results are counter-
intuitive given that research in the boundary-spanning lit-
erature has indicated the importance of boundary-
spanning individuals in managing the interfirm relation-
ship (Friedman and Podolny 1992).
8
To investigate the apparent anomaly of negotiation
costs increasing with interpersonal trust, we constructed
a three-dimensional surface plot of negotiation costs
against interpersonal trust, controlling for interorganiza-
tional trust (see Figure 4). Although the pattern is not
readily apparent, the surface of the plot is saddle-shaped
and rises to the maximum in both the foreground and the
background. Examination of the plot confirms the in-
verted U-shaped relationship between negotiation costs
and interpersonal trust. Further, the third dimension of the
plot, interorganizational trust, provides a possible expla-
nation for the increasing part of the U-shaped curve. Spe-
cifically, where interpersonal trust is low, we find that
interorganizational trust remains high. High interorgani-
zational trust appears to compensate for low interpersonal
trust and thus to explain the low level of negotiation costs.
At the same time, the high negotiation costs associated
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
156 O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998
with intermediate levels of interpersonal trust are accom-
panied by lower levels of interorganizational trust as in-
dicated by the high point in the foreground, which cor-
responds to interorganizational trust at its lowest level and
both interpersonal trust and negotiation costs at their
highest levels. This pattern of findings indicates that even
though the individuals across the dyad may not trust each
other, as long as the institutionalized structures accom-
panying high interorganizational trust are in place, ne-
gotiating costs will be kept down. Further, high interor-
ganizational trust and low interpersonal trust can coexist
in the same relationship given that boundary-spanning in-
dividuals come and go, whereas the institutionalized
structures and processes accompanying interorganiza-
tional trust are more stable and enduring (Ring and Van
de Ven 1994).
We conducted an additional analysis to confirm the
curvilinear relationship between interpersonal trust and
negotiation costs by regressing negotiation costs on in-
terpersonal trust and its squared term. The results show a
positive and significant coefficient for the main term, con-
sistent with the LISREL analysis, and a negative and sig-
nificant coefficient for the squared term.
9
The negative
sign for the squared term indicates an inverted U-shaped
relationship between interpersonal trust and negotiation
costs. In other words, the costs of negotiation appear to
be lowest at both low and high levels of interpersonal
trust and are highest at intermediate levels of interper-
sonal trust representing the inflection point of the inverted
U-shaped curve. We calculated the inflection point of the
curve to be at 4.30 for interpersonal trust, suggesting that
negotiation costs increase with increasing interpersonal
trust up to the value 4.30, after which they decline.
Those observations are strong indications that inter-
personal trust by itself is insufficient for lowering nego-
tiation costs. Nevertheless, interpersonal trust is by no
means unimportant. At the point where interorganiza-
tional trust is at its maximum, negotiation costs arehigher
when interpersonal trust is low, as indicated by the rela-
tively high point in the background. From that point down
the upper right edge of the surface, increasing interper-
sonal trust clearly corresponds to lowering negotiation
costs. This observation makes it apparent that interper-
sonal trust plays a distinct, though subordinate, role in
affecting the costs of negotiation when examined in con-
junction with interorganizational trust.
Implications
In comparison with interpersonal trust, interorganiza-
tional trust emerges as the dominant influence on ex-
change processes and outcomes. The pattern of results
suggests that institutionalized practices and routines for
dealing with a partner organization, as captured by inter-
organizational trust, transcend the influence of the indi-
vidual boundary spanner. In considering the role of trust
in relational exchange, we argue that firms must recog-
nize the impersonal (Shapiro 1987) structures, processes,
and routines that create a stable context within which in-
terpersonal trust can develop and persist (Heide and
Miner 1992, Parkhe 1993). The stability of interfirm ex-
change is not created and maintained solely by boundary
spanning individuals, but rather is institutionalized in the
interorganizational relationship. As one purchasing man-
ager we interviewed asserted:
Even though I may leave, the relationship [between our firms]
will continue since what we’ve been doing goes beyond one or
two people. Over the years, there are a host of people who have
worked together.
We do not by any means suggest that the social com-
ponent of exchange is unimportant. On the contrary, our
data indicate that interpersonal trust, in conjunction with
interorganizational trust, plays a unique role in relational
exchange, over and above the effects of governance struc-
ture. However, we suggest that when exchange is carried
out between organizations with an institutionalized pat-
tern of dealings, the interorganizational context becomes
more prominent.
