PirsonMalhotra STKTrust

Organization Science
Vol. 22, No. 4, July–August 2011, pp. 1087–1104

issn 1047-7039 eissn 1526-5455 11 2204 1087 doi 10.1287/orsc.1100.0581
© 2011 INFORMS
Foundations of Organizational Trust: What Matters to

Different Stakeholders?
Michael Pirson
Schools of Business, Fordham University, New York, New York 10019, pirson@fordham.edu
Deepak Malhotra
Harvard Business School, Boston, Massachusetts 02163, dmalhotra@hbs.edu
rior research on organizational trust has not rigorously examined the context specificity of trust nor distinguished
P between the potentially varying dimensions along which different stakeholders base their trust. As a result, dominant
conceptualizations of organizational trust are overly generalized. Building on existing research on organizational trust and
stakeholder theory, we introduce a more nuanced perspective on the nature of organizational trust. We develop a framework
that distinguishes between organizational stakeholders along two dimensions: depth of the relationship (deep or shallow)
and locus (internal or external). The framework identifies which of six dimensions of trustworthiness (benevolence, integrity,
managerial competence, technical competence, transparency, and identification) will be relevant to which stakeholder type.
We test the predictions of our framework using original survey data from 1,298 respondents across four stakeholder groups
from four different organizations. The results reveal that the relevant dimensions of trustworthiness vary systematically
across different stakeholder types and provide strong support for the validity of the depth and locus dimensions.
Key words: trust; stakeholders; organizational trust; trustworthiness; integrity; benevolence; competence; identification;
transparency
History: Published online in Articles in Advance October 22, 2010.
Trust, defined as the willingness to be vulnerable to accept vulnerability to the actions of the organization
to the discretionary actions of another party, has been (Zaheer et al. 1998). As we discuss in greater detail
widely recognized as a key enabler of organizational below, this implicates an individual-to-organization level
success (Davis et al. 2000). Trust facilitates efficient of analysis, wherein trust in the organization may be
business transactions (Nooteboom 1996, Williamson based in part on attributions that the stakeholder makes
1993), increases customer satisfaction (Doney and regarding the trustworthiness of relevant organizational
Cannon 1997, Morgan and Hunt 1994), and enhances actors.
employee motivation and commitment (Brockner et al. We take as a point of departure the perspective
1997, Tyler 2000). More generally, trust promotes that trust entails positive expectations regarding another
cooperative behavior within organizations (Gulati and party’s behavior and intentions (e.g., Rousseau et al.
Westphal 1999, Williams 2001) and between organiza- 1998) and that these expectations are based on the attri-
tional stakeholder groups (Jensen 2003, Uzzi 1997), as it butions the trustor makes regarding the trustworthiness
fosters commitment (Ganesan 1994), motivation (Dirks of the other party (Malhotra and Murnighan 2002, Mayer
1999), creativity, innovation, and knowledge transfer et al. 1995, McAllister 1995). For example, Mayer et al.
(Tsai and Ghoshal 1998). As such, by strengthening rela- (1995), in their seminal review of trust research, iden-
tionships between the firm and its various stakeholders tify attributions regarding “ability,” “benevolence,” and
(e.g., employees, customers, and investors), trust serves “integrity” as three primary dimensions along which the
as a source of competitive advantage for the organization trustworthiness of the target may be evaluated. Mishra
(Barney and Hansen 1994, Nahapiet and Ghoshal 1998). (1996) includes attributions regarding “openness” as a
With the impact of trust gaining increased substantia- relevant dimension, whereas Sitkin and Roth (1993) and
tion, a vast and growing literature has begun to focus Lewicki and Bunker (1996) introduce the role of “iden-
on identifying the foundations of trust in organizational tification.” Still others include attributions of discre-
contexts (Kramer 1999). tion, predictability, and reliability in their models (Rotter
This paper examines whether and how the basis of 1971, Sheppard and Sherman 1998).
trust in organizations differs across different types of Whereas trustworthiness is widely considered to
stakeholders. As conceptualized here, stakeholder trust be multidimensional (Butler 1991, Tschannen-Moran
in an organization entails a willingness on the part of the and Hoy 2000), which dimension is most relevant
individual stakeholder (e.g., a customer, employee, etc.) in a particular situation can vary (Coleman 1990,
1087
Pirson and Malhotra: Foundations of Organizational Trust
1088 Organization Science 22(4), pp. 1087–1104, © 2011 INFORMS
Schoorman et al. 2007). Boersma et al. (2003) find that We begin by defining trust, and we then characterize
in different stages of a joint-venture relationship, differ- trust as it pertains to an organization’s stakeholder rela-
ent dimensions of trustworthiness are critical. Similarly, tionships. Next, we identify those dimensions of trust-
Sheppard and Sherman (1998) argue that in different worthiness that are likely to play a role in stakeholder
types of relationships, entirely different dimensions of relationships. We then develop a framework that identi-
trustworthiness may be relevant. As we elaborate below, fies two dimensions along which different stakeholders
the relevant dimension(s) of trustworthiness in any rela- might be categorized and use this framework to develop
tionship are likely to be a function of the type of vul- and test hypotheses regarding stakeholder-specific foun-
nerability the trustor faces. dations of organizational trust.
To examine the contingent nature of stakeholder trust
(Zey 1998), we focus on the type of the stakeholder
relationship (see Friedman and Miles 2002, Wicks et al. The Attributional Basis of Trust
1999). Drawing on stakeholder theory, which suggests Although definitions of trust vary across disciplines,
that stakeholders differ greatly in their expectations and most conceptualizations of trust include the element of
interests (Donaldson and Preston 1995, Schneper and risk or vulnerability (Rousseau et al. 1998); trust exists
Guillen 2004), we posit that different stakeholders will when parties are willing to make themselves vulnerable
also look for different signals regarding the trustworthi- to the discretionary behavior of others. More specifically,
following Rousseau et al. (1998), we define trust as the
ness of the organizations with whom they interact. For
psychological willingness of a party to be vulnerable
example, customers may trust an organization because
to the actions of another party (individual or organiza-
they perceive its workers to be competent, whereas
tion) based on positive expectations regarding the other
employees base their trust in the organization on whether
party’s motivation and/or behavior (Ferrell 2004, Mayer
management is perceived to be benevolent. This raises
et al. 1995).
an important and, as yet, unanswered theoretical (as well
Trust, so defined, should be distinguished from dimen-
as empirical) question regarding the basis of organiza-
sions of trustworthiness (Mayer and Davis 1999, p. 134);
tional trust: Which dimensions of trustworthiness are
attributions along these dimensions are what create in
most relevant to which stakeholder groups?
the trustor a willingness to accept vulnerability (Branzei
Although prior research has identified a host of dimen- et al. 2007, Kim et al. 2006, Mayer and Davis 1999,
sions—e.g., competence, benevolence, integrity, trans- Weber et al. 2005). This conceptualization is consistent
parency, etc.—which are potentially relevant across a with earlier work that has focused on the attributional
variety of relationships, there has been relatively little underpinnings of trust (Ferrin and Dirks 2003, Jarvenpaa
effort to integrate this important body of work. As a et al. 1998, Mayer and Davis 1999, McKnight et al.
result, little is known regarding which of these dimen- 1998, Tomlinson and Mayer 2009, Weber et al. 2005).
sions are most critical in building and sustaining stake- Leveraging attribution theory, Ferrin and Dirks (2003)
holder trust, and whether the relevance of the various argue that individuals seek out information that will
dimensions is contingent upon the nature of the stake- allow them to draw inferences regarding the disposition
holder relationship (Bigley and Pearce 1998). Further- and motivation of those they are in a position to trust.
more, as we discuss below, this lack of integration has Kramer (1999) delineates a number of sources for such
made it difficult to identify the precise nature of orga- information, including the target’s reputation as gleaned
nizational trust as it is built (or undermined) across from third parties and the target’s behavior in previous
stakeholders. interactions. Prior research on the attributional basis of
This paper seeks to remedy this in three ways. First, trust has not, however, attempted to identify which types
we bridge organizational trust theory and organizational of attributions (i.e., along which dimensions) will be rel-
stakeholder theory to create a contingency framework evant in which types of relationships.
that identifies which dimensions of trustworthiness are
most relevant for which types of organizational stake-
holders. Second, we leverage a large, original survey Conceptualizing Stakeholder Trust in
of 1,298 stakeholders from four different stakeholder Organizations
groups across four differently structured organizations to Trust may be targeted towards an individual, a group,
test the absolute relevance and the relative importance or an organization. In this paper, we consider the degree
of a comprehensive set of trustworthiness dimensions. to which an individual (stakeholder) trusts an organi-
Third, whereas previous analyses have largely focused zation, and we do so by drawing on the attributional
on within-stakeholder comparisons (e.g., Davis et al. model of trust described above. Consistent with this
2000, Lusch et al. 2003, Mayer and Davis 1999, Mayer approach, Zaheer et al. (1998) argue that the origin of
and Gavin 2005), we evaluate how the relevance of trust- trust is always grounded in an individual perspective,
worthiness dimensions differs across stakeholder types. even if individuals belonging to a certain group (such
Organization Science 22(4), pp. 1087–1104, © 2011 INFORMS 1089
as a particular stakeholder type) may share a similar versus shallow) and “form” (dependent versus inter-
orientation (p. 143). They further argue that the trust dependent). As a consequence, they propose four dis-
referent—i.e., the target of trust—may be an organi- tinct “grammars” of trust in which different dimensions
zation. Likewise, Currall and Inkpen (2002) argue that of trustworthiness are relevant. In shallow-dependent
the individual-to-organization interaction is an appro- relationships, for example, they argue that perceptions
priate level of analysis when studying trust in organi- of discretion, reliability, and competence are the key
zational contexts. Accordingly, we leverage the Zaheer determinants of trust. As relationship depth—defined
et al. (1998) approach to distinguish between interper- by Sheppard and Sherman (1998, p. 423) as being a
sonal trust and organizational trust as it pertains to “product of the importance, range and number of points
stakeholder perceptions: interpersonal trust describes the of contact among parties”—increases, integrity, con-
extent to which individuals (origin) trust other indi- cern, and benevolence become crucial. Meanwhile, a
viduals (referent) along relevant trustworthiness dimen- shift from dependence towards interdependence leads to
sions; organizational trust, in contrast, describes the an increased role for empathy, foresight, and intuition
extent to which individuals (origin) trust an organi- (p. 427). Sheppard and Sherman’s framework, although
zation (referent). Stakeholder trust in organizations, it has not been empirically examined, represents an
then, entails the willingness of individuals (customers, important step in examining the contingent nature of
employees, etc.) to accept vulnerability to the actions of trust across relationship types.
an organization based on positive expectations. The application of Sheppard and Sherman’s (1998)
Also consistent with Zaheer et al. (1998), whereas categorization to our context, however, reveals some
we focus on the degree to which “the organization” is limitations. In particular, our sample makes it difficult
trusted, we acknowledge that stakeholder expectations to apply the form dimension, because all organizational
regarding an organization’s likely behavior may stem, stakeholders would be categorized as “dependent”
in part, from expectations regarding the likely behav- according to Sheppard and Sherman’s framework.
ior of relevant organizational actors (e.g., Currall and According to Sheppard and Sherman (1998, p. 424),
Judge 1995, Inkpen and Currall 1998). Fombrun (1996) dependent relationships are those in which “one’s
argues that attributions regarding an organization consti- outcomes are contingent upon the actions of another,”
tute a reputational “umbrella” for the individual or col- and interdependent relationships are those that require
lective actions of those who represent the organization. the parties to “coordinate behavior in order to achieve
This perspective is particularly appropriate in the current desired goals.” Sheppard and Sherman’s (coordination-
context because we are studying trust across multiple based) view of interdependence renders it applicable to
stakeholder groups; different stakeholders are likely to too few stakeholder relationships. Meanwhile, their char-
interact with different organizational members and be acterization of dependence—which includes unilateral as
subject to vulnerability based on different types of orga- well as mutual dependence relations—would subsume
nizational behaviors. all of the relationships in our sample, rendering the clas-
sification moot; indeed, according to stakeholder theory,
Toward a Framework of a stakeholder–organization relationship is characterized
Stakeholder-Specific Trust by its inherent mutual dependence. We therefore take
Two distinct lines of research on trust development Sheppard and Sherman’s model as a useful point of
appear to dominate the literature. One builds on the departure and guidance in our study of stakeholder trust.
premise that perceived trustworthiness, in any relation- We build on Sheppard and Sherman’s (1998) efforts
ship and at any given time, is multidimensional; the goal by (a) leveraging research on stakeholder theory to iden-
of the research is to delineate these different dimen- tify another dimension that, along with depth, helps
sions (e.g., Lewis and Weigert 1985, Mayer et al. to meaningfully distinguish between stakeholder types;
1995, McAllister 1995, Williams 2001). The second line (b) leveraging trust research, as well as in-depth inter-
of research investigates which dimensions are relevant views of organizational stakeholders, to deductively and
at various stages of a relationship (e.g., Lewicki and inductively identify key dimensions of trustworthiness
Bunker 1996, Ring and Van de Ven 1994, Rousseau et al. that we expect will be relevant to stakeholders; and
1998, Shapiro et al. 1992, Whitener et al. 1998). (c) bringing together these two lines of research to for-
Less extensive has been the effort to identify the mulate a stakeholder-specific model of organizational
potentially varying trust needs in distinct relationships. trust that yields a series of empirically testable proposi-
A notable exception is the framework developed by tions regarding which dimension of trustworthiness will
Sheppard and Sherman (1998). Based on Fiske’s (1990) be relevant to which type of stakeholder.
relationship categorization (communal sharing, author-
ity ranking, equality matching, and market pricing), Dimensions of Trustworthiness
Sheppard and Sherman argue that relationships can be As Rousseau et al. (1998, p. 398) posit, “Trust takes
meaningfully distinguished in terms of “depth” (deep different forms in different relationships.” Depending on
the situation, there are several potential attributions on to describe their relationship with the target organiza-
which trust may be based (Boersma et al. 2003). In their tion and to explain how they defined or interpreted the
seminal review of the literature, Mayer et al. (1995) sug- notion of trust as applied to their relationship with the
gest that would-be trustors seek to formulate attributions organization. As we describe in more detail in the Meth-
along three key dimensions: ability, benevolence, and ods section, the responses converged significantly with
integrity. Attributions regarding ability gauge whether our definition of trust. (See Appendix B for examples of
a trustee has the requisite level of competence to per- trust definitions provided by the stakeholders we inter-
form the tasks entrusted in him. Attributions regarding viewed.) Then, to validate our dimensionality of trust-
benevolence gauge whether the trustee exhibits goodwill worthiness, we employed the critical incident technique
toward the trustor and has concern for the trustor’s well- (Woolsey 1986), which involved the elicitation of thick
being. Attributions of integrity gauge whether a trustee is descriptions of trust-related events to explore the dif-
perceived as forthcoming, honest, and of requisite moral ferent types of attributions that seemed to underlie the
character. interviewee’s trust in the organization.
In this paper, we take the Mayer et al. (1995) abil- The interviews validated the four dimensions we had
ity, benevolence, and integrity (ABI) framework as the identified ex ante but led to two further modifica-
starting point for our analysis of the determinants of tions of our framework. First, as described in detail
stakeholder trust. Mayer et al. (1995) suggest that attri- later, we found two different subcategories of references
butions regarding ability, benevolence, and integrity rep- to ability. These corresponded closely with Madhavan
resent an exhaustive list of relevant trustworthiness and Grover’s (1998) distinction between managerial
dimensions for both interpersonal and organizational and technical competence. Second, we found a number
contexts. However, for our analysis, we added a fourth of references to the notion of transparency. Although
dimension: identification. According to Lewicki and Mayer et al. (1995) nominally subsumed transparency
Bunker (1996), identification-based trust stems from the (openness) under integrity, their empirical examinations
understanding and internalization of the interests and of integrity (e.g., Mayer and Davis 1999, Mayer and
Gavin 2005) exclude any mention of transparency or
intentions of the other party, based on shared values
openness. We therefore decided to include transparency
and commitment.1 There were three reasons for this
as a separate dimension in our analysis. Consistent with
expansion of the ABI framework. First, whereas Mayer
this, several scholars have argued that transparency, or
et al. (1995) subsume elements of identification, or value
the perceived willingness to share trust-relevant infor-
congruence, as part of integrity, they admit that others
mation with vulnerable stakeholders, is a distinct crit-
(e.g., Sitkin and Roth 1993) have studied identification
ical dimension of trustworthiness (e.g., Mishra 1996,
as a separate construct when the organization was the
Tschannen-Moran 2000). (See Appendix B for examples
trust referent (p. 720). Second, in their first empirical of trustworthiness dimensions mentioned by the stake-
analysis of the ABI framework, Mayer and Davis (1999) holders we interviewed.)
devote only one item in their integrity scale to identifica- Based on the ABI model as an organizing framework,
tion. Finally, whereas Mayer et al. (1995) argue that the and on subsequent refinements stemming from our qual-
ABI dimensions are universally relevant, Lewicki and itative interviews, we examined attributions along six
Bunker’s (1995) work suggests that identification-based key dimensions of trustworthiness: ability (disaggregated
trust is formed only very late in a relationship, and that into managerial competence and technical competence),
identification, or perceived value congruence, may be a benevolence, integrity, transparency, and identification.
feature of very few close relationships. Our slight mod-
ification of the ABI framework creates four dimensions Dimensions of Stakeholder Relationships
of stakeholder trust. Although stakeholders may differ along many differ-
Despite the Mayer et al. (1995) suggestion that the ent dimensions (e.g., the degree to which they have
ABI framework is relevant to both individual and orga- power or legitimacy; Mitchell et al. 1997), we follow
nizational relationships, we found no evidence of its the lead of network theorists who study organizational
prior application in the study of organizational trust stakeholders in focusing on (a) the degree of interac-
across different stakeholder groups. As such, we sought tion between stakeholders and organizations and (b) the
to further validate our adaptation of the ABI frame- structural positioning of stakeholders vis-à-vis the orga-
work through qualitative interviews of actual organi- nization as critical determinants of stakeholder influence
zational stakeholders. Prior to conducting the expan- and organizational behavior (see, e.g., Rowley 1997).
sive survey that represents the core of this paper, we These two dimensions are particularly appropriate when
conducted 32 semistructured interviews of four differ- studying organizational trust because each of them influ-
ent stakeholder groups: employees, customers, suppli- ences the type and degree of vulnerability a stakeholder
ers, and investors.2 During these interviews, which lasted faces, the type of information a stakeholder can access,
between 25 and 90 minutes each, we asked respondents and the attributions the stakeholder is likely to make
regarding the organization (Dervitsiotis 2003, Shankar of vulnerabilities (Ogden and Watson 1999). For exam-
et al. 2002, Swift 2001). Consistent with Sheppard and ple, the emergence of a strong competitor may threaten
Sherman (1998), our first dimension is depth. Depth is internal stakeholders of an organization (e.g., employees
a measure of the extent and intensity of a stakeholder’s and investors), even as it makes life better for external
interaction with an organization. The second dimension, stakeholders (e.g., suppliers and customers) who stand
which we call locus, captures the stakeholder’s posi- to benefit from the competition. Internal and external
tion (internal versus external) vis-à-vis the organization stakeholders also seek to access information regarding
(Jones and Wicks 1999, Levin et al. 2006, Schneper different organizational behaviors (Brickson 2005, Scott
and Guillén 2004, Schoorman et al. 2007). Limiting our and Lane 2000).
attention in this paper to these two stakeholder dimen-
sions seems further warranted because one of these The Depth Dimension
dimensions (i.e., depth) has firm grounding in the orga- Relationships can be differentiated according to the
nizational trust research and the other (i.e., locus) is degree of uncertainty involved (Luhmann 1979, 2000).
well researched by stakeholder theorists. These are the In shallow relationships, where interactions between the
two literatures that we are bringing together in this organization and the stakeholder are sparse and the dura-
framework. tion of contact has been short, uncertainty about the
Relationship depth refers to the intensity and intimacy behavior of the other party is likely to be high. Accord-
of a relationship. Although it is often difficult to assess ing to Mayer et al. (1995), attributions of integrity are
the precise depth of a relationship, it can be approx- likely to play an important role in trust development in
