INITIAL TRUST FORMATION IN NEW ORGANIZATIONAL RELATIONSHIPS
D. HARRISON MCKNIGHT
Florida State University—College of Business
Tallahassee, Florida 32306-1042 USA
Phone: (850) 644-1044
Fax: (850) 644-8225
LARRY L. CUMMINGS
NORMAN L. CHERVANY
University of Minnesota--Curtis L. Carlson School of Management
271-19th Avenue South
Minneapolis, MN 55455 USA
Phone: (612) 624-1520
Fax: (612) 626-1316
Acknowledgments: The authors wish to express appreciation to Ellen Berscheid, Shawn Curley,
Fred Davis, Gerald Smith and Aks Zaheer for their helpful reviews and comments on earlier
versions of this article. Special thanks go to the anonymous reviewers and guest editor of the
Academy of Management Review.
INITIAL TRUST FORMATION IN NEW ORGANIZATIONAL RELATIONSHIPS
Arguably, the most critical timeframe for developing trust between organizational participants is
at the beginning of their relationship. Using primarily a cognitive approach, this article
addresses factors and processes that enable relatively high trust to form initially between two
organizational parties. A model of specific relationships is proposed among several trust-related
constructs and two cognitive processes. The model helps explain the paradoxical finding of high
initial trust levels in new organizational relationships.
Several trust theorists have stated that trust develops gradually over time (e.g., Blau,
1964; Rempel, Holmes & Zanna, 1985; Zand, 1972). When contrasted with some recent
empirical findings, these theories present an interesting paradox. By positing that trust grows
over time, trust theorists implicitly assume that trust levels start small and gradually increase.
Expecting this, some researchers have been surprised at how high their subjects’ early trust
levels were, both in survey- and experimental studies (e.g., Berg, Dickhaut & McCabe, 1995;
Kramer, 1994). For example, economics-based researchers Berg et al. expected subjects to
exhibit low to medium trust in each other when faced with a trust dilemma. Instead, subjects
frequently exhibited high trust by passing to a second specific subject dollars that they were
given during the first part of the experiment. This occurred without any reason to expect that
their generosity would be reciprocated. Kramer (1994) surveyed MBAs who were new to each
other. Because the MBAs had no interaction history, one would expect them to have low trust
levels. However, Kramer found surprisingly high trust levels among them.
The Paradox of High Initial Trust Levels
High initial trust findings are paradoxical because some trust theories predict low initial
trust. By initial, we mean when parties first meet or interact. As an example of initial trust
predictions, economics- or calculative-based trust researchers (e.g., Coleman, 1990; Williamson,
1993) theorize that individuals make trust choices based on rationally-derived costs and benefits
(Lewicki & Bunker, 1995; Shapiro, Sheppard & Cheraskin, 1992). Calculative-based trust
theory would predict that the lack of incentives (benefits) of Berg et al.’s subjects would result in
low levels of trusting behavior among them. Berg et al.’s results did not agree with this
prediction. As another example, knowledge-based trust theory proposes that trust develops over
time as one accumulates trust-relevant knowledge through experience with the other person
(Holmes, 1991; Lewicki & Bunker, 1995). From this perspective, Kramer’s study participants
would require time and an interaction history to develop a high level of trust in each other.
Kramer’s results contradicted what knowledge-based trust theory would predict. Thus, studying
initial trust formation is important because the results from such studies require additional
explanation beyond what calculative-based and knowledge-based trust theories provide.
This article argues that the paradox of high trust in initial relationships may be explained
by identifying the ‘hidden’ factors and processes that enable trust to be high when people in
organizations first meet. A model of initial trust formation is developed to explain why trust
may be high when members of organizations barely know each other.
Increasingly Common New Work Relationships
In today’s work environment, the experience of interacting with a new manager or with
new coworkers is becoming commonplace. These situations involve initial trust. By definition,
the parties have not worked together long enough to develop an interaction history. Such
‘initial’ trust situations occur naturally when an employee or manager is newly hired or
transferred, when cross-functional teams are formed, when salespeople or consultants call, when
mergers bring two sets of employees together, or when a new joint venture begins. Such initial
trust situations are becoming more common because of increased mergers or acquisitions, and
because widespread corporate restructuring and lower employee loyalty have increased the
typical turnover rate of organizational workers and managers (e.g., Evans, Gunz & Jalland,
1996). The nature of tasks is also increasing new work encounters as temporary task teams or
project engagements become the norm in organizations. Meyerson, Weick and Kramer (1996)
explained that such environmental factors as outsourcing and labor/skill shortages have increased
the number of temporary task teams. Further, communication technology now enables millions
of people (Henry & Hartzler, 1997) to work on virtual task teams (Lipnack & Stamps, 1997), in
which participants are often new to each other. Because working together well requires some
level of trust (Bromiley & Cummings, 1995), increasingly common new work encounters
demand that the parties come to trust each other quickly (Meyerson et al., 1996). Thus, the need
exists for a model of how trust initially forms.
A Model of Initial Trust Formation
Initial trust between parties will not be based on any kind of experience with, or first-
hand knowledge of, the other party. Rather, initial trust will be based on one’s disposition to
trust or on institutional cues that enable one to trust the other without first-hand knowledge.
Figure 1 depicts the initial trust formation model the article discusses in more detail. The model
applies only to new encounters between people. Therefore, the model excludes experiential
processes (e.g., observing the trustees’ behavior), but includes cognitive processes and factors
that lead to initial trust. As explained below, the model uses constructs from four trust research
Insert Figure 1 about here
We define trust to mean that one believes in, and is willing to depend on, another party
(e.g., Mayer, Davis & Schoorman, 1995). This high level trust concept decomposes into two
constructs: a) trusting intention, meaning that one is willing to depend on the other person in a
given situation (e.g., Currall & Judge, 1995); and b) trusting beliefs, meaning that one believes
the other person is benevolent, competent, honest, or predictable in the situation (Mayer et al.,
1995). Distinguishing two constructs that constitute trust is important because the word trust is
so confusing (Shapiro, 1987a) and broad (Williamson, 1993) that is almost defies careful
definition (e.g., Gambetta, 1988). The distinction between trusting beliefs and trusting intention
follows the Fishbein & Ajzen (1975) typology that separates constructs into beliefs, attitudes,
intentions, and behaviors. Attitudes and behaviors are excluded here in order to focus the article
on cognitive concerns. At a summary level, the model (Figure 1) implies that trust forms because
of one’s disposition to trust, one’s institution-based trust, and two cognitive processes discussed
later. Disposition to trust refers to a tendency to be willing to depend on others. Institution-
based trust means one believes that impersonal structures support one’s likelihood for success in
the situation. To avoid the vagueness of discussing the formation of ‘trust,’ we use the model to
explore the specific formation of trusting beliefs and trusting intention.