Limitations and Suggestions for Future Research
Our study makes important theoretical and empirical con-
tributions to the literature on the role of trust in interor-
ganizational relations. However, the validity of the im-
plied causal links of our model is limited by the
cross-sectional nature of our research design. We there-
fore encourage longitudinal research on the development
and consequences of trust in interorganizational ex-
change. Further investigation of our findings based on in-
depth case studies would also enhance our understanding
of how interorganizational trust builds up over time and
its dynamic relationship with interpersonal trust.
In our study, we inferred from a specific buyer-supplier
setting to interorganizational relationships more gener-
ally. Clearly replication of our findings in other interfirm
settings, such as joint ventures and strategic alliances, is
needed to establish their external validity. Further, wesee
an interesting extension of the research in cross-national
settings where cultural differences may alter the out-
comes of trust in relational exchange.
We also encourage more research specifically exam-
ining the “downside” of trust in interorganizational rela-
tions. Much has been made of the positive effects of trust,
but there is clearly potential for abuse of trust by ex-
change partners. Barnes (1981), for example, suggests
that excessive levels of trust can result in exclusive and
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 157
dysfunctional reliance on soft rather than hard data to
inform decision making.
10
Finally, our understanding of
the phenomenon of interorganizational trust and its con-
sequences would be greatly enhanced by a study of con-
textual antecedents of trust—such as the extent to which
exchange partners share similar organizational structures,
policies, and mindsets or cultures—and its immediate
outcomes such as negotiating costs, conflict, and rela-
tionship performance.
Concluding Remarks
Overall, our research makes several contributions to re-
search on relational exchange by examining the nature of
trust in buyer-supplier dyads. Through the use of multiple
organizational respondents, our study recognizes the dis-
tinction between interorganizational and interpersonal
trust. Although associated with each other, these con-
structs are shown to relate differently to costs of negoti-
ation and conflict in the interfirm relationship. Further,
enhanced supplier performance, lowered costs of nego-
tiation, and reduced conflict are shown to be related to
high levels of interorganizational trust.
Acknowledgments
A preliminary version of this paper, entitled “With a Little Help from
My Friends: Exploring the Links Between Interpersonal and Interor-
ganizational Trust,” was presented at the Academy of Management
Meetings, Dallas, in August 1994. The authors are most grateful to Phil
Bromiley, Larry Cummings, George John, Sue McEvily, Tom Murtha,
Peter Ring, N. Venkatraman, and Sri Zaheer for comments and sug-
gestions on previous versions of the paper. Clearly, those individuals
are not responsible for its contents. The research was supported by
grants to the first author from the International Program Development
Office of the Carlson School of Management, University of Minnesota,
and from the Graduate School Grants-in-Aid program of theUniversity
of Minnesota. Both are gratefully acknowledged. Iris Leibowitz Dori
provided admirable research assistance. The National Association of
Purchasing Managers kindly provided access to membership data.
Endnotes
1
Whereas reliability includes an element of trust as competence (Barber
1983), predictability simply refers to the degree of consistency in in-
tended behavior.
2
Clearly, as one reviewer points out, contextual factors other than trust
also may affect negotiation costs, such as the extent to which exchange
partners have similar organizational structures, policies, and mindsets
or cultures.
3
We define exchange performance as the extent to which the supplier
has fulfilled the buyer’s requirements in terms of price, timeliness of
delivery, input quality, and supplier flexibility.
4
We believe that in buyer-supplier relationships, though there may be
a degree of congruence about what the goals of the relationship are,
the buyer organization is the final arbiter of the extent to which ex-
change goals have been met to a satisfactory degree.
5
There is a well-established precedent for this approach in the orga-
nizational literature. Most importantly, Bagozzi and Phillips (1982)
explicitly detail the use of the multitrait-multimethod (MTMM) matrix
with two informants as different methods. As those authors indicate,
“[i]nformants one and two are treated as methods” (p. 469).
6
The test was carried out with data from the same respondent, thereby
providing a stronger test of discriminant validity than using data from
two different respondents for interpersonal and interorganizational
trust, respectively. The latter approach is considered a less stringent
test because method and trait variance are confounded.
7
For each of the models estimated, we report four standard fit indices.
The goodness of fit index (GFI) reflects the relative amount of variance
and covariance jointly explained by the model. The adjusted goodness
of fit index (AGFI) is the same as the GFI, but adjusts for the number
of degrees of freedom in the model. The normed fit index (NFI)
(Bentler and Bonett 1980) represents the point at which the model
being evaluated falls on a scale from a null model (specifying mutual
independence among indicators) to a perfect fit. The comparative fit
index (CFI) (Bentler 1990) is similar to the NFI, but corrects for small
sample size by subtracting the degrees of freedom from their corre-
sponding chi-square values. Each of these indices range from zero to
1.00, with values closer to 1.00 indicating a good fit. A commonly
accepted rule of thumb is that a fit index should be greater than 0.90.