imated, at least in part, by interaction frequency and such relationships. Likewise, Granovetter (1985) argues
relationship duration (Kramer 1999, Levin et al. 2006, that generalized morality—or integrity—plays a crucial
Lewicki and Bunker 1995, Macy and Skvoretz 1998, role in the sustenance of “weak tie” (i.e., low intensity)
Mayer et al. 1995, Sheppard and Sherman 1998). Like- relationships. Thus, in shallow relationships that entail
wise, our depth dimension consists of a measure based high levels of uncertainty, perceptions of integrity may
on the interaction frequency and relationship duration be necessary to induce the degree of trust required for
of a stakeholder with an organization and distinguishes coordination and cooperation. Similarly, Whitener et al.
between those that have shallow versus deep relations
(1998) argue that trust initiation is based on perceiv-
with the organization (Becerra and Gupta 2003, Kenning
ing the organization (i.e., relevant organizational deci-
2001, Lewicki and Bunker 1996, Lewis and Weigert
sion makers) as honest and forthcoming. Mayer et al.
1985, McAllister 1995, Sheppard and Sherman 1998).
(1995, p. 722) make the point succinctly: the “effect of
This classification is largely consistent with Granovet-
integrity on trust will be most salient early in the rela-
ter’s (1985) categorization of strong versus weak ties.
tionship prior to the development of meaningful benev-
Depth may influence stakeholder trust for two reasons:
olence data.”
(1) it affects the degree to which a stakeholder is vul-
Transparency is also likely to be of relevance to stake-
nerable to organizational actions, and (2) it affects the
holders in shallow relationships (Dervitsiotis 2003,
degree of risk-related information exchange that is pos-
DiPiazza and Eccles 2002, Sheppard and Sherman
sible and likely between the stakeholder and the organi-
zation (Eisenhardt 1989, Rousseau et al. 1998, Sheppard 1998, Turnbull 2002). According to Hardin (2002) and
and Sherman 1998). McKnight et al. (1998), when there is little previous
Our locus dimension, which focuses on the posi- interaction and when information asymmetry is high, all
tion of the stakeholder relative to the organization, trust-relevant information is sought and scrutinized. This
is based on extant stakeholder research that distin- should accentuate the importance of transparency. For
guishes between internal and external stakeholders example, corporate communication initiatives and newly
(Freeman 1984, Donaldson and Preston 1995). Follow- developed reporting standards (e.g., the Global Report-
ing Schneper and Guillén (2004), we view employees ing Initiative) are aimed at building trust with stake-
and investors as internal stakeholders. Stakeholder the- holders (e.g., investors) who might otherwise not have
ory argues that internal stakeholders represent the orga- access to information regarding organizational behaviors
nization (e.g., Blair 1995, Freeman 1984): employees and intentions. We therefore hypothesize the following.
do so because they work for the organization (Aguilera Hypothesis 1. Stakeholder trust in shallow relation-
and Jackson 2003), and investors do so because they are ships will be based on perceptions of transparency and
owners of the organization (Goodpaster 1991).3 In con- integrity.
trast, customers and suppliers are external stakeholders:
neither group works for or owns the organization, and Whereas stakeholders in deep relationships might also
neither is typically seen as representing the organization. need to see a “track record” of ethical and honest behav-
Locus is likely to influence stakeholder trust because ior (Elangovan and Shapiro 1998, Mishra and Spreitzer
internal and external stakeholders face different types 1998, Sheppard and Sherman 1998), their demands
extend beyond those of stakeholders in shallow rela- ent vulnerabilities, there are different capabilities that
tionships (Rempel et al. 1985, Rempel and Ross 2001). an organization must develop and signal to build trust.
Deep relationships entail not only the need for but also Based on our qualitative analysis, described further
the capacity for more information exchange. As con- below, we follow Madhavan and Grover (1998) and dis-
tact with the organization increases and duration of the tinguish between two types of competence—managerial
contact persists, a stakeholder’s vulnerability is likely to competence and technical competence—and argue that
increase (see also Gottman and Levenson 1992), as does the relevance of each type depends upon the locus
its ability to better evaluate the impact of organizational of the stakeholder (see also Hodson 2004, Tan and
behaviors. This reduces the need for perceived integrity Libby 1997). Managerial competence is implicated in
(Mayer et al. 1995, Lewicki and Bunker 1996, Rousseau the organization’s ability to make strategic decisions
et al. 1998) but increases the role of perceived orga- and manage stakeholder relationships. Technical compe-
nizational benevolence (McAllister 1995, Shaw 1997, tence is implicated in the organization’s ability to deliver
Sheppard and Sherman 1998). Unlike stakeholders in high-quality products and services (see, e.g., Miles and
shallow relationships, stakeholders in deep relationships Snow 1992). External stakeholders are more likely to
are not likely to see themselves as “just another stake- be directly impacted by, and also more likely to have
holder” and will thus have a greater need for (and information that allows them to assess, technical com-
access to) information that signals organizational benev- petence rather than managerial competence. Consistent
olence (Mayer and Davis 1999, McAllister 1995). Con- with this, Parmigiani and Mitchell (2005) have found
sistent with the above logic, Mayer et al. (1995, p. 722) that the extent to which suppliers (external stakehold-
argue that the effect of perceived benevolence on trust ers) trust the organization is highly dependent on the
increases “over time as the relationship between the par- technical expertise of the buying organization. Similarly,
ties develops.” Morgan and Hunt (1994) argue that customer trust (also
Whereas integrity refers to an organization’s gen- external) is based on satisfaction with the quality of the
eral tendency (or propensity) to act fairly and ethically, product or the service offered, which again implicates
benevolence refers to the organization’s concern for their the technical aspect of competence.
stakeholders’ well-being (see also Whitener et al. 1998). Internal stakeholders, on the other hand, such as em-
Organizational stakeholders perceive benevolence when ployees and investors, may care relatively more about—
organizations (or relevant organizational actors) express and be in a better position to judge—managerial com-
concern, care, and interest (Edmondson 1999, Mayer petence, such as decision-making ability and strate-
et al. 1995), even when fairness or equity does not gic vision, which are essential for long-term survival
demand it (e.g., will the employee be laid off during and competitiveness. For example, Shockley-Zalabak and
an economic downturn?). Stakeholders in deep relation- Morley (1994) argue that employees (internal stakehold-
ships may continue to value integrity, albeit less so than ers) evaluate organizations on whether they will survive
those in shallow relationships, but their demand for per- and be able to compete. Likewise, Mayer and Gavin
ceived benevolence will increase (Mayer et al. 1995). (2005), Davis et al. (2000), and Hodson (2004) show that
We therefore hypothesize the following. employees trust organizations more because of high man-
agerial competence and enduring success in the market
Hypothesis 2. Stakeholder trust in deep relation-
place, and less because of technical expertise or product
ships will be based on perceptions of integrity and
quality. Investor trust has also been shown to be based in
benevolence.
large part on the perceived competence of the firm’s man-
Hypothesis 3. Perceived integrity will be less rele- agement team (e.g., Manigart et al. 2002, Warner et al.
vant to stakeholder trust in deep relationships than in 1988). Thus, we hypothesize the following.
shallow relationships.
Hypothesis 5. Perceived managerial competence
Hypothesis 4. Perceived benevolence will be rele- will be more relevant to trust among internal stakehold-
vant to stakeholder trust in deep relationships, but not ers than among external stakeholders.
in shallow relationships.
Hypothesis 6. Perceived technical competence will
be more relevant to trust among external stakeholders
The Locus Dimension
than among internal stakeholders.
Stakeholder trust is based not only on the perceived
intention of the organization (as captured, for example, Locus can also affect trust development in another
by integrity and benevolence attributions regarding rel- way. Whereas external stakeholders are, by definition,
evant organizational actors) but also on the perceived not a part of the organization (Brickson 2005, Freeman
ability of the organization to behave in ways that 1984, Schneper and Guillén 2004), many internal stake-
benefit the stakeholder (McAllister 1995, Mayer and holders will have, at some point, made conscious deci-
Davis 1999). Because different stakeholders face differ- sions to join, and thereby represent, the organization.
This increases their vulnerability: the internal stake- internal database or via publicly available databases
holder faces not only financial risks at the discretion of containing personal profiles and organizational affilia-
organizational decision makers, but also identity risk— tions. Investors included institutional investors, invest-
i.e., the threat to one’s self-perceptions based on the ment advisors, venture capitalists, and private investors.
fact that one’s social identity may include an association All stakeholders were contacted via e-mail or through
with the organization, its behaviors, and its espoused direct contact (in which case they were asked to fill
values (Giddens 1991, Sitkin and Roth 1993, Tajfel and out paper surveys). Those respondents contacted via
Turner 1979). This will be particularly true for those e-mail were also asked to forward the survey to fel-
internal stakeholders who have frequent interaction with low stakeholders (snowball sampling). An introductory
the organization and are hence “closest” to it—i.e., deep sheet described the survey and explained the process
internal stakeholders (Lewicki and Bunker 1996). As used to ensure anonymity. The data for each organiza-
Rousseau et al. (1998, p. 18) posit, “There is a tendency tion were collected during a period of one to two months
for repeated interactions to 0 0 0 give rise to a psychologi- to reduce the potential effects of organizational change.
cal identity (Gaertner et al. 1996).” To mitigate identity The overall data collection process spanned a period of
risks, these stakeholders will have to evaluate whether five months.
their own values are congruent with those of the organi- Because of the snowball sampling procedure for
zation (Schein 1985, Sitkin and Roth 1993). customers, suppliers, and investors, a response rate is
Like Rousseau et al. (1998), we argue that the rela- difficult to establish. We also cannot exclude the pos-
tionship between an organization and its deep inter- sibility, albeit remote, that some respondents may have
nal stakeholders will be based in part on perceptions responded as a stakeholder for more than one orga-
of value congruence, and more generally, on identifica- nization (e.g., as a client for one organization and as
tion (Enz 1988, Lewicki and Bunker 1996, Yaniv and an investor for another). The response rate for employ-
Farkas 2005). Consistent with this, Shockley-Zalabak ees contacted through the organization ranged from 8%
et al. (1999) have posited that identification is a critical to 10%, except for Organization 1, where 63% of the
aspect of the trusting relationship between employees employees responded. Although we cannot fully dis-
and their organizations, and Lewicki and Bunker (1996) count the possibility of response bias in such a survey
have argued that identification-based trust is the highest (e.g., that the most and least trusting stakeholders were
form of trust and is characteristic of relatively few close more likely to answer the survey), such concerns are
relationships. We therefore hypothesize the following. mitigated for two reasons. First, we conducted supple-
mental analyses (e.g., checking for kurtosis in the distri-
Hypothesis 7. Identification will be relevant to trust bution of trust responses) to test whether our respondents
only among deep internal stakeholders. exhibited a bipolar distribution on trust. The results of
these analyses argue against such a distribution. Second,
Method although response bias could affect the mean level of
trust in our sample relative to the overall population, our
Sample and Data Collection analysis does not focus on mean trust levels. Rather, we
We administered surveys to stakeholders from four dif- test the relevance of different trustworthiness dimensions
ferent organizations in western Europe. To have as to different stakeholders, which is arguably less likely to
representative a sample as possible, we chose organi- be affected by self-selection or representativeness issues.
zations of different sizes (ranging from small/regional Finally, supplemental analyses with our data revealed no
to large/multinational), with different ownership struc- evidence that high versus low trust respondents based
tures (privately held, publicly held, government owned), their trust on different types of attributions.
and from different industries (manufacturing, logistics, Overall, 1,298 usable responses were received. EM
consulting, and research). Organization 1 is a small to imputation was used to deal with missing data (Allison
medium-sized manufacturing firm in Switzerland, Orga- 2001).4 Customers were the largest group (N = 601),
nization 2 is a large logistical company in Germany, followed by employees (N = 423), suppliers (N = 141),
Organization 3 is a western European branch of an inter- and investors (N = 133). (Note that because Organiza-
national consulting firm, and Organization 4 is a public tion 4, as a public university, may be qualitatively differ-
university in Switzerland. ent from other organizations in the sample, and because
The stakeholders we surveyed were investors (inter- Organization 4 only allowed us to reliably identify and
nal), employees (internal), customers (external), and sup- reach one type of stakeholder (employees), we also ran
pliers (external). Employees were primarily contacted by a separate set of analyses in which Organization 4 was
the organization itself via e-mails that encouraged par- eliminated. This did not affect the results in any of the
ticipation in the anonymous process and contained a link hypothesis tests we conducted.)
to the survey. Customers, suppliers, and investors were Of the respondents, 74% were male; the age groups
sampled randomly based either on an organization’s of 18–30 (43.3%) and 31–45 (41.9%) were most highly
represented. Of the respondents, 51% reported that they in another set of interviews and in survey pretests with
had been in contact with the organization for more than 260 stakeholders of the target organizations. Appendix A
seven years, 23.3% reported four to seven years of con- lists the items for each measure.
tact, and 18.6% reported one to three years. Of the
respondents, 59% reported more than 100 prior interac- Independent Measures. To validate and/or modify,
tions with the organization; 14.7% reported between 50 in our empirical context, the trustworthiness dimen-
and 100 interactions. sions suggested by the ABI framework, we interviewed
The data on prior interactions and the length of the 32 stakeholders with regard to their understanding of
relationship between stakeholders and their organization trust in the organizational context and with regard to
were used to classify stakeholders as being in shal- their specific stakeholder experiences. We asked them
low versus deep relationships. We classified stakehold- to provide us with a definition of organizational trust
ers who reported more than 100 interactions and more and, following the critical incident technique (Woolsey
than three years of contact with the organization as hav- 1986), to tell us about incidents in which their trust
ing a deep relationship. Stakeholders with fewer than in an organization had increased or decreased. Some
100 interactions or less than three years of contact were interviewees provided responses from more than one
classified as low-depth stakeholders.5 This classifica- stakeholder perspective, and we therefore ended with
tion resulted in relatively equal sample sizes between 16 narratives regarding employee relationships, 18 nar-
the shallow (N = 665 stakeholders) and deep (N = 633 ratives regarding customer relationships, and 5 narratives
stakeholders) categories. In addition, N = 556 stakehold- each for supplier and investor relationships. We coded
ers were classified, by definition, as internal (employees the interview transcripts using template analysis (King
and investors) and N = 742 stakeholders were classified 1998, Maznevski and Chudoba 2000).6 As predicted, we
as external (customers and suppliers). found significant overlap with the ABI framework pro-
vided by Mayer et al. (1995). However, as described
Measures previously, we deconstructed the Mayer et al. construct
Although it would have been preferable to use exist- of integrity to include transparency and identification as
ing scales for measuring trust among our respondents, separate constructs. We also found that stakeholders who
a review of existing trust scales revealed two problems mentioned ability or competence seemed to clearly dis-
(see also Dietz and Den Hartog 2006). First, most estab- tinguish between managerial and technical dimensions
lished scales (e.g., Mayer and Davis 1999, Robinson of competence. We then created survey items to mea-
1996, Spreitzer and Mishra 1999)—even those designed sure the six trustworthiness dimensions. Responses were
to measure trust in organizational contexts—are aimed marked using a five-point scale with endpoints labeled
at assessing trust perceptions within one specific stake- “strongly disagree” (1) and “strongly agree” (5). Follow-
holder group (usually employees). Second, and even ing a procedure similar to that of Hoy and Tschannen-
more problematic, existing scales consist of items that Moran (1999), for each dimension of trustworthiness,
already presume which dimensions of trustworthiness we identified two to four items that demonstrated high
will be relevant (Dietz and Den Hartog 2006). In other convergent and discriminant validity in pretest samples
words, whereas we distinguish between trust (our depen- using exploratory factor analysis (see also Ross and
dent measure) and dimensions of trustworthiness (our Lacroix 1996).
independent measures), existing scales typically com- The convergent and discriminant validity of the mea-
bine these constructs. sures was then calculated using the full participant sam-
To the extent possible, we based our independent ple. Using exploratory factor analysis, we found that
measure items—i.e., those relating to trustworthiness all item-factor loadings were significant (p < 0005) and
dimensions—on suitable, existing scales (e.g., Mishra greater than 0.40, supporting the convergent validity
1996, Shockley-Zalabak et al. 1999, Tschannen-Moran criterion. We used confirmatory factor analysis (CFA)
and Hoy 2000) and followed Dillman’s (2000) tailored based on AMOS 7 (Arbuckle 2006) to compare the fit
design method to adjust items as needed to (a) orient of our hypothesized model (which predicted six sepa-
them to trust in organizational relationships, (b) broaden rate constructs) to a number of alternative models (see
their scope so they apply to all stakeholder groups, Table 1). A test for multivariate normality had been
and (c) maintain the ability to create a separate depen- conducted prior to the analysis, which yielded positive
dent measure for trust. Because the survey was con- results (skewness and kurtosis of all items below one).
ducted in German-speaking parts of Europe, we had the Additionally, estimating the CFA model using maxi-
items translated and back-translated to ensure accuracy mum likelihood estimation (MLE) and asymptotically
(Chapman and Carter 1979). We then used university distribution-free (ADF) interval estimation did not reveal
faculty and doctoral students to reassess the face validity any significant differences, thus making it safe to assume
of the items (DeVellis 1991). This assessment yielded that the model was stable even if the data violated
a set of revised items that were then tested for clarity assumptions of multivariate normality. The fit indexes
Table 1 Results of Confirmatory Factor Analysis for Study Variables
Model 2 df NFI RFI IFI TLI CFI RMSEA
Six-factor model 6400794 104 00963 00951 00969 00959 00969 00063
Five-factor model a 9450774 109 00945 00932 00951 00939 00951 00077
One-factor model 313410024 119 00806 00778 00812 00785 00812 00144
Notes. The chosen five-factor model was the best-fitting five-factor model. NFI, normed fit index; RFI, relative noncentrality index; IFI,
incremental fit index; TLI, Tucker-Lewis index; RMSEA, root mean square error of approximation.
a
Managerial and technical competences combined.
reveal that the hypothesized six-factor model fit much organization. This was done by evaluating their willing-
better than any of the alternative models, including a ness to recommend the organization to others. Ideally,
three-factor model based on the Mayer et al. (1995) ABI in evaluating respondents’ willingness to accept vul-
framework (see Table 1). Despite some high correla- nerability, we would have asked them to provide their
tions, the data support sufficient discriminant validity of own willingness to interact with the organization. How-
the items. The scale reliabilities are high, with Cron- ever, because all respondents were already interacting
bach’s alphas ranging from 0.85 to 0.93. Supplemen- with the organization (as stakeholders), we instead used
tal analysis, in which the model was tested separately the two “recommendation” items as proxies. One was
for customers, suppliers, employees, and investors, was a general recommendation item (“I would recommend
also supportive (comparative fit indices (CFIs) ranging the organization because I predict a good future for the
from 0.92 to 0.96), suggesting model stability despite organization”); the other was a targeted, stakeholder-
the variance in sample size across stakeholders. Table 2 specific recommendation item, with respondents being
displays descriptive statistics and correlations for all asked whether they would recommend to friends that
variables. they buy, invest, work for, or work with the organization.
Dependent Measure. To measure the stakeholder’s The alpha for these three items was 0.84.
trust in the organization without referencing specific We then conducted a battery of tests and assessments
trustworthiness dimensions that may be expected to to ensure the validity of our measure. First, we tested
impact trust (integrity, benevolence, etc.), we created a to ensure that the concept of trust was not understood
three-item trust measure based on our definition of stake- differently across stakeholder groups. To do so, we con-
holder trust. The first item asked whether the respondent ducted a measurement invariance test based on the survey
trusted the organization (“I trust the organization”). The data, which suggested no significant effect of stakeholder
second and third items sought to proxy for the respon- group on the interpretation of “trust.” We also tested for
dents’ willingness to be vulnerable to the actions of the discriminant validity of the dependent (trust) measures
Table 2 Descriptive Statistics, Correlations Among Key Variables, and Alpha Coefficients
Variable Mean S.D. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1. Managerial 3016 1014 400875