Theoretical Foundations for the Initial Trust Model
A significant body of knowledge from five research streams sheds light on how initial
trust forms. In addition to knowledge-based and calculative-based trust research, three other
research streams have been employed: personality-based, institution-based, and cognition-based.
To personality-based trust researchers, trust develops during childhood as the infant seeks and
receives help from its benevolent caregiver (Bowlby, 1982; Erikson, 1968), resulting in a
general tendency to trust others (Rotter, 1967). Institution-based trust researchers maintain that
trust reflects the security one feels about the situation because of guarantees, safety nets, or other
structures (Shapiro, 1987a; Zucker, 1986). Cognition-based trust researchers say that trust relies
on rapid, cognitive cues or first impressions, rather than on personal interactions (Brewer, 1981;
Lewis & Weigert, 1985b; Meyerson et al., 1996). The personality-, institution-, and cognition-
based research streams can each help explain the paradox of high initial trust levels. Personality-
based trust theorists, for example, would say that subjects in Kramer’s study with high
disposition to trust would have high initial trust. But this may not explain the overall high level
of trust, because it is unlikely that nearly all of Kramer’s subjects had high disposition to trust
levels. Institution-based theorists would argue that the structure of the classroom situation
supported high levels of initial trust by enabling subjects to feel secure in the situation (e.g.,
Lewis & Weigert, 1985a; Shapiro, 1987a). Kramer himself used a cognitive explanation for his
results, attributing the high trust levels to the favorable views the MBAs had towards those of
their own kind.
Whereas each of these three trust research streams can partially explain high initial trust,
focusing on only one of the three presents the danger that the other two will act as hidden
confounds. The danger of hidden confounds exists because, in a given context, all three types of
factors-- personality, institutional and cognitive-- may be present. Sitkin & Pablo (1992)
demonstrated this in the prediction of risk behavior. They identified a hidden personality-related
construct (risk propensity) that explained the paradoxical empirical results researchers found
when predicting risk behavior with situational and organizational variables. By identifying the
hidden construct, Sitkin & Pablo were able to specify more fully the antecedents of risk
behavior. To explain the paradox of high initial trust, the initial trust formation model (Figure 1)
will be detailed and justified.
The detailed model’s constructs and processes come from four of the five research
streams mentioned above (see Figure 2): personality (faith in humanity); institutional
(institution-based trust); calculative (trusting stance); and cognitive (categorization processes and
illusions of control process). Trusting beliefs and trusting intention come from more than one
research stream. The fifth research stream, knowledge-based trust, assumes the parties have
first-hand knowledge of each other, based on an interaction history. Because initial
relationships, by definition, have no interaction history, first-hand knowledge-based trust does
not apply to initial relationships. Hence, first-hand knowledge-based trust formation processes
lie outside the scope of this article. Second-hand knowledge, such as reputation, is addressed as
a categorization process.
Insert Figure 2 about here
The model’s four trusting beliefs were chosen because they are the most commonly used
trusting beliefs in the literature (e.g., Mayer et al., 1995). Each trusting belief will be more
highly correlated with other trusting beliefs than with the model’s other constructs, indicating a
type of convergent validity. The constructs in institution-based trust and disposition to trust will
also display convergent validity. Trusting beliefs will be more highly related to trusting intention
than they are to institution-based trust and disposition to trust, as reflected by the box grouping
them together as a composite trust concept.
The model’s constructs are defined at the individual level of analysis. Hence, the
model’s constructs are internally consistent, even though one set of constructs (institution-based
trust) reflects group- or organization-level phenomena. In limiting the article’s scope to the
individual level of analysis, we are not disallowing social and organizational effects. In fact, we
believe organizational level constructs affect trust. However, these lie outside this article’s scope.
The article’s primary contribution is to create a model that explains why trust can
initially be high. But the model also addresses a second paradox. Whereas trust has often been
termed fragile (e.g., Worchel, 1979), it has also been described as robust (e.g., Luhmann, 1979).
As an added contribution, we later discuss the model’s implications for understanding why trust
is considered both fragile and robust. Research implications for the model are included in the
A MODEL EXPLAINING THE HIGH INITIAL TRUST PARADOX
This section discusses how a model of trust-related concepts explains the high initial trust
Disposition to Trust Affects Trusting Beliefs and Trusting Intention
A person exhibits disposition to trust to the extent that s/he demonstrates a consistent
tendency to be willing to depend on others across a broad spectrum of situations and persons. This
personality construct builds on the work of Erikson (1968) and Rotter (e.g., 1967). In contrast to
others, we distinguish between two types of disposition to trust, each of which affects trusting
intention in a different way: faith in humanity and trusting stance. Reflecting the traditional
view of personality-based trust, faith in humanity means that one believes that others are
typically well-meaning and reliable (e.g., Rosenberg, 1957; Wrightsman, 1991). Trusting stance
means that one believes that irrespective of whether people are reliable or not, one will obtain
better interpersonal outcomes by dealing with people as though they are well-meaning and
reliable. Because it reflects a conscious choice, trusting stance derives from the calculative-
based trust research stream (e.g., Riker, 1971). As an example of trusting stance, one of the
authors asked a respondent if she trusted her new manager, who had just been hired from outside
the company. She replied “yes,” explaining that she generally trusts new people until they give
her some reason not to trust them. Hence, she exhibited a personal strategy to trust newcomers.