8
In order to verify that these results are not a methodological artifact
we performed an additional analysis. We found a near identical pattern
of results using partial correlation analyses which is analogous to mul-
tiple regression analysis. Specifically, controlling for interorganiza-
tional trust we found that the partial correlation coefficient between
interpersonal trust and conflict while negative, was nonsignificant (par-
tial correlation coefficient 410.09, p40.16) and the partial corre-
lation coefficient between interpersonal trust and costs of negotiation
was positive and marginally significant (partial correlation coefficient
40.15, p40.066). This pattern of results is quite consistent with
those that we obtained using LISREL.
9
The standardized beta coefficient for the squared term is 11.273 (t-
value 412.85; p,0.01).
10
We are indebted to an anonymous reviewer for suggesting this point.
References
Anderson, E. and B. Weitz (1989), “Determinants of Continuity in
Conventional Industrial Channel Dyads,” Marketing Science, 8,
4, 310–323.
Anderson, J. C. and D. W. Gerbing (1988), “Structural Equation Mod-
eling in Practice: A Review and Recommended Two-Step Ap-
proach,” Psychological Bulletin, 103, 411–423.
—— and J. A. Narus (1990), “A Model of Distributor Firm and Man-
ufacturer Firm Working Partnership,” Journal of Marketing, 54,
January, 42–58.
Arrow, K. (1974), The Limits of Organizations, New York: Norton.
Bagozzi, R. P. (1993), “Assessing Construct Validity in Personality
Research: Applications to Measures of Self-Esteem,” Journal of
Research and Personality, 27, 1, 49–87.
—— and T. F. Heatherton (1994), “A General Approach to Repre-
senting Multifaceted Personality Constructs: Application to State
Self-Esteem,” Structural Equation Modeling, 1, 35–67.
—— and L. W. Phillips (1982), “Representing and Testing Organi-
zational Theories: A Holistic Construal,” Administrative Science
Quarterly, 27, 459–489.
——, Y. Yi, and L. W. Phillips (1991), “Assessing Construct Validity
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
158 O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998
in Organizational Research,” Administrative Science Quarterly,
36, 421–458.
Barber, B. (1983), The Logic and Limits of Trust, New Brunswick, NJ:
Rutgers University Press.
Barnes, L. B. (1981), “Managing the Paradox of Organizational Trust,”
Harvard Business Review, 59, 107–116.
Barney, J. B. and M. H. Hansen (1995), “Trustworthiness as a Source
of Competitive Advantage,” Strategic Management Journal, 15,
Special Issue, 175–190.
Bentler, P. M. (1990), “Comparative Fit Indexes in Structural Models,”
Psychological Bulletin, 107, 238–246.
—— and D. G. Bonett, (1980), “Significance Tests and Goodness of
Fit in The Analysis of Covariance Structures,” Psychological Bul-
letin, 88, 588–606.
Bradach, J. L. and R. G. Eccles (1989), “Price, Authority, and Trust:
From Ideal Types to Plural Forms,” Annual Review of Sociology,
15, 97–118.
Bromiley, P. and L. L. Cummings (1995), “Transaction Costs in Or-
ganizations with Trust,” in R. Bies, B. Sheppard, and R. Lewicki
(Eds.), Research on Negotiation in Organizations, Greenwich,
CT: JAI Press.
Campbell, D. T. and D. W. Fiske (1959), “Convergent and Discrimi-
nant Validation by the Multitrait-Multimethod Matrix,” Psycho-
logical Bulletin, 56, 81–105.
Coase, R. (1937), “The Nature of the Firm,” Economica, 4, 386–405.
Coleman, J. S. (1990), The Foundations of Social Theory, Cambridge,
MA: Harvard University Press.
Corazzini, R. (1977), “Trust as a Complex Multi-Dimensional Con-
struct,” Psychological Reports, 40, 75–80.
Cummings, L. L. and P. Bromiley (1996), “The Organizational Trust
Inventory (OTI): Development and Validation,” in R. N. Kramer
and T. R. Tyler (Eds.), Trust in Organizations, Thousand Oaks,
CA: Sage.