competence
2. Technical 3054 1005 0068∗∗ 400855
competence
3. Integrity 3025 1002 0061∗∗ 0063∗∗ 400855
4. Transparency 2087 0091 0064∗∗ 0057∗∗ 0069∗∗ 400875
5. Benevolence 2097 0096 0060∗∗ 0061∗∗ 0076∗∗ 0071∗∗ 400885
6. Identification 2096 1020 0065∗∗ 0064∗∗ 0067∗∗ 0063∗∗ 0070∗∗ 400935
7. Trust 2065 0080 0070∗∗ 0071∗∗ 0068∗∗ 0063∗∗ 0068∗∗ 0077∗∗ 400845
8. Gender 0026 0044 −0005∗∗ −0006 −0003 −0001 −0005 −0009 −0009∗∗
9. Age 2072 0076 0005 0004 −0002 −0002 −0003 0001 0000 −0014∗∗
10. Multiple 0029 0045 0018∗∗ 0018∗∗ 0012∗∗ 0010∗∗ 0014∗∗ 0022∗∗ 0017∗∗ −0007∗∗ 0002
stakeholder
∗∗ ∗∗ ∗∗ ∗∗ ∗∗ ∗∗ ∗∗ ∗∗
11. Organization 1 0007 0026 0023 0014 0018 0022 0026 0017 0019 −0012 0018∗∗ −0006∗
12. Organization 2 0063 0048 −0065∗∗ −0042∗∗ −0036∗∗ −0036∗∗ −0035∗∗ −0041∗∗ −0045∗∗ 0005 −0005 −0018∗∗ −0036∗∗
13. Organization 3 0021 0041 0046∗∗ 0029∗∗ 0020∗∗ 0019∗∗ 0013∗∗ 0022∗∗ 0026∗∗ −0003 −0007∗ 0016∗∗ −0014∗∗ −0071∗∗
14. Organization 4 0008 0028 0024∗∗ 0018∗∗ 0017∗∗ 0014∗∗ 0018∗∗ 0022∗∗ 0021∗∗ 0007∗ 0003 0014∗∗ −0008∗∗ −004∗∗ −0016∗∗
15. Locus 0057 0050 −0050∗∗ −0044∗∗ −0036∗∗ −003∗∗ −0035∗∗ −0051∗∗ −0046∗∗ 0001 0003 −0032∗∗ −0005 0051∗∗ −0034∗∗ −0035∗∗
16. Depth 0049 0050 0002 0002 −0007∗ −0008∗∗ −0008∗∗ 0007∗ 0005 −0007 00095∗∗ 0006∗ −00076∗ 00064∗ −0006∗ 0004∗ −0006∗
Notes. N = 11298. Alpha coefficients are on the diagonal in parentheses.