Both faith in humanity and trusting stance are dispositional because they reflect tendencies that
apply across various situations. The term ‘dispositional’ refers to personal tendencies, not traits.
Mixed empirical findings. Researchers have experienced mixed results when using
disposition to trust to predict trust. From their research, Johnson-George and Swap (1982)
concluded that constructs like disposition to trust “do not accurately determine an individual’s
trust in another under particular circumstances.” (1982: 1307) By contrast, others have found
disposition-related constructs to be important predictors. For example, Goldsteen, Schorr and
Goldsteen (1989) found that disposition-related trust had a statistically significant effect on
peoples’ mistrust of federal nuclear plant authorities whom they did not know personally.
Mayer, Davis & Schoorman (1995) reviewed additional evidence that disposition-based trust is
important to other trust constructs. However, Holmes (1991) said that although researchers
assume disposition to trust is a contributor to the development of relationship-specific trust, this
link has not been proven.
Resolving the research findings. These mixed results can be resolved by our initial
trust model. The timeframe of the relationship is important in predicting the effects of
disposition to trust. Although other variables will swamp the effects of a person’s trusting
tendency in ongoing relationships, disposition to trust will likely have a significant effect on a
person’s trusting beliefs and trusting intention in new organizational relationships. Johnson-
George and Swap (1982) cited evidence that disposition to trust predicts what they called
trusting behavior only when parties are new to each other in "highly ambiguous, novel, or
unstructured situations, where one's generalized expectancy is all one can rely on" (Johnson-
George & Swap, 1982: 1307; cf. Rotter, 1980). A transferred worker’s relationship with a new
manager in an unfamiliar area would exemplify such a situation, because roles and relationships
would not yet be clear. Goldsteen et al.’s finding--that dispositional trust was related to specific
trust of federal authorities they had never met--supports the idea that dispositional trust is salient
when people don’t know each other.
Faith in humanity effects. Because faith in humanity reflects the extent to which one
believes that nonspecific others are trustworthy, faith in humanity will probably affect one's
initial trusting beliefs (Kramer, 1994). Already-developed patterns of thinking about
relationships in general tend to transfer to a specific initial relationship. This is particularly true
if the person cannot draw on other reasons (e.g., trusting beliefs, institution-based trust) because
the situation, the type of relationship, and the type of other person are new. In other words, if no
more specific situational information is available, one would rely on one's basic beliefs about
human nature (Wrightsman,1991), as reflected in faith in humanity. This is similar to the
argument presented by Mullins & Cummings (1995: 6) that “weak situations” display ambiguity
in terms of the meaning of the situation. In weak situations, the person's disposition will be more
salient than the situation. Initial trust-related situations may be ambiguous because the parties’
roles or task may be new. Rotter (1971) said that the novelty of the situation affects how salient
dispositional trust will be. In a novel situation, then, faith in humanity will enable trusting beliefs
to be high.
Proposition 1: In the initial relationship, to the extent the situation is novel and
ambiguous, faith in humanity will lead to trusting beliefs.
Trusting stance effects. Trusting stance influences one to be intentionally willing to
depend on the other without respect to beliefs in the other. One with high trusting stance
probably believes that things turn out best when one is willing to depend on others, even though
others may, or may not, be trustworthy. Thus, trusting stance does not lead to beliefs about the
other person; rather, it directly supports one's willingness to depend on that person.
Proposition 2: In the initial relationship, trusting stance will lead to trusting intention.
The effects of trusting stance on trusting intention will not be mediated by trusting beliefs.
Institution-based Trust Affects Trusting Intention
Institution-based trust means one believes that the necessary impersonal structures are in
place to enable one to act in anticipation of a successful future endeavor (e.g., Shapiro, 1987a;
Zucker, 1986). Initial relationship trusting intention may be high because of high institution-
based trust levels. Two types of institution-based trust are discussed in the literature: (a)
situational normality—defined as the belief that success is likely because the situation is normal,
and (b) structural assurances—defined as the belief that success is likely because contextual
conditions like promises, contracts, regulations and guarantees are in place. Later, we discuss
the effects of institution-based trust on trusting beliefs. This section looks at direct effects on
Situational normality belief effects. Situational normality belief is based on the
appearance that things are normal (Garfinkel, 1963: 188) or "customary" (Baier, 1986: 245), or
that "everything seems in proper order," (Lewis & Weigert, 1985a: 974). Garfinkel's
experiments demonstrated that trust between people breaks down when people face inexplicable,
abnormal situations. For example, one subject told the experimenter he had a flat tire on the
way to work. The experimenter responded, “What do you mean, you had a flat tire?” (1963:
221) At this point, trust broke down because of the illogical question. Situational normality
involves a properly ordered setting that appears likely to facilitate a successful interaction. For
example, one who enters a bank tends to expect a setting conducive to both customer service and
fiduciary responsibility that is reflected in the workers’ professional appearance, the prosperous
and secure physical setting, and the friendly yet safe money-handling procedures. One’s belief
that the situation is normal helps one feel comfortable enough to rapidly form a trusting
intention towards the other party in the situation. Situational Normality can also relate to one's
comfort with one's own roles and the other person’s roles in that setting (Baier, 1986). Socially
constructed roles create a shared understanding among members of the social system that
facilitates trusting intention among them.