Currall, S. C. and T. A. Judge (1995), “Measuring Trust Between Or-
ganizational Boundary Role Persons,” Organizational Behavior
and Human Decision Processes, 64, 151–170.
Dilman, D. A. (1978), Mail and Telephone Surveys: The Total Design
Method, New York: Wiley.
Dore, R. (1983), “Goodwill and the Spirit of Market Capitalism,” Brit-
ish Journal of Sociology, 34, 459–482.
Dwyer, F. R., P. H. Schurr, and S. Oh (1987), “Developing Buyer-
Seller Relationships,” Journal of Marketing, 51, April, 11–27.
Fisher, R. and W. Ury (1991), Getting To Yes, New York: Penguin
Books.
Friedman, R. A. and J. Podolny (1992), “Differentiation of Boundary
Spanning Roles: Labor Negotiations and Implications for Role
Conflict,” Administrative Science Quarterly, 37, March, 28–47.
Giddens, A. (1979), Central Problems in Social Theory: Action, Struc-
ture and Contradiction in Social Analysis, Berkeley, CA: Uni-
versity of California Press.
Granovetter, M. (1985), “Economic Action and Social Structure: The
Problem of Embeddedness,” American Journal of Sociology, 91,
3, 481–510.
Gulati, R. (1995), “Does Familiarity Breed Trust? The Implications of
Repeated Ties for Contractual Choice in Alliances,” Academy of
Management Journal, 38, 85–112.
Heide, J. B. and G. John (1988), “The Role of Dependence-Balancing
in Safeguarding Transaction-Specific Assets in Conventional
Channels,” Journal of Marketing, 52, 20–35.
—— and —— (1990), “Alliances in Industrial Purchasing: The De-
terminants of Joint Action in Buyer-Supplier Relationships,”
Journal of Marketing Research, 27, February, 24–36.
—— and A. S. Miner (1992), “The Shadow of the Future: Effects of
Anticipated Interaction and Frequency of Contact on Buyer-Seller
Cooperation,” Academy of Management Journal, 35, 2, 256–291.
—— and R. L. Stump (1995), “Performance Implications of Buyer-
Supplier Relationships in Industrial Markets: A Transaction Cost
Explanation,” Journal of Business Research, 32, 57–66.
Helper, S. and D. Levine (1992), “Long-Term Supplier Relations and
Product Market Structure,” Journal of Law Economics and Or-
ganization, 8, 561–581.
John, G. (1984), “An Empirical Investigation of Some Antecedents of
Opportunism in a Marketing Channel,” Journal of Marketing Re-
search, 21, 278–89.
Jo¨reskog, K. G. (1971), “Statistical Analysis of Congeneric Tests,”
Psychometrika, 36, 109–133.
—— and D. So¨rbom, (1989). LISREL 7: A Guide to the Program and
Applications (2nd edition), Chicago, IL: SPSS.
Katz, D. and R. Kahn (1978), The Social Psychology Of Organizations
(revised edition), New York: Wiley.
Kumar, N., L. W. Stern, and J. C. Anderson (1993), “Conducting In-
terorganizational Research Using Key Informants,” Academy of
Management Journal, 36, 6, 1633–1651.
Lewis, J. D. and A. J. Weigert (1985a), “Trust as a Social Reality,”
Social Forces, 63, 966–985.
—— and —— (1985b), “Social Atomism, Holism, and Trust,” Socio-
logical Quarterly, 26, 455–471.
Luhmann, N. (1979), Trust and Power, New York: Wiley.
Macauley, S. (1963), “Non-contractual Relations in Business,” Amer-
ican Sociological Review, 28, 55–67.
Macneil, I. R. (1980), The New Social Contract, New Haven, CT: Yale
University Press.
Malmgren, H. (1961), “Information, Expectation, and the Theory of
the Firm,” Quarterly Journal of Economics, 75, 399–421.
Milgrom, P. and J. Roberts (1992), Economics, Organization and Man-
agement, Englewood Cliffs, NJ: Prentice Hall.
Noordewier, T. G., G. John, and J. R. Nevin (1990), “Performance
Outcomes of Purchasing Arrangements in Industrial Buyer-
Vendor Relationships,” Journal of Marketing, 54, October, 80–
93.
Nunnally, J. (1978), Psychometric Theory, New York: McGraw Hill.
Parkhe, A. (1993), “Strategic Alliance Structuring: A Game Theoretic
and Transaction Cost Examination of Interfirm Cooperation,”
Academy of Management Journal, 36, 4, 794–829.
Pfeffer, J. (1982), Organizations and Organization Theory, Boston,
MA: Pitman.