∗
Correlations significant at 0.05; ∗∗ correlations significant at 0.01.
and the independent measures. To do that, we included Model 1). We found that all dimensions are signifi-
the dependent measures as a seventh factor in the CFA cant predictors of trust except for transparency, which is
model comparing the different models tested before. The only marginally significant (p < 0010). The results are
seven-factor model fit significantly better than any other unchanged when we include control variables. Although
model combining the dependent variable with indepen- multicollinearity exists, as would be expected given that
dent variables (as indicated by modification indices >100 all of the dimensions are related to trust as a second-
and the reduction of 2 > 500). The discriminant patterns order construct, it does not confound our results because
of the model fits remained stable, indicating discriminant variance inflation factors among the main constructs do
validity of the dependent and independent measures. not exceed 3.16, which is well below the critical level
Finally, we did a qualitative analysis of the text of our (>5) for multicollinearity to be a concern (McDaniels
interviews. Although the interviews were semistructured, 2005). In addition, although the trustworthiness dimen-
many interviewees were asked to define trust as the term sions were correlated, we found them to be meaningfully
would be used in the surveys, i.e., with an organization distinct, as demonstrated by the CFA described earlier.
as the referent. Of the 32 interviewees, 20 provided a Finally, a test of the discriminant validity of constructs
definition of trust. Of these responses, 85% (N = 17) that would have been combined if using the original ABI
seemed to correspond closely (or precisely) with our def- framework (technical with managerial competence, iden-
inition of trust explicitly, or implicitly citing the willing- tification with integrity, and transparency with integrity)
ness to take risk or accept vulnerability in interactions revealed that modification indexes were well above 100
with the organization. Appendix B contains examples of when constructs were combined and that our model fit
these responses from each stakeholder group. significantly better than ABI.
Control Measures. To ensure the robustness of our
Depth: The Relevance of Integrity,
results, we included a number of control variables: age,
Benevolence, and Transparency
gender, organization type, and whether the respondent
interacted with the organization in more than one capac- Our model predicted that integrity and transparency
ity (e.g., as a customer and an employee). We also would be relevant for stakeholder trust in shallow
controlled for stakeholder type (employee, customer, relationships (Hypothesis 1) and that integrity and
investor, or supplier). Respondents checked one of five benevolence would be relevant for stakeholder trust
age brackets: “under 18,” “18–30,” “31–45,” “46–60,” in deep relationships (Hypothesis 2). We tested these
or “61 or older.” Gender was measured as a standard predictions, separately, for internal and external stake-
dichotomous variable (1 = female, 0 = male). Dummy holders (see Table 4). Hypothesis 1 received mixed
variables were included for the four organizations in support: as predicted, trust in shallow relationships
the sample. was based on perceptions of integrity for both inter-
Because many stakeholders (especially in larger orga- nal (b = 00478, p < 00001) and external stakeholders
nizations) belonged to more than one stakeholder group, (b = 00296, p < 0001). However, perceived transparency
we created a dummy variable labeled “multidimensional did not predict trust in shallow relationships for inter-
stakeholder.” After having identified their “primary” nal stakeholders (b = 00095, p = 00444) and did so
relationship with the organization, respondents were only marginally for external stakeholders (b = 00203,
asked to answer yes (1) or no (0) to the following p < 0010). As predicted by Hypothesis 2, benevolence
question: “Do you belong to more than one stakeholder was a significant predictor of trust in deep relationships
group?” Of the sample, 28.9% identified themselves for internal (b = 00325, p < 0001) and external stake-
as multidimensional stakeholders. Finally, we included holders (b = 00288, p = 00067). Contrary to Hypothe-
dummy variables for stakeholder type, which respon- sis 2, but consistent with Hypothesis 3, integrity was
dents selected from a given list (including employee, not a significant predictor of trust in deep relationships
investor, supplier, and customer). In a separate analy- (internal: b = −00083, p = 00485; external: b = 00190,
sis, we also controlled for the moderating effects of the p = 00148).
organization on the independent variables. The results Hypothesis 3 predicted that integrity would be more
did not reveal any significant effects, supporting the relevant to shallow stakeholders than to deep stakehold-
results below. ers. Consistent with this, as discussed above, integrity
is a significant predictor of trust in shallow but not
in deep relationships. In an additional set of analyses
Results (see Table 3), we tested the relative strength of trust-
To test our hypotheses, we ran several multiple regres- worthiness dimensions (e.g., integrity) across stakeholder
sion models (see Tables 3 and 4). In the first analysis, groups (e.g., deep versus shallow) using interaction terms
we regressed trust perceptions on the six dimensions of for each combination of stakeholder category and each
trustworthiness, using the entire group of stakeholders in dimension of trustworthiness. As predicted (Table 3,
the sample and without any control variables (Table 3, Model 2), integrity is more relevant for stakeholders in
Table 3 Relative Relevance of Trust Dimensions for Different Stakeholder Groups
Model 0 Model 1 Model 2 Model 3
B S.E. B S.E. B S.E. B S.E.
(Constant) 20353∗∗∗ 4000765 00393∗∗∗ 4000475 10183∗∗∗ 4002795 10001∗∗∗ 4002565