Proposition 3: In the initial relationship, situational normality belief will lead to trusting
Structural assurance belief effects. Shapiro (1987b: 204) referred to structural
safeguards in terms of institutional "side bets," such as regulations, guarantees, and legal
recourse. Regulations enable people to feel assured about their expectations of the other party's
future behavior (e.g., Sitkin, 1995). For example, a company subcontracting the construction of
metric-sized engine parts relies on the other party to use the same measures the company itself
uses, because they are specified by well-accepted metric standards makers. Guarantees mitigate
the perceived risk involved in forming trusting intention. Hence, structural assurance belief
leads to trusting intention. Zaheer, McEvily & Perrone (in press) found empirical evidence that
structure-based trust is positively related to interpersonal trust. Legal recourse (i.e., regarding
contracts or promises) is related to trusting intention for two reasons (Baier, 1986). First, the
trustor feels comfortable that the promise has the type of significance in the particular setting
such that the trusted person will make every effort to fulfill it, or reap sanctions through social
disapproval or legal action (e.g., Sitkin, 1995). Second, the trustor feels comfortable that the
trusted person, out of either socially-learned behavior patterns or fear of sanctions, will act
according to the norms surrounding promise in the social setting. Structural assurance belief will
be more influential in the initial relationship than it will later because information about the
other person is very incomplete when the relationship begins, making situational information
Proposition 4: In the initial relationship, structural assurance belief will lead to trusting
Structural Assurance Belief and Situational Normality Belief Affect Trusting Beliefs
Structural assurance belief is likely to affect trusting beliefs for three reasons, the last two
of which are shared by situational normality belief. First, believing that the situation is bounded
by safeguards enables one to believe that the individuals in the situation are trustworthy. For
example, a new employee can better believe in the boss’s benevolence if the employee believes
that the workplace has procedures that punish abusive managerial treatment. An employee is
more likely to believe a new coworker is competent if they believe that the department’s hiring
process is sound. Second, the institutions in the situation reflect the actions of the people
involved, and therefore beliefs about the institutions will help form beliefs about the people who
are involved in the institutions. For example, within the bounds set by corporate practices and
procedures, the boss is the chief administrator of fairness in the workplace. Therefore, one’s
belief about the structures supporting fairness in the workplace will support one’s belief in the
boss’s benevolence. Similarly, a situational normality belief would imply that the people in the
situation will act normally, and can therefore be trusted. Third, structural assurance belief and
situational normality belief are beliefs, and, based on many studies of cognitive consistency
(Abelson, Aronson, McGuire, Newcomb, Rosenberg & Tannenbaum, 1968), will probably stay
consistent with related beliefs, such as trusting beliefs. Cognitive consistency is probably even
more salient during the initial relationship, before beliefs about the other person and the situation
become highly differentiated through experiential knowledge (Sitkin & Roth, 1993).
Proposition 5: In the initial relationship, the trusting beliefs will be a function of
structural assurance belief and situational normality belief.
Trusting Beliefs Affect Trusting Intention
Evidence for the link between trusting beliefs and trusting intention is reviewed
elsewhere (Mayer et al., 1995). Dobing (1993) also found a strong relationship between his
trusting beliefs and trust (willingness to depend) constructs. Logically, if one believes that the
other party is benevolent, competent, honest, and predictable, one is likely to form a trusting
intention toward that person. Therefore, trusting beliefs will positively impact trusting intention.
At a more general level, the literature that links beliefs to intentions supports this relationship
(e.g., Davis, 1989; Fishbein & Ajzen, 1975). Ajzen (1988) discussed evidence that beliefs and
intentions tend to stay consistent. They should be especially consistent at first, when one has no
experiential basis not to believe the other person is trustworthy.
Proposition 6: In the initial relationship, trusting intention will be a function of
benevolence belief, competence belief, honesty belief, and predictability belief.
Faith in Humanity and Trusting Stance Affect Structural Assurance Belief
Faith in humanity reflects one’s lifelong experiences with others (e.g., Rotter, 1967). A
person who believes in the honesty and benevolence of people generally will probably have
stronger beliefs in the security afforded by human institutions. Stated another way, one’s
structural assurance belief is probably partly based on how one feels about people in general,
because people play roles that relate to how secure the situation is. Hence, one’s feelings about
people in general will likely influence their structural assurance belief. This is more likely to be
true in the initial relationship, when beliefs about the situation are based more on assumptions
than on facts.
Trusting stance will also affect initial structural assurance belief. One with high trusting
stance believes that trusting others facilitates success irrespective of their beliefs about specific
people. This assumption is consistent with a perception that safeguards or safety nets will
protect them from bad consequences that people cause. In other words, developing a high level
of structural assurance belief is facilitated when one has a high level of trusting stance. When
parties first meet, the person’s already-formed high trusting stance level will tend to encourage a
corresponding high level of structural assurance belief. Over time, the relationship between
these constructs may be reciprocal, but at the initial meeting, a person will rely on their prior
tendencies, in terms of trusting stance, to form structural assurance belief.
Proposition 7: In the initial relationship, faith in humanity and trusting stance will lead
to structural assurance belief.
Categorization Processes Enable High Levels of Trusting Beliefs
In a new relationship, a person may use three types of categorization processes to
develop trusting beliefs: (a) unit grouping, (b) reputation categorization, and (c) stereotyping.
Unit grouping means to put the other person in the same category in which one places oneself.
Reputation means that one assigns attributes to another person based on second hand information
about the person. Stereotyping means to place another person into a general category of persons.
Unit Grouping. Because those who are grouped together tend to share common goals
and values, they tend to perceive each other in a positive light (Kramer, Brewer & Hanna, 1996).
Hence, one group member will be more likely to form trusting beliefs towards another group
member. For example, Zucker, Darby, Brewer & Beng (1996) found that being a member of the
same organization generated the trust needed to collaborate on research. Brewer & Silver (1978)
found that people perceived in-group members to be more trustworthy than out-group members.
These studies provide evidence that Unit Grouping leads to high levels of trusting beliefs. Such
beliefs develop quickly when Unit Grouping occurs. For example, in studying prospective dating
couples who had never met, Darley & Berscheid (1967) found that the knowledge that one
would be paired with the other tended to enhance one's beliefs about the other's characteristics.
Applied to a new task team, Unit Grouping would enable one member to immediately form
trusting beliefs about another team member.
Reputation. Those with good reputations are categorized as trustworthy individuals.
Reputation may reflect professional competence (Barber, 1983; Powell, 1996) or the other
trusting beliefs: benevolence (Dasgupta, 1988), honesty, and predictability. The person may be
perceived as a competent individual because s/he is a member of a competent group (Dasgupta,
1988) or because of her/his own actions. So, because the other party has a good reputation, one
will quickly develop trusting beliefs about that individual, even without first-hand knowledge.