Phillips, L. W. (1981), “Assessing Measurement Error in Key Infor-
mant Reports: A Methodological Note on Organizational Anal-
ysis in Marketing,” Journal of Marketing Research, 18, 395–415.
Rempel, J. K. and J. G. Holmes (1986), “How Do I Trust Thee?”
Psychology Today, February, 28–34.
——, —— and M. P. Zanna (1985), “Trust in Close Relationships,”
Journal of Personality and Social Psychology, 49, 95–112.
Ring, P. S. and A. H. Van de Ven (1992), “Structuring Cooperative
AKBAR ZAHEER, BILL McEVILY, AND VINCENZO PERRONE Does Trust Matter?
O
RGANIZATION
S
CIENCE
/Vol. 9, No. 2, March–April 1998 159
Relationships Between Organizations,” Strategic Management
Journal, 13, 483–498.
—— and —— (1994), “Developmental Processes of Cooperative In-
terorganizational Relationships,” Academy of Management Jour-
nal, 19, 90–118.
Rotter, J. B. (1971), “Generalized Expectancies of Interpersonal Trust,”
American Psychologist, 26, 443–452.
Rousseau, D. M. (1985), “Issues of Level in Organizational Research,”
in L. L. Cummings and B. M. Staw (Eds.), Research in Organi-
zational Behavior, 7, Greenwich, CT: JAI Press, 1–37.
Sako, M. (1992), Prices, Quality, and Trust: Interfirm Relations in
Britain and Japan, New York: Cambridge University Press.
Shapiro, S. P. (1987), “The Social Control of Impersonal Trust,” Amer-
ican Journal of Sociology, 93, 3, 623–658.
Sitkin, S. B. and N. L. Roth (1993), “Explaining the Limited Effec-
tiveness of Legalistic ‘Remedies’ for Trust/Distrust,” Organiza-
tion Science, 4, 3, 367–392.
Smith, K. G., S. J. Carroll, and S. J. Ashford (1995), “Intra- and Inter-
organizational Cooperation: Toward a Research Agenda,” Acad-
emy of Management Journal, 38, 7–23.
Van de Ven, A. H. and D. L. Ferry (1980), Measuring and Assessing
Organizations, New York: Wiley.
Venkatraman, N. (1989), “Strategic Orientation of Business Enter-
prises: The Construct, Dimensionality and Measurement,” Man-
agement Science, 35, 8, 942–962.
Walker, G. (1994), “Asset Choice and Supplier Performance in Two
Organizations—US and Japanese,” Organization Science, 5, 4,
583–593.
—— and D. Weber (1984), “A Transaction Cost Approach to Make
Versus Buy Decisions,” Administrative Science Quarterly, 29,
373–391.
Walton, R. E. and R. B. McKersie (1965), A Behavioral Theory of
Labor Negotiations, New York: McGraw-Hill.
Williams, L. J. and J. T. Hazer (1986), “Antecedents and Consequences
of Satisfaction and Commitment in Turnover Models: A Reanal-
ysis Using Latent Variable Structural Equation Methods,” Journal
of Applied Psychology, 71, 2, 219–231.
—— and L. R. James (1994), “Causal Models in Organizational Be-
havior Research: From Path Analysis to LISREL and Beyond,”
in J. Greenberg (Ed.), Organizational Behavior: The State of the
Science, Hillsdale, NJ: Erlbaum, 181–205.
Williamson, O. E. (1975), Markets and Hierarchies, New York: The
Free Press.
—— (1985), The Economic Institutions of Capitalism, New York: The
Free Press.
Zaheer, A. and N. Venkatraman. (1995), “Relational Governance as an
Interorganizational Strategy: An Empirical Test of the Role Trust
in Economic Exchange,” Strategic Management Journal, 19, 5,
373–392.
Zajac, E. J., and C. P. Olsen (1993), “From Transaction Cost to Trans-
action Value Analysis: Implications for the Study of Interorga-
nizational Strategies,” Journal of Management Studies, 30, 1,
131–145.
Zucker, L. G. (1977), “The Role of Institutionalization in Cultural Per-
sistence, 1840–1920,” American Sociological Review, 42, 726–
743.
—— (1986), “Production of Trust: Institutional Sources of Economic
Structure, 1840–1920,” in L. L. Cummings and B. M. Staw (Eds.),
Research in Organizational Behavior, 8, Greenwich, CT: JAI
Press, Inc, 53–111.
Accepted by Jay B. Barney; received June 26, 1995. This paper has been with the authors for three revisions.