Managerial competence 00113∗∗∗ 4000165 00447∗∗∗ 4001085 00269∗∗∗ 4000635
Technical competence 00173∗∗∗ 4000175 00341∗∗ 40015 00559∗∗∗ 4000575
Integrity 00077∗∗∗ 4000205 00335∗∗ 4001155 00327∗∗∗ 4000675
Transparency 00041† 4000215 0018 4001215 00092 4000715
Benevolence 00071∗∗∗ 4000225 00078 4001285 00162∗ 4000785
Identification 00241∗∗∗ 4000165 0075∗∗∗ 4000975 00697∗∗∗ 4000575
Age −00088 4000255 −00073 4000495 −00073 4000495
Gender −00112 4000435 −00123 4000855 −00139 4000845
Multiple stakeholdership −00034 4000445 −0004 4000855 −00037 4000855
Organization 1a 00686∗∗∗ 4000775 −00394∗ 4001635 00399∗ 4001625
Organization 3 00434∗∗∗ 4000535 −00062 4001255 00103 4001225
Organization 4 00392∗∗∗ 4000785 0014 4001615 00177 4001595
Employeesb 00601∗∗∗ 4000525 −00264 4003755 00494∗∗ 4001585
Customer 00158∗∗ 4000645 −00317† 4001275 00308∗ 4001265
Investors 00357∗∗∗ 4000725 −00181 4003795 00413∗ 4001795
Depth 00113 4001315 00159 4001435
Deep/Internal 00159 4001475
Managerial competence × Depth −00046 4001015
Technical competence × Depth −00061 40015
Integrity × Depth −00211† 4001215
Transparency × Depth −00063 4001245
Benevolence × Depth 00267∗ 4001355
Identification × Depth −00104 4000965
Managerial competence × Locus −00218∗ 4001075
Technical competence × Locus 00256∗ 4001095
Integrity × Locus 00021 4001255
Transparency × Locus −00121 4001295
Benevolence × Locus −00026 4001365
Identification × Locus 00019† 4001055
Managerial competence × Deep–Internal 00108 4001245
Technical competence × Deep–Internal −00121 4001275
Integrity × Deep–Internal −00424∗∗ 4001555
Transparency × Deep–Internal 00269† 4001555
Benevolence × Deep–Internal 00140 4001535
Identification × Deep–Internal 00047 4001295
Adjusted R2 00293 00708 00717 00717
a
Organization 2 is the base case.
b
Suppliers are the base case. Stakeholder controls are introduced in Models 2 and 3 instead of a separate locus variable.
†
p < 0010; ∗ p < 0005; ∗∗ p < 0001; ∗∗∗ p < 00001.
shallow relationships than in deep relationships, as indi- These results lend considerable support to the model.
cated by the marginally significant Integrity × Depth As predicted, stakeholders in shallow relationships base
interaction term (b = −00211; p < 0010).7 their trust on integrity (not on benevolence), whereas
Finally, as predicted by Hypothesis 4, benevolence was stakeholders in deep relationship base their trust on
relevant to trust for stakeholders in deep relationships benevolence and not integrity. Consistent with this, we
(internal: b = 00326, p < 0001; external: b = 00288, p = also find that perceptions of integrity matter significantly
more in shallow relationships than in deep relationships.
000671) but not those in shallow relationships (internal:
The one result that represents a significant departure
b = 00047, p = 00729; external: b = 00089, p = 00513).
from our model pertains to the role of transparency:
Because Hypothesis 4 predicted, in part, a null result, we contrary to expectations, transparency was not a sig-
conducted a supplemental analysis in which we tested the nificant predictor of trust in shallow relationships. In
relative strength of benevolence in deep versus shallow a post hoc analysis, we considered the possibility that
relationships (Table 3, Model 2). Consistent with Hypoth- transparency might play a role but only when stakehold-
esis 4, benevolence mattered significantly more in high- ers have extremely few interactions with the organiza-
depth relationships (b = 00267, p < 0005). tion (i.e., in very shallow relationships). To test this, we
Table 4 Absolute Relevance of Trust Dimensions for Different Stakeholder Groups
Internal External
Shallow Deep Shallow Deep
Variable B S.E. B S.E. B S.E. B S.E.
Intercept 20181∗∗∗ 00497 00948 00474 00840 00455 10383∗∗ 00464

Managerial competence 00433∗∗∗ 00113 00465∗∗∗ 00119 00194† 00110 00169 00129
Technical competence 00311∗∗ 00100 00504∗∗∗ 00109 00580∗∗∗ 00102 00694∗∗∗ 00108
Integrity 00478∗∗∗ 00114 −00083 00119 00296∗∗ 00118 00190 00131
Transparency 00095 00124 00347∗∗ 00117 00203 00121 00027 00137
Benevolence 00047 00135 00325∗∗ 00117 00089 00136 00288† 00157
Identification 00687∗∗∗ 00100 00757∗∗∗ 00099 00677∗∗∗ 00100 00726∗∗∗ 00107
Gender −00109 00153 00005 00158 −00160 00162 −00218 00192
Age −00181 00109 −00068 00100 00089 00100 −00160† 00094
Multiple stakeholdership −00071 00139 −00089 00135 00044 00201 −00058 00219
Organization 1 −00077 00284 00231 00290 00692∗ 00300 00474 00505
Organization 2 −00052 00207 −00091 00212
Organization 3 00051 00243 −00099 00418
Organization 4 00020 00224 00037 00206
Employee (base case for internal)
Supplier (base case for external)
Investor −00165 00171 00005 00180
Customer 00198 00194 00566∗ 00216
Adjusted R2 00713 00722 00591 00625
N 265 291 400 342
†
p < 0010; ∗ p < 0005; ∗∗
p < 0001; ∗∗∗
p < 00001.
analyzed a subsample of stakeholders who reported the Identification and Trust

fewest interactions with the organization (1–10) and the We predicted that identification would only be a signif-
shortest duration (<1 year); again, no significant effects icant predictor for trust among internal stakeholders in
of transparency were found. deep relationships with the organization (Hypothesis 7).
Contrary to this prediction, an analysis of the absolute
relevance of trustworthiness dimensions across stake-
Stakeholder Locus: The Relevance of holder types revealed that identification plays a signifi-
Managerial and Technical Competence cant role in all stakeholder relationships (see Table 4):
The model predicted that internal stakeholders will shallow–internal (b = 00687; p < 00001), deep–internal
value managerial competence more so than external (b = 00757, p < 00001), shallow–external (b = 00677,
stakeholders (Hypothesis 5) and that external stake- p < 00001), and deep–external (b = 00726, p < 00001).
holders will value technical competence more so than This intriguing result runs contrary to prior research
internal stakeholders (Hypothesis 6). The results reveal (Lewicki and Bunker 1996, Shockley-Zalabak et al.
that trust among internal stakeholders (employees and 1999), which suggests that identification is a relevant
investors) is based on perceptions of managerial compe- factor in very few relationships. Furthermore, there
tence (shallow: b = 00433, p < 00001; deep: b = 00465, was no significant interaction between identification
p < 00001) and technical competence (shallow: b = and deep–internal stakeholders (b = 00047, p > 001),
00311, p < 0001; deep: b = 00504, p < 00001). Mean- suggesting that identification was not more relevant
while, trust among external stakeholders (customers and to deep–internal stakeholders than to others (Table 3,
suppliers) is based only on perceptions of technical Model 3). Figure 1 summarizes all of our findings.
competence (shallow: b = 00580, p < 00001; deep: b =
00694, p < 00001). Consistent with predictions, we find Discussion
that internal stakeholders rely significantly more so than Prior research on trust within and between organiza-
do external stakeholders on perceptions of managerial tions has not sought to distinguish between the poten-
competence, supporting Hypothesis 5; meanwhile, exter- tially varying dimensions on which stakeholders base
nal stakeholders rely significantly more so than do inter- their trust (e.g., Lusch et al. 2003, Mayer and Davis
nal stakeholders on perceptions of technical competence, 1999, Mayer et al. 1995, Morgan and Hunt 1994). More-
supporting Hypothesis 6 (both interaction terms are sig- over, the surveys that researchers typically employ when
nificant at p < 0005; see Table 3, Model 2). studying stakeholder trust tend to presume, rather than
Figure 1 Relevance of Trustworthiness Dimensions Across constitute the three primary dimensions of trustworthi-
Stakeholder Types ness. Our results support the broad relevance of (at least
some aspect of) ability. We do not find support for the
Benevolence Integrity general relevance of integrity or benevolence; rather,
External
+ +
identification emerges as an independent, fundamental
Technical competence Technical competence
Identification Identification component of organizational trust.
LOCUS
Another intriguing finding concerns perceptions of

Transparency Integrity transparency. As reported above, there was little
Benevolence
Internal
evidence for the relevance of transparency in shallow