Stereotypes. Stereotyping may be done either at broad levels, such as gender (e.g.,
Orbell, Dawes, and Schwartz-Shea, 1994), or at more specific levels, such as prejudices against
occupational groups like used-car salespeople. Stereotypes may be formed at the first meeting
between the parties, based on the other's voice (e.g. male/female, domestic/foreign) (Baldwin,
1992) or physical appearance (Dion, Berscheid & Walster, 1972; Riker, 1971). Stereotyping
enables one to quickly form positive trusting beliefs about the other by generalizing from the
favorable category into which the person was placed.
Proposition 8: In the initial relationship, categorization processes that place the other
person in a positive grouping will tend to produce high levels of trusting beliefs.
Illusions of Control Process: Interactive Effects that Elevate Trusting Beliefs
This section explains why the illusions of control process will interact with
categorization processes, faith in humanity, and structural assurance belief to produce high levels
of trusting beliefs. People in an uncertain situation will perform small actions to try to assure
themselves that things are under their personal control (Langer, 1975). This results in
unrealistically inflated perceptions of personal control (Taylor & Brown, 1988), which Langer
(1975) termed ‘illusions of control.’ Illusions involve perceptions that differ from reality.
Considerable evidence demonstrates the presence of illusion in cognitive processing (e.g., Fiske
& Taylor, 1984).
The illusions of control process that helps build trusting beliefs may be similar to how
people become overconfident of their judgments, as reported in Langer (1975) and Paese &
Sniezek (1991). First, one forms a tentative belief. Second, one watches for clues that confirm
the belief. Even without evidence, the effort of watching tends to overinflate confidence in one’s
judgment, (Davis & Kottemann, 1994). Similarly, even a slight effort to confirm one’s tentative
trusting beliefs in the other may overinflate one's confidence that high levels of trusting beliefs
Token control efforts. As an example, people may make an initial effort to think about
the other person’s trustworthiness. Or, they may, upon meeting the person, immediately attempt
to gauge whether or not they can influence them in some small way (e.g., make them smile).
We term such actions ‘token control efforts.’ The person is not trying to categorize the other,
but, rather, to test their ability to successfully deal with them. Initially, a person is likely to use
token control efforts because s/he does not know from experience whether or not the other has
the attributes needed to be considered trustworthy. After making such small control efforts, the
person may form an unjustifiably strong confidence that one’s positive categorization, faith in
humanity, and structural assurance beliefs are correct, and, therefore, that the other person is
trustworthy. Langer (1975) found, for example, that token control efforts to improve one’s
chances in a lottery situation (i.e., by choosing their own lottery ticket) made one very
overconfident of winning.
Trust theorists have already posited that trust-building involves illusion (Holmes, 1991;
Meyerson et al., 1996). The results of one empirical trust study support the application of
illusion to trust. Kramer (1994) found that ruminating for a few minutes about the others'
motives and intentions increased a person's confidence in the accuracy of their judgments of the
others. Mental assessments would tend to increase one’s confidence because, as Langer (1975)
found, they move a task further from the realm of chance and closer to the realm of a skill-based
judgment task. Hence, token control efforts can support a person's confidence in their trust-
Interactive effects. We propose that token control efforts will interact with
categorization processes, faith in humanity, and structural assurance belief, strengthening their
capacity to form trusting beliefs. Token control efforts will give one the illusion that one’s
positive faith in humanity can apply to the particular other party by convincing one that one is
applying skill, not just chance, to one’s trust-related judgment of the other party. Similarly,
token control efforts will: a) build one’s confidence that one’s positive categorization of the
other person is correct; and b) bolster one’s confidence that structural safeguards make the
environment secure, and by association, the people involved trustworthy.
Proposition 9: In the initial relationship, token control efforts will strengthen the
tendency of categorization processes, faith in humanity, and structural assurance belief to
produce high levels of trusting beliefs.
MODEL IMPLICATIONS FOR TRUSTING INTENTION FRAGILITY / ROBUSTNESS
Paradoxically, whereas trust has often been termed fragile (e.g., Worchel, 1979), it has
also been described as robust, in that it progresses in upward spirals (e.g., Zand, 1972) or
becomes more fully developed over time (Gabarro, 1978; Sitkin & Roth, 1993). The model of
initial trust development improves our understanding of how initial trusting intention, the
model’s ultimate consequent, may be either fragile or robust under different conditions.
Conditions under which Initial trusting intention will Likely be Fragile
Initial trusting intention is likely to be fragile under three conditions: a) inadequate
support from Figure 2’s antecedent constructs, b) the tentative and assumption-based nature of
the antecedent constructs, and c) high perceived risk. The term ‘fragile’ refers to a trusting
intention level that is subject to the likelihood of large changes during a given timeframe. That
is, as time passes and conditions change, to what extent is the level of trusting intention likely to
change? Although researchers’ use of the term fragile typically refers to a high trust level
suddenly becoming low, we define fragile so that it could apply to either high or low trust levels.
At high levels, fragile trust is subject to rapid decreases. At low levels, fragile trust is subject to
rapid increases. In either case, fragile trust is unstable, quickly changeable, or easily influenced.
‘Robust’ (the opposite of fragile) means a trusting intention level that does not change
dramatically during a given timeframe. As an example of the difference between robust and
fragile, if trusting intention’s level moved dramatically downward, it would be termed fragile,
rather than robust.
Inadequate antecedent support. Intuitively, the more highly positive antecedents
trusting intention has, the less fragile trusting intention will be. A weak combination of trusting
intention antecedents would result in a fragile trusting intention. That is, if trusting intention is
associated with only one highly positive antecedent, it is likely to change downward soon after
the initial period. For example, suppose a home buyer with low disposition to trust and low
trusting beliefs in the home builder enters into a contract to have a house built, based on the
belief that the legal processes provide a safety net (structural assurance belief). Further, the
home buyer has no cues by which to categorize the home builder positively. In this situation, the
only highly positive antecedent of trusting intention is structural assurance belief. In the case of
such agreements or contracts, Dasgupta (1988) said that the trust one person has in another to
fulfill a contract rests precariously upon the power of the agencies that are able to enforce
contracts. Therefore, "If your trust in the enforcement agency falters, you will not trust persons
to fulfill their terms of an agreement and thus will not enter into that agreement...It is this
interconnectedness which makes trust such a fragile commodity." (1988: 50) On the other hand,
if the home buyer’s structural assurance belief is accompanied by trusting beliefs in the home
builder, they may be able to maintain trusting intention even when their structural assurance
belief wavers. In general, then, when any particular antecedent of trusting intention is the only
one at a high level, trusting intention would likely be fragile. For example, the home buyer’s
honesty belief is an antecedent of trusting intention that can crumble quickly (Dasgupta, 1988) if
experience with the builder indicates that the belief is mistaken. We previously discussed how
disposition to trust may be a weak trusting intention antecedent in the presence of a strong
situation. The home buyer’s high disposition to trust will not hold up if they have reason to
believe structural assurances are missing or the home builder may be dishonest. Situation- or
person-specific beliefs supporting trusting intention quickly become stronger than dispositional
support as the buyer gains experience with the situation.
Tentative, assumption-based antecedents. Second, trusting intention will be fragile in
the initial relationship because of the tentative and assumption-based nature of its antecedents
(Figure 2). Initial trust is not so much based on evidence as on lack of contrary evidence
(Gambetta, 1988). Illusions of control is almost completely assumption-based. If the illusion
crumbles, the constructs bolstered by it could rapidly decrease, negatively affecting trusting
intention. Riker (1971: 78) pointed out that stereotyping categorization is tentative because
"only rarely do [the applied categories] effectively discriminate between the trustworthy and the
untrustworthy." Hence, trusting beliefs (leading to trusting intention) produced by
categorization are subject to abrupt corrections. Initial institution-based trust, founded on
assumptions about the situation, is subject to rapid deterioration as facts become known.
Disposition to trust assumes things will work out successfully, and will only be salient until
situational or personal facts are uncovered (Johnson-George & Swap, 1982).
Experience supplies facts that can quickly displace illusions and assumptions. Fazio &
Zanna (1981) pointed out two reasons why experience-based facts readily replace assumptions.
First, people consider behavioral experience information to be more reliable than indirectly
obtained information. More reliable information reduces uncertainty, making it highly desirable
(Smith, Benson & Curley, 1991). Second, an individual’s direct experience in forming an
attitude or judgment makes that attitude more readily accessible in memory (cf. Riker, 1971:78).
High perceived risk. Some risk is perceived even when people trust each other (e.g.,
Mayer et al., 1995). High levels of perceived risk make it more likely that the trustor will pay
attention to the other’s behavior, seeking validating information. Kramer (1996) found this to be
true in relationships between students and faculty members. The more one attends to new
information about the other person, the more likely it is that contrary evidence will be found,
negatively affecting trusting intention.
Proposition 10: In the initial relationship, high trusting intention is likely to be very
fragile when: (a) it is supported by only one or two antecedents; (b) it relies almost exclusively
on assumptions; and (c) perceived risk is high.
Several of the above fragility examples demonstrate that the interdependent nature of the
model constructs is itself a reason trusting intention may be fragile. When one construct’s level
is reduced, it is likely to negatively affect a related construct. For example, whereas a decrease
in trusting stance directly affects trusting intention, it also affects institution-based trust.
Whereas institution-based trust will have a direct effect on trusting intention, it will also affect
trusting intention through trusting beliefs. Hence, the model portrays the possibility that trusting
intention is like the roof on the proverbial ‘house of cards’ that collapses when one structurally
key antecedent slips.
Conditions under which Trusting Intention will Likely be Robust
Trusting intention is likely to be robust for three reasons: a) adequate antecedent support; b)
belief-confirming cognitive mechanisms; and c) social mechanisms.
Adequate antecedent support. Given our model, the most obvious reason trusting
intention will continue high is that in many cases, several of the antecedents shown in Figure 2
will exist at high levels. Further, it is more probable that the antecedents will exist at a
consistently high level rather than a combination of high and low levels. Researchers in
cognitive consistency have found evidence that related beliefs tend to stay consistent with each
other because people keep their various cognitions reconciled (Abelson et al., 1968; Luhmann,
1979). Hence, for a given subject, we expect to find relatively consistent levels among trust
constructs, especially as the relationship begins. A consistently high set of antecedents will be
less likely to cause a trusting intention dissolution than will an inconsistently high set.
For example, in a classroom setting, it is likely that, through past experiences, students
feel comfortable in their own, and the instructor’s, roles (as in institution-based trust). Hence,
they may have formed structural assurance beliefs within class situations generally. Hopefully,
they have heard enough about the instructor’s reputation that, along with early cues, they can
categorize her/him as having trustworthy attributes (e.g., competence), thus forming one or more
of the trusting beliefs. Under these circumstances, the student is likely to have high initial
trusting intention toward the instructor. This trusting intention will probably prove stable, given
that is likely supported by trusting beliefs and institution-based trust. Unless the instructor does
something to violate seriously the positive expectations a student develops from the initial class
interaction, a student’s trusting intention is likely to endure throughout the course. But trusting
intention is even more likely to stay high if the student also has high disposition to trust, which
will reinforce their institution-based trust and trusting beliefs. A person with high disposition to
trust is more likely to see the good points and to overlook flaws in the other person or situation
that would threaten high trusting intention levels by lowering trusting beliefs or institution-based
Belief-confirming cognitive mechanisms. Attentional cognitive processes play a role in
sustaining initial trust. Not all information is attended to. Unless it is attended to, it will not
affect trust-related constructs. Peoples’ beliefs and preconceived notions tend to filter the
information that is attended to. Good (1988) said that evidence contrary to one’s beliefs is
seldom sought and often ignored. Good cited psychological experimental evidence (e.g.,
Mitroff, 1974; Tajfel, 1969; Wason, 1960) for this confirmation bias effect. Similarly, Taylor
and Brown (1988) cited evidence that people “generally select, interpret, and recall information
to be consistent with their prior beliefs or theories (see Fiske & Taylor, 1984; Greenwald, 1980;
Taylor & Crocker, 1981, for reviews).” Taylor and Brown (1988) applied this effect to positive
illusion-based beliefs: “Consequently, if a person’s prior beliefs are positive, cognitive biases
that favor conservatism generally will maintain positive illusions more specifically.” (1988: 202)
Taylor & Brown (1988) cited additional evidence that peoples’ preconceptions guide which
information they considered relevant, and therefore, to which information they attended
(Howard & Rothbart, 1980; Nisbett & Ross, 1980).