Managerial competence Managerial competence
relationships. In a follow-up analysis, and contrary
Technical competence Technical competence
Identification Identification to predictions, we found that perceptions of trans-
parency were only significant predictors of trust for
High Low
DEPTH stakeholders in deep, internal relationships (b = 00347,
Note. Benevolence was significant at p = 00067 (CI = 90%) in high/ p < 0001). This result seems interesting for at least
external relationships, and managerial competence was significant two reasons. First, based on prior research (Dervitsiotis
at p = 00078 (CI = 90%) in external/low relationships. 2003, DiPiazza and Eccles 2002, Sheppard and Sherman
+
Indicates that a specific trustworthiness dimension matters rel- 1998), we would have expected transparency to mat-
atively more in this type of relationship than in another type of
relationship.
ter more so in shallow relationships, where stakeholders
have limited access to first-hand information regarding
the organization’s likely behavior. Second, of the two
examine, which dimensions of trustworthiness are rel- types of internal stakeholders that are in our sample—
evant to trust. The results of our analyses are consis- employees and investors—the latter group is the one that
tent with the perspective that trust is relationship-specific is most often associated with a desire for transparency
(Zey 1998). Moreover, there are identifiable categoriza- (DiPiazza and Eccles 2002, Turnbull 2002). Follow-up
tions along which stakeholders vary, and these categories analysis comparing employees and investors in our sam-
meaningfully distinguish between more and less relevant ple reveals, however, that transparency is a significant
dimensions of trustworthiness. predictor of employee trust (b = 00217, p < 0005) but
Both of the dimensions we studied—depth and locus— not of investor trust. One possible interpretation of these
had significant predictive power. Stakeholders who had results is that whereas transparency may be most needed
shallow relationships with the organization based their by those who lack first-hand information regarding the
trust in the organization largely on perceptions of organization (e.g., those in shallow or external relation-
integrity, whereas trust among deep stakeholders was ships), it may be most valued by those who have the
based on perceptions of benevolence. Thus, whereas most at stake in their relationships with the organization
some research (e.g., Dirks and Ferrin 2001) has com- (i.e., those in deep, internal relationships).
bined benevolence and integrity and treated the combi- Our study also informs stakeholder theory. In addi-
nation as one construct (as in “character”), our results tion to further validating the locus dimension (Freeman
suggest that these are meaningfully distinct (see Mayer 1984, Huse 1998) as a relevant means of distinguish-
ing between stakeholders, we tested and validated depth
et al. 1995), at least in the context of stakeholder trust.
as a relevant dimension in characterizing organizational
The locus dimension also provides insight into how
stakeholders. This paper also represents what may be
organizations might best manage trust across different
the first attempt to theoretically and empirically examine
stakeholders. We found that trust among internal stake- stakeholder-specific foundations of trust and, in doing
holders depended significantly more on perceptions of so, takes a step towards bridging trust research and
managerial competence, whereas trust among external stakeholder theory.
stakeholders was based significantly more on percep- Taken together, our findings extend, and in some
tions of technical competence. cases challenge, existing theories of organizational trust
An interesting—and unexpected—result involved the and stakeholder theory. The results refine and provide
relevance of identification-based trust to all stakeholder nuance to existing conceptualizations of organizational
groups in the sample. Lewicki and Bunker (1996) have trust by highlighting and carefully examining contex-
argued that few relationships are likely to reach a point tual contingencies. In particular, the stakeholder-specific
at which identification is relevant to trust. Our results framework of organizational trust we have developed
suggest that the decision to trust others—even in rela- allows for more accurate predictions regarding the basis
tionships that are not particularly intense—may be more on which stakeholder trust in organizations is actually
personal and more relevant to one’s self-identity than structured.
is typically assumed. This result also challenges and The results also suggest practical implications. In par-
extends the Mayer et al. (1995) model of organizational ticular, organizational actors that are interested in man-
trust that argues that ability, benevolence, and integrity aging trust with various stakeholders might be well
advised to consider the type of relationship they have there are limitations, our analyses challenge some exist-
with the target stakeholder. Rather than to assume the ing beliefs regarding organizational trust and offer a
generalized need to enhance transparency, to engage in number of intriguing avenues for future research. In
acts of benevolence, or to signal competence, organi- particular, the results suggest a need to better under-
zations should seek to understand the specific types of stand the seemingly expansive role of identification, to
attributions that are relevant to the stakeholder whose examine the targeted role of benevolence, to more care-
trust is sought. For example, a company that tries to fully study the distinction between managerial and tech-
project an image that it cares about each of its indi- nical competence, and to further our understanding of
vidual customers or investors (i.e., benevolence) might when transparency might be important. We believe the
be wasting resources; if these are shallow relationships, framework we have developed may provide guidance in
it might more effectively build trust by signalling that approaching and answering these and other important
its management has high ethical standards (i.e., has questions.
integrity). As another example, organizations might try
to focus on building identification across all stakehold- Acknowledgments
ers (and not simply with their key employees and cus- The authors thank J. Keith Murnighan, Michael Jensen, Bruno
tomers). More generally, organizational actors would Fey, and three anonymous reviewers for their feedback and
benefit from better understanding and appreciating the guidance to make the prior versions of this paper even
stronger.
types of vulnerabilities different stakeholders face and
the kinds of information that can most clearly signal Appendix A. Scale Items Measuring Each Construct
trustworthiness to them.
Some limitations of the study warrant mention. First, Managerial Competence
The organization 0 0 0
because this study is based on cross-sectional data,
• can successfully adapt to changing demands.
it is not possible to infer causality. As a result, we • is able to reach set goals.
have focused on understanding the stakeholder-specific
nature of trust (i.e., the dimensions of trustworthi- Technical Competence
ness that matter to different stakeholders) and not on
• is very competent in its area.
temporal antecedents of trust. Second, inherent in the • generally has high standards.
methodology employed (i.e., one survey that yields both
independent and dependent measures) is the risk of com- Integrity
mon method bias. Following the advice of Podsakoff The organization 0 0 0
• does not try to deceive.
et al. (2003), we followed survey procedures to minimize
• has high moral standards.
common method bias and conducted statistical analyses • treats its stakeholder with respect.
that revealed a low risk of such bias in our data, but
such risks cannot be entirely eliminated. Third, whereas Benevolence
our research leverages organizational trust research and
• is caring.
stakeholder theory to identify two key dimensions along • listens to my needs.
which stakeholder trust may vary, there are other rel- • does not abuse stakeholders.
evant dimensions that were not included here. One
such dimension, implicated in the work of Sheppard Identification
• I can identify with the organization.
and Sherman (1998) and Weber et al. (2005), relates
• My personal values match the values of the
to the level of dependence. Whereas all stakeholder organization.
relationships entail interdependence with the organiza- • I feel connected with the organization.
tion, future work could analyze how trustworthiness
Transparency
dimensions vary based on the degree of dependence
or the asymmetry in dependence between the stake- • explains its decisions.
holder and organization. Fourth, our survey included • says if something goes wrong.
stakeholders and organizations from western Europe; • is transparent.
the extent to which the results extend to other institu- • openly shares all relevant information.
tional and cultural contexts is worth studying. Finally,
although beyond the scope of this paper, future research Appendix B. Sample Trust Definitions and
might examine the specific organizational actors or Trustworthiness Dimensions from Interviews
actions stakeholders evaluate when making trust judg-
ments about organizations. Sample Trust Definitions from Stakeholders
Our framework takes an initial step towards creating Employee 1: 0 0 0 it means making myself vulnerable, increasing
a stakeholder model of organizational trust. Although my risk.
Employee 2: Trust is the confidence that someone places in not look attractive from an economic stance, but the values
you to be able to do a task of relevance. match and you want to promote this cause. (I)
(2) I trust (Bank X) more because it has 0 0 0 principles 0 0 0 the
Supplier 1: Trust is reduced when you risk assigning a task to
bank does a lot to support local businesses and they run a
somebody, who [may] not do it.
lot of classes for people [in the community] and [host] local
Supplier 2: The more trust there is, the less need for a
activities. (C).
contract 0 0 0 0
Investor 1: Trust is when you don’t need a 20-page MOU or Endnotes
120-page contract; a handshake is enough. 1
Identification (i.e., a belief regarding shared values) differs
Investor 2: I don’t trust [certain] places 0 0 0 we don’t invest from—and can vary independently of—perceived benevolence
there because it is too risky 0 0 0 0 and integrity. For example, politicians from opposing parties
Customer 1: Depends on the situation 0 0 0 The difference is the can be close friends and believe in each other’s integrity with-
risk involved. out perceiving shared values.
2
Customer 2: One thing that’s important for me to trust is that Of the 32 interviews, 20 were conducted with stakeholders
I feel secure that I’m not taken advantage of. of the four organizations that were targeted for our survey:
six customers, seven employees, four investors, and three sup-
Sample Trustworthiness Dimensions from Employees (E), pliers. The remaining interviews were conducted with stake-
Suppliers (S), Investors (I), and Customers (C) holders of other organizations who were participating in an
unrelated project.
Managerial Competence 3
We understand that there are many types of investors and
(1) The layoff wasn’t large enough to fix the problem. Basi- that a majority owner of a firm is very different from some-
cally [management] was incompetent. (E) one who owns a handful of shares through a mutual fund.
(2) When we invest we only look at companies that are Similarly, there may be suppliers who have few and infre-
managed well. The quality of the management team is very quent dealings with an organization, whereas others are so
important to us, especially in smaller companies. (I) closely tied to the organization that they share an office. In
Technical Competence our framework, and consistent with the work of Schneper and
(1) I trust (Organization X) because the [service works] Guillen (2004), these differences do not impact categorization
well and they get you from A to B safely. (C) on the internal–external dimension. Rather, they are accounted
(2) I trust (Organization X) because they really know their for by our depth dimension.
4
stuff. I could imagine telling my friends to work there, since EM imputation is more appropriate than listwise deletion
they are competent to deliver what they promise, unlike when, as in our data, more than 5% of the cases have one
others 0 0 0 0 (E) or more missing data points, the data are missing at random,
and multivariate normality can be assumed (Allison 2001).
Transparency Notably, listwise deletion and EM imputation provide similar
(1) I cherish open communication. In one case the founder results in our analyses.
presented new numbers, and one of my investment colleagues 5
We tested an alternative classification in which we character-
jumped in and said, “Oh yeah I am already sold,” and then ized as shallow those stakeholders who reported fewer than 50
the founder said, wait one second until you have heard all the interactions or less than three years of contact. Results of this
downsides to this 0 0 0 . (I) analysis were similar to those reported below. We focus on the
(2) When the “s∗∗∗ hits the fan,” I want to know. Will they proposed categorization because this more evenly divides the
tell me? I want to know from them and not via the newspapers. sample between low and high depth.
I want to know everything that is relevant, like in a family. (S) 6
Template analysis uses a research template (such as ABI) for
Integrity the set of initial codes rather than an open-ended coding sys-
(1) Honesty 0 0 0 this is what counts in professional life. Then tem. The coding categories may be revised, added, or deleted
[the relationship] works. (S) during the process based on the data analysis (Maznevski and
(2) What made me trust (Organization X) less is they tried Chudoba 2000).
7
to come up with new projects to sell [the client] and I was As an alternative approach to testing such hypotheses, we can
not sure if there was a need. They tried to take advantage of compare factor coefficients for the trustworthiness dimensions
people to get contracts. (E) across stakeholder types. We ran additional analyses using this
approach and replicated our results.
Benevolence
(1) My computer broke while on warranty, but (Organiza- References
tion X) did not seem to care. I had to call a lot 0 0 0 the computer Aguilera, R. V., G. F. Jackson. 2003. The cross-national diversity
was fixed, but they gave me such a hard time I decided to not of corporate governance: Dimensions and determinants. Acad.
buy there again. (C) Management Rev. 28(3) 447–465.
(2) I trust the organization because in the end they take Allison, P. D. 2001. Missing Data. Sage, Thousand Oaks, CA.
care of you. (E)
Arbuckle, J. 2006. AMOS 7.0 User’s Guide. AMOS Development
Identification Corporation, Mount Pleasant, SC.
(1) There are situations where you feel you should invest Barney, J. B., M. H. Hansen. 1994. Trustworthiness as a source
based on numbers. Still there is the feeling not to since values of competitive advantage. Strategic Management J. 15(S1)
do not match. The reverse also occurred. The opportunity did 175–190.
Becerra, M., A. K. Gupta. 2003. Perceived trustworthiness within Eisenhardt, K. M. 1989. Agency theory: An assessment and review.
the organization: The moderating impact of communication fre- Acad. Management Rev. 14(1) 57–74.
quency on trustor and trustee effects. Organ. Sci. 14(1) 32–44. Elangovan, A. R., D. L. Shapiro. 1998. Betrayal of trust in organiza-
Bigley, G. A., J. L. Pearce. 1998. Straining for shared meaning in tions. Acad. Management Rev. 23(3) 547–567.
organization science: Problems of trust and distrust. Acad. Man-
Enz, C. A. 1988. The role of value congruity in intraorganizational
agement Rev. 23(3) 405–421.
power. Admin. Sci. Quart. 33(2) 284–304.
Blair, M. M. 1995. Ownership and Control: Rethinking Corporate
Ferrell, O. C. 2004. Business ethics and customer stakeholders. Acad.
Governance for the Twenty-First Century. Brookings Institution,
Management Executive 18(2) 126–129.
Washington, DC.
Ferrin, D. L., K. T. Dirks. 2003. The use of rewards to increase
Boersma, M. F., P. J. Buckley, P. N. Ghauri. 2003. Trust in interna-
and decrease trust: Mediating processes and differential effects.
tional joint venture relationships. J. Bus. Res. 56(12) 1031–1042.
Organ. Sci. 14(1) 18–31.
Branzei, O., I. Vertinsky, R. D. Camp. 2007. Culture-contingent signs
Fiske, A. P. 1990. Relativity within Moose culture: Four incommen-
of trust in emergent relationships. Organ. Behav. Human Deci-
surable models for social relationships. Ethos 18(2) 180–204.
sion Processes 104(1) 61–82.
Fombrun, C. J. 1996. Reputation: Realizing Value from the Corporate
Brickson, S. L. 2005. Organizational identity orientation: Forging a