Empirical studies have confirmed that much counter-belief evidence is simply ignored.
An example is the 1986 NASA Challenger disaster. Starbuck and Milliken (1988) found that
NASA and Morton-Thiokol managers ignored or explained away evidence that rocket booster
O-rings would erode in a low temperature take-off. Contrary evidence is especially ignored
when things are perceived to be going well. Sitkin (1992: 232) explained that “small successes
may unintentionally induce low levels of attention and reduced information search.” By
contrast, “The greater the incidence of small prior failure, the more attention will be paid to and
the more deeply will [be] the processing of information about potential problems.” (1992: 240)
The tendency to ignore counter-belief evidence should be true of trusting beliefs as well as other
beliefs. However, in the initial relationship, the desire to feel assured through experiential
evidence (Fazio & Zanna, 1981) will, to some extent, mitigate this cognitive bias. This desire
would be heightened by low levels of disposition to trust or institution-based trust or high levels
of perceived risk, because the person would require additional assurances that their initial beliefs
When people attend to information that disconfirms their views, they often discount it as
inaccurate or uninformative (Markus, 1977; Ross & Anderson, 1992; Swann & Read, 1981;
Taylor & Brown, 1988) or reinterpret it positively (Holmes, 1991; Robinson, 1996). Ross and
Anderson (1992: 144) pointed out that “beliefs...are remarkably resilient in the face of empirical
challenges that seem logically devastating.” People tend to accept belief-supportive information
uncritically, but slowly acknowledge disconfirming evidence (Tetlock, 1985). Ross & Anderson
(1992) said people subject disconfirming evidence to considerable scrutiny. People do not go
back and update or re-evaluate evidence relevant to their beliefs based on disconfirming
evidence. People also search their memories to find ways to explain their existing beliefs (Ross
& Anderson, 1992). Similarly, Kahneman and Tversky (1973) found significant evidence that
people will interpret ambiguous or incomplete information to agree with their pre-existing
beliefs. If one has a high level of trusting intention toward another, for example, then one can
view specific trust violations as isolated exceptions or as a personal quirk (Sitkin & Roth, 1993;
Zucker, 1986), with no resulting negative effect on trusting beliefs. A high faith in humanity
level would provide another reason to ignore or discount such behavior, because it assumes that
most people are basically good. Having a high faith in humanity level would facilitate quick
forgiveness of trust violations.
Good (1988) suggested that another reason for this ‘cognitive inertia’ is the set effect.
The set effect refers to the continued use of mentally stored procedures to handle a situation even
when the situation changes. That is, once people have developed a situational strategy, they tend
to continue to use it, even when it does not work. Applied to trust, this would mean that one
would continue to trust another even when they breach one’s trust--at least for a while. Good
(1988) used the set effect to reinterpret some of the results of experimentalists who studied trust
with such procedures as the Trucking Game. Note the interactive effect of this cognitive
tendency with the model’s constructs: high institution-based trust and disposition to trust levels
would jointly encourage a person to believe such a course of action is not very risky.
Luhmann (1979) suggested a related reason trust may be resilient. He said that people
build up mechanisms to handle refutations of their trust decisions. This is especially true when
the felt security associated with trusting intention is not strong: “Insecure expectations, however
paradoxical it may at first appear, are psychologically more stable [than secure ones]....[They
are] normalized, stereotyped and thus in various ways immunized against the refutation.
Explanations of disappointment are built into [them] in such a way that a particular case of
disappointment presents no problem but rather confirms the structure of the expectation as a
whole.” (Luhmann, 1979: 79) Hence, it appears that people develop mechanisms that enable
them to absorb disappointments as part of their expectation of the other person, thus reducing the
effect of the disappointment on trusting intention. We speculate that individuals with high
disposition to trust levels will absorb disappointments better than those with low levels.
Proposition 11: In the initial relationship, high trusting intention is likely to be robust
when: (a) a combination of several of its antecedents encourages the trustor to ignore,
rationalize, or absorb the negative actions of the other; (b) continued success or low perceived
risk of failure consequences cause little critical attention to be paid to the other’s behavior.
Subjects with high disposition to trust or institution-based trust levels are likely to pay even less
Social mechanisms. Good (1988) noted that being around the other person will generally
increase favorable beliefs about them. This occurs because the interpersonal cues from the other
person are generally harder to misconstrue face-to-face, and because the pair can more easily go
beyond surface information about the other to more substantive levels of mutual understanding.
Hence, high trusting intention levels will likely be sustained as people interact in cooperative
ways. If they hold positive beliefs about the other party, they are not likely to decrease
interaction (Darley & Fazio, 1980). Therefore, the trusting cycle becomes self-sustaining.
Social interaction will also tend to uphold early trusting intention because people in
social situations tend to confirm their beliefs about themselves (Swann, 1983) and about the
other party. For example, one who trusts another will tend to express that trust by their actions
toward the other. Because initially extended trust is usually reciprocated (e.g., Burt & Knez,
1996), the other party will also express trust. This confirms the first party’s trusting beliefs, as
Dasgupta (1988) noted, which, in turn, supports continued high levels of trusting intention.