Image. Harvard Business School Press, Boston.
link between organizational identity and organization’s relations
with stakeholders. Admin. Sci. Quart. 50(4) 576–609. Freeman, R. E. 1984. Strategic Management: A Stakeholder
Approach. Pitman, Boston.
Brockner, J., P. A. Siegel, J. P. Daly, T. R. Tyler, C. Martin. 1997.
When trust matters: The moderating effect of outcome favora- Friedman, A. L., S. Miles. 2002. Developing stakeholder theory.
bility. Admin. Sci. Quart. 42(3) 558–583. J. Management Stud. 39(1) 1–21.
Butler, J. K. 1991. Toward understanding and measuring conditions Gaertner, S. L., J. F. Dovidio, B. A. Bachman. 1996. Revisiting the
of trust: Evolution of a conditions of trust inventory. J. Manage- contact hypothesis: The induction of a common group identity.
ment 17(3) 643–663. Internat. J. Intercultural Relations 20(3–4) 271–290.
Chapman, D. W., J. F. Carter. 1979. Translation procedures for the Ganesan, S. 1994. Determinants of long-term orientation in buyer-
cross cultural use of measurement instruments. Ed. Eval. Policy seller relationships. J. Marketing 58(2) 1–19.
Anal. 1(3) 71–76. Giddens, A. 1991. Modernity and Self-Identity: Self and Society in
Coleman, J. S. 1990. Foundations of Social Theory. Harvard Univer- the Late Modern Age. Polity Press, Cambridge, UK.
sity Press, Cambridge, MA.
Goodpaster, K. E. 1991. Business ethics and stakeholder analysis.
Currall, S. C., A. C. Inkpen. 2002. A multilevel approach to trust in Bus. Ethics Quart. 1(1) 53–73.
joint ventures. J. Internat. Bus. Stud. 33(3) 479–495.
Gottman, J. M., R. W. Levenson. 1992. Marital processes predictive of
Currall, S. C., T. A. Judge. 1995. Measuring trust between organi- later dissolution: Behavior, physiology, and health. J. Personality
zational boundary role persons. Organ. Behav. Human Decision Soc. Psych. 63(2) 221–233.
Processes 64(2) 151–170.
Granovetter, M. 1985. Economic action and social structure: The
Davis, J. H., F. D. Schoorman, R. C. Mayer, H. H. Tan. 2000. The problem of embeddedness. Amer. J. Sociol. 91(3) 481–510.
trusted general manager and business unit performance: Empir-
Gulati, R., J. D. Westphal. 1999. Cooperative or controlling? The
ical evidence of a competitive advantage. Strategic Manage-
effects of CEO-board relations and the content of interlocks
ment J. 21(5) 563–576.
on the formation of joint ventures. Admin. Sci. Quart. 44(3)
Dervitsiotis, K. N. 2003. Beyond stakeholder satisfaction: Aiming for 473–506.
a new frontier of sustainable stakeholder trust. Total Quality
Hardin, R. 2002. Trust and Trustworthiness. Russell Sage Foundation,
Management Bus. Excellence 14(5) 515–528.
New York.
DeVellis, R. F. 1991. Scale Development: Theory and Applications.
Sage, Newbury Park, CA. Hodson, R. 2004. Organizational trustworthiness: Findings from the
population of organizational ethnographies. Organ. Sci. 15(4)
Dietz, G., D. N. Den Hartog. 2006. Measuring trust inside organisa- 432–445.
tions. Personnel Rev. 35(5) 557–588.
Hoy, W. K., M. Tschannen-Moran. 1999. Five faces of trust: An
Dillman, D. A. 2000. Mail and Internet Surveys: The Tailored Design empirical confirmation in urban elementary schools. J. School
Method. John Wiley & Sons, New York. Leadership 9(3) 184–208.
DiPiazza, S. A., R. G. Eccles. 2002. Building Public Trust: The Future Huse, M. 1998. Researching the dynamics of board-stakeholder rela-
of Corporate Reporting. John Wiley & Sons, New York. tions. Long Range Planning 31(2) 218–226.
Dirks, K. T. 1999. The effects of interpersonal trust on work group Inkpen, A. C., S. C. Currall. 1998. The nature, antecedents, and con-
performance. J. Appl. Psych. 84(3) 445–455. sequences of joint venture trust. J. Internat. Management 4(1)
Dirks, K. T., D. L. Ferrin. 2001. The role of trust in organizational 1–20.
settings. Organ. Sci. 12(4) 450–467. Jarvenpaa, S. L., K. Knoll, D. E. Leidner. 1998. Is anybody out there?
Donaldson, T., L. E. Preston. 1995. The stakeholder theory of the Antecedents of trust in global virtual teams. J. Management
corporation: Concepts, evidence, and implications. Acad. Man- Inform. Systems 14(4) 29–64.
agement Rev. 20(1) 65–91. Jensen, M. 2003. The role of network resources in market entry:
Doney, P. M., J. P. Cannon. 1997. An examination of the nature of Commercial banks’ entry into investment banking, 1991–1997.
trust in buyer-seller relationships. J. Marketing 61(2) 35–51. Admin. Sci. Quart. 48(3) 466–497.
Edmondson, A. 1999. Psychological safety and learning behavior in Jones, T. M., A. C. Wicks. 1999. Convergent stakeholder theory. Acad.
work teams. Admin. Sci. Quart. 44(2) 350–383. Management Rev. 24(2) 206–221.
Kenning, P. 2001. Customer trust management—Ein Beitrag zum Ver- McAllister, D. J. 1995. Affect- and cognition-based trust as founda-
trauensmanagement im Lebensmitteleinzelhandel. Ph.D. disser- tions for interpersonal cooperation in organizations. Acad. Man-
tation, Universität Münster, Münster, Germany. agement J. 38(1) 24–59.
Kim, P. H., K. T. Dirks, C. D. Cooper, D. L. Ferrin. 2006. When more McDaniels, T. 2005. Regression analysis II. ICPSR Summer Program,
blame is better than less: The implications of internal vs. exter- Inter-University Consortium for Political and Social Research,
nal attributions for the repair of trust after a competence- vs. University of Michigan, Ann Arbor.
integrity-based trust violation. Organ. Behav. Human Decision McKnight, D. H., L. L. Cummings, N. L. Chervany. 1998. Initial trust
Processes 99(1) 49–65. formation in new organizational relationships. Acad. Manage-
King, N. 1998. Template analysis. C. C. G. Symon, C. Cassell, ment Rev. 23(3) 473–490.
eds. Qualitative Methods and Analysis in Organizational Miles, R. E., C. C. Snow. 1992. Causes of failure in network organi-
Research: A Practical Guide. Sage, Thousand Oaks, CA, zations. Calif. Management Rev. 34(4) 53–72.
118–134.
Mishra, A. K. 1996. Organizational responses to crisis: The centrality
Kramer, R. M. 1999. Trust and distrust in organizations: Emerg- of trust. R. M. Kramer, T. R. Tyler, eds. Trust in Organizations:
ing perspectives, enduring questions. Annual Rev. Psych. 50 Frontiers of Theory and Research. Sage, Thousand Oaks, CA,
569–598. 261–287.
Levin, D. Z., E. M. Whitener, R. Cross. 2006. Perceived trustworthi- Mishra, A. K., G. M. Spreitzer. 1998. Explaining how survivors
ness of knowledge sources: The moderating impact of relation- respond to downsizing: The roles of trust, empowerment, justice,
ship length. J. Appl. Psych. 91(5) 1163–1171. and work redesign. Acad. Management Rev. 23(3) 567–588.
Lewicki, R. J., B. B. Bunker. 1995. Trust in relationships: A model of Mitchell, R. K., B. R. Agle, D. J. Wood. 1997. Toward a theory of
trust development and decline. B. B. Bunker, J. Z. Rubin, eds. stakeholder identification and salience: Defining the principle
Conflict, Cooperation, and Justice: Essays Inspired by the Work of who and what really counts. Acad. Management Rev. 22(4)
of Morton Deutsch. Jossey-Bass, San Francisco, 133–173. 853–887.
Lewicki, R. J., B. B. Bunker. 1996. Developing and maintaining Morgan, R. M., S. D. Hunt. 1994. The commitment-trust theory of
trust in work relationships. R. M. Kramer, T. R. Tyler, eds. relationship marketing. J. Marketing 58(3) 20–38.
Trust in Organizations: Frontiers of Theory and Research. Sage,
Nahapiet, J., S. Ghoshal. 1998. Social capital, intellectual capital,
Thousand Oaks, CA, 114–139. and the organizational advantage. Acad. Management Rev. 23(2)
Lewis, J. D., A. Weigert. 1985. Trust as a social reality. Soc. Forces 242–266.
63(4) 967–985. Nooteboom, B. 1996. Trust, opportunism and governance: A process
Luhmann, N. 1979. Trust: A mechanism for the reduction of social and control model. Organ. Stud. 17(6) 985–1010.
complexity. N. Luhmann, ed. Trust and Power. John Wiley & Ogden, S., R. Watson. 1999. Corporate performance and stakeholder
Sons, New York, 4–103. management: Balancing shareholder and customer interests in
Luhmann, N. 2000. Vertrauen: Ein Mechanismus Zur Reduktion the U.K. privatized water industry. Acad. Management J. 42(5)
Sozialer Komplexität, 4th ed. Lucius & Lucius, Stuttgart, 526–538.
Germany. Parmigiani, A., W. Mitchell. 2005. How buyers shape supplier per-
Lusch, R. F., M. O’Brien, B. Sindhav. 2003. The critical role of trust formance: Can governance skills substitute for technical exper-
in obtaining retailer support for a supplier’s strategic organiza- tise in managing out-sourcing relationships? Acad. Management
tional change. J. Retailing 79(4) 249–258. Proc. C1–C6.
Macy, M. W., J. Skvoretz. 1998. The evolution of trust and cooper- Podsakoff, P., S. MacKenzie, J.-Y. Lee, N. Podsakoff. 2003. Com-
ation between strangers: A computational model. Amer. Sociol. mon method biases in behavioral research: A critical review of
Rev. 63(5) 638–660. the literature and recommended remedies. J. Appl. Psych. 88(5)
879–903.
Madhavan, R., R. Grover. 1998. From embedded knowledge to
embodied knowledge: New product development as knowledge Rempel, J. K., J. G. Holmes, M. P. Zanna. 1985. Trust in close rela-
management. J. Marketing 62(4) 1–12. tionships. J. Personality Soc. Psych. 49(1) 95–112.
Malhotra, D., J. K. Murnighan. 2002. The effects of contracts on Rempel, J. K., M. Ross, J. G. Holmes. 2001. Trust and commu-
interpersonal trust. Admin. Sci. Quart. 47(3) 534–559. nicated attributions in close relationships. J. Personality Soc.
Psych. 81(1) 57–64.
Manigart, S., A. Korsgaard, R. Folger, H. Sapienza, K. Bayens. 2002.
The impacts of trust on private equity contracts. Vlerick Work- Ring, P. S., A. H. van de Ven. 1994. Developmental processes of
ing Papers 2002/1, Vlerick Leuven Gent Management School, cooperative interorganizational relationships. Acad. Management
Gent, Belgium. Rev. 19(1) 90–118.
Mayer, R., C., J. H. Davis. 1999. The effect of the performance Robinson, S. L. 1996. Trust and breach of the psychological contract.
appraisal system on trust in management: A field quasi- Admin. Sci. Quart. 41(4) 574–599.
experiment. J. Appl. Psych. 84(1) 123–136. Ross, W., J. Lacroix. 1996. Multiple meanings of trust in negotiation
Mayer, R. C., M. B. Gavin. 2005. Trust in management and perfor- theory and research: A literature review and integrative model.
mance: Who minds the shop while the employees watch the Internat. J. Conflict Management 7(4) 314–360.
boss? Acad. Management J. 48(5) 874–888. Rotter, J. B. 1971. Generalized expectancies for interpersonal trust.
Mayer, R. C., J. H. Davis, F. D. Schoorman. 1995. An integrative Amer. Psych. 26(5) 443–452.
model of organizational trust. Acad. Management Rev. 20(3) Rousseau, D. M., S. B. Sitkin, R. S. Burt, C. Camerer. 1998. Not
709–734. so different after all: A cross-discipline view of trust. Acad.
Maznevski, M. L., K. M. Chudoba. 2000. Bridging space over time: Management Rev. 23(3) 393–404.
Global virtual team dynamics and effectiveness. Organ. Sci. Rowley, T. J. 1997. Moving beyond dyadic ties: A network theory of
11(5) 473–492. stakeholder influences. Acad. Management Rev. 22(4) 887–910.
Schein, E. H. 1985. Organizational Culture and Leadership: Tyler, T. R. 2000. Cooperation in Groups: Procedural Justice,
A Dynamic View. Jossey-Bass, San Francisco. Social Identity, and Behavioral Engagement. Psychology Press,
Schneper, W. D., M. F. Guillén. 2004. Stakeholder rights and cor- Philadelphia.
porate governance: A cross-national study of hostile takeovers. Uzzi, B. 1997. Social structure and competition in interfirm networks:
Admin. Sci. Quart. 49(2) 263–295. The paradox of embeddedness. Admin. Sci. Quart. 42(1) 35–67.
Schoorman, F. D., R. C. Mayer, J. H. Davis. 2007. An integrative Warner, J. B., R. L. Watts, K. H. Wruck. 1988. Stock prices and top
model of organizational trust: Past, present, and future. Acad. management changes. J. Financial Econom. 20(January–March)
Management Rev. 32(2) 344–354. 461–492.
Scott, S. G., V. R. Lane. 2000. A stakeholder approach to organiza-
Weber, J. M., D. Malhotra, J. K. Murnighan. 2005. Normal acts of
tional identity. Acad. Management Rev. 25(1) 43–62.
irrational trust: Motivated attributions and the trust development
Shankar, V., G. L. Urban, F. Sultan. 2002. Online trust: A stake- process. Res. Organ. Behav. 26 75–101.
holder perspective, concepts, implications, and future directions.
J. Strategic Inform. Systems 11(3–4) 325–344. Whitener, E. M., S. E. Brodt, M. A. Korsgaard, J. M. Werner. 1998.
Managers as initiators of trust: An exchange relationship frame-
Shapiro, D., B. Sheppard, L. Cheraskin. 1992. Business on a hand- work for understanding managerial trustworthy behavior. Acad.
shake. Negotiation J. 8(4) 365–377. Management Rev. 23(3) 513–530.
Shaw, R. B. 1997. Trust in the Balance: Building Successful
Wicks, A. C., S. L. Berman, T. M. Jones. 1999. The structure of opti-
Organizations on Results, Integrity and Concern. Jossey-Bass,
mal trust: Moral and strategic implications. Acad. Management
San Francisco.
Rev. 24(1) 99–116.
Sheppard, B. H., D. A. Sherman. 1998. The grammars of trust:
A model and general implications. Acad. Management Rev. Williams, M. 2001. In whom we trust: Group membership as an affec-
23(3) 422–437. tive context for trust development. Acad. Management J. 28(3)
377–396.
Shockley-Zalabak, P., D. D. Morley. 1994. Creating a culture: A lon-
gitudinal examination of the influence of management and Williamson, O. E. 1993. Calculativeness, trust, and economic organi-
employee values on communication rule stability and emer- zation. J. Law Econom. 36(1, Part 2) 453–486.
gence. Human Comm. Res. 20(3) 334–335. Woolsey, L. K. 1986. The critical incident technique: An innovative
Shockley-Zalabak, P., K. Ellis, R. Cesaria. 1999. Measuring orga- qualitative method of research. Canadian J. Counselling 20(4)
nizational trust—Trust and distrust across cultures—The orga- 242–254.
nizational trust index. Report, IABC Research Foundation,
Yaniv, E., F. Farkas. 2005. The impact of person-organization fit on
San Francisco.
the corporate brand perception of employees and of customers.
Sitkin, S. B., N. L. Roth. 1993. Explaining the limited effectiveness of J. Change Management 5(4) 447–461.
legalistic “remedies” for trust/distrust. Organ. Sci. 4(3) 367–392.
Zaheer, A., B. McEvily, V. Perrone. 1998. Does trust matter? Explor-
Spreitzer, G. M., A. K. Mishra. 1999. Giving up control without losing ing the effects of interorganizational and interpersonal trust on
control: Trust and its substitutes’ effects on managers’ involving performance. Organ. Sci. 9(2) 141–159.
employees decision making. Group Organ. Management 24(2)
155–187. Zey, M. 1998. Rational Choice Theory and Organizaiton Theory: A
Critique. Sage, Thousand Oaks, CA.
Swift, T. 2001. Trust, reputation and corporate accountability to stake-
holders. Bus. Ethics: A Eur. Rev. 10(1) 16–26.
Tajfel, H., J. Turner. 1979. An integrative theory of social contact. Michael Pirson is an assistant professor of management
W. Austin, S. Worchel, eds. The Social Psychology of Intergroup systems at Fordham Schools of Business, Fordham Univer-
Relations. Brooks/Cole, Monterey, CA, 33–47. sity, New York. His research focuses on trust and well-being
Tan, H.-T., R. Libby. 1997. Tacit managerial versus technical knowl- in organizational contexts, as well as sustainability manage-
edge as determinants of audit expertise in the field. J. Accounting ment and social entrepreneurship. His work appears in a wide
Res. 35(1) 97–113. range of scholarly journals and various other academic and
Tomlinson, E. C., R. C. Mayer. 2009. The role of causal attribu- practitioner-oriented outlets. He is also a partner at the Human-
tion dimensions in trust repair. Acad. Management Rev. 34(1) istic Management Institute and a founder of the Humanistic
85–104. Management Network.
Tsai, W., S. Ghoshal. 1998. Social capital and value creation: The role Deepak Malhotra is an associate professor of business
of intrafirm networks. Acad. Management J. 41(4) 464–477. administration at the Harvard Business School. His research
Tschannen-Moran, M. 2000. Collaboration and the need for trust. focuses on negotiation strategy, trust development, competitive
J. Ed. Admin. 39(4) 308–331. escalation, and ethnopolitical dispute resolution. He is coau-
Tschannen-Moran, M., W. K. Hoy. 2000. A multidisciplinary analysis thor (with Max Bazerman) of Negotiation Genius: How to
of the nature, meaning and measurement of trust. Rev. Ed. Res. Overcome Obstacles and Achieve Brilliant Results at the Bar-
70(4) 547–593. gaining Table and Beyond, which was awarded the 2008 Out-
Turnbull, S. 2002. A New Way to Govern: Organizations and Society standing Book Award by the International Institute for Conflict
After Enron. New Economics Foundation, London. Prevention and Resolution.
Copyright 2011, by INFORMS, all rights reserved. Copyright of Organization Science: A Journal of the
Institute of Management Sciences is the property of INFORMS: Institute for Operations Research and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.

PirsonMalhotra STKTrust

Uploaded by

Copyright:

Available Formats

PirsonMalhotra STKTrust

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PirsonMalhotra STKTrust

Uploaded by

Copyright:

Available Formats

Organization Science

Vol. 22, No. 4, July–August 2011, pp. 1087–1104

Foundations of Organizational Trust: What Matters to

Table 1 Results of Confirmatory Factor Analysis for Study Variables

Model 2 df NFI RFI IFI TLI CFI RMSEA

1. Managerial 3016 1014 400875

Notes. N = 11298. Alpha coefficients are on the diagonal in parentheses.

Table 3 Relative Relevance of Trust Dimensions for Different Stakeholder Groups

Model 0 Model 1 Model 2 Model 3

B S.E. B S.E. B S.E. B S.E.

(Constant) 20353∗∗∗ 4000765 00393∗∗∗ 4000475 10183∗∗∗ 4002795 10001∗∗∗ 4002565

Table 4 Absolute Relevance of Trust Dimensions for Different Stakeholder Groups

Shallow Deep Shallow Deep

Variable B S.E. B S.E. B S.E. B S.E.

Intercept 20181∗∗∗ 00497 00948 00474 00840 00455 10383∗∗ 00464

analyzed a subsample of stakeholders who reported the Identification and Trust

Another intriguing finding concerns perceptions of

evidence for the relevance of transparency in shallow

You might also like

Model 2 df NFI RFI IFI TLI CFI RMSEA