Social interaction also upholds early trusting intention because of reputation effects. A
person’s reputation spreads gradually (Dasgupta, 1988) through social interaction (Burt & Knez,
1996). One party remembers the other party’s previous encounters. This interactional history
cumulates, along with information about the person’s background (Dasgupta, 1988), and is
transmitted to others. When many people perceive that one has a good reputation, it is harder
for a negative event to significantly reduce a high level of trusting beliefs.
We believe that social interaction also sustains trusting intention through institution-
based trust. When parties interact in a cordial way, they establish a feeling and appearance that
everything is normal or in proper order (Lewis & Weigert, 1985a). Hence, social interaction
sustains situational normality belief. Through situational normality belief, the parties’
interaction strengthens trusting intention. If one has high levels of situational normality belief,
one is also likely to believe that structural assurances will operate properly in the situation.
Structural assurance will, in turn, positively influence trusting beliefs and trusting intention, as
Proposition 12: In the initial relationship, high trusting intention is likely to be robust
when: (a) the parties interact face-to-face and frequently in positive ways; or (b) the trusted party
has built a widely-known good reputation. Social interaction affects trusting intention by its
positive effects on trusting beliefs and institution-based trust.
RESEARCH IMPLICATIONS AND CONCLUSION
Future Research Implications
Whereas the interplay between trusting intention and its antecedents helps enlighten the
paradox of high trust in new relationships and situations (e.g., Kramer, 1994), the model can also
help explain “disturbing” results in game theory research (Baxter, 1972: 100), a well-spring for
trust theory historically. Baxter’s review of the two-person game theory research indicated that
researchers had found no solid link between trusting personality and trusting behavior in the
Prisoner’s Dilemma game. Because our model ties dispositional variables to beliefs and
intentions, it suggests that personality variables may be too distal from behaviors to be
predictive. Our model’s time boundary suggests that disposition-related trust will only be salient
when the parties first meet in an ambiguous situation. Further, the institution-based trust
constructs help explain why no solid link was found by game theorists. Whereas game theorists
assume that the game context effectively restricts the effects of situational variables (Baxter,
1972), our model’s use of institution-based trust suggests that what a subject believes about the
situation is an empirical question that should be measured as a potential confound (e.g., Erez,
1992), rather than being assumed away. Over twenty-five years ago, Kee & Knox (1970: 365)
recommended that Prisoner's Dilemma research should include "continuous" measures of
subjects’ cognitions in addition to their behavioral choices. Very little trust research has been
done in this manner. Our model's constructs provide a reasoned theoretical basis for doing so.
We suggest that researchers replicate early behavioral experiments while measuring our
constructs. This will enable researchers to reinterpret historical results.
Significant empirical work is needed to gather evidence regarding this article’s
propositions. First, reliable and valid instruments reflecting the constructs in this article should
be developed. These instruments should then be used to test various subsets of the model
through questionnaire studies. For example, Propositions 1-2 and 4-7 could be tested in one
study by measuring trusting intention, faith in humanity, trusting stance, structural assurance
belief, and benevolence belief. Second, the categorization and illusion propositions should be
tested in laboratory settings, incorporating disposition to trust and institution-based trust
constructs as control variables. Third, researchers should test the model’s interactive effects.
Scholars should test whether disposition to trust constructs moderate the effects of categorization
on trusting beliefs. Based on Proposition 1, one could test whether or not structural assurance
belief moderates the link between faith in humanity and trusting beliefs. Trusting stance and
institution-based trust may also have an interactive effect on trusting intention. The combination
of a high situational normality belief and a high structural assurance belief would probably
produce higher trusting beliefs than either would alone.
Finally, the fragility/robustness of initial trust should be tested, based on our
propositions, which argued that initial trust may be either fragile or robust under certain
conditions. For example, one could test through longitudinal laboratory experiments the extent
to which social interaction increases the robustness of trusting intention when: a) the
relationship is successful/unsuccessful, or b) risk is high/low. Many more conditions could also
be posited. This article’s distinction between trust levels and trust fragility/robustness raises
important issues that merit further conceptual and empirical work. As an example, a high initial
level of trusting intention may be quite fragile (subject to change), whereas a low initial level of
trusting intention may not be very fragile. In this respect, an interesting research question is,
‘under what conditions are low initial levels of trusting intention less fragile than high initial
levels of trusting intention?’
The article’s primary contribution is to explain the high initial trust paradox by
synthesizing a model of constructs and processes from diverse trust research streams. Because
initial trusting intention is not always high, the model resolves the paradox by pointing out why
it could be high initially, but may not be high because of the situation or the persons involved.
Thus, the model is predictive based upon specified conditions related to the antecedents of
trusting intention. This model, along with the discussion of the conditions producing fragile
versus robust initial trust, should generate significant amounts of research. The delineation of
two specific types of institution-based trust and disposition to trust constructs provides
researchers construct definitions that lend themselves to consistent empirical measurement.
Using both higher level constructs (e.g., trusting beliefs) and lower level constructs (e.g.,
honesty belief) helps organize the “confusing potpourri” (Shapiro, 1987a: 625) of trust construct
definitions with which researchers must grapple.
Our model integrates different aspects of trust that have not previously been linked, and
does so within a specific time parameter. Some have called for more integrative models that
simultaneously address dispositional and situational constructs (Davis-Blake & Pfeffer, 1989;
Sitkin & Pablo, 1992). Answering this call, this article’s model brings together dispositional,
situational, and interpersonal constructs from four divergent research streams. Poole & Van de
Ven (1989) said that adding the temporal dimension can improve theories. This article explains
that the processes by which trust forms initially are not the same as those by which it forms later.
In particular, a model of continuing trust would emphasize experiential knowledge while de-
emphasizing assumptions and dispositions. In this way, the model’s temporal lens highlights the
unique aspects of how trust forms at the earliest stage of an organizational relationship.
Abelson, R. P., Aronson, E., McGuire, W. J., Newcomb, T. M., Rosenberg, M. J. &
Tannenbaum, P. H. (Eds.). 1968. Theories of cognitive consistency: A sourcebook.
Ajzen, I. 1988. Attitudes, personality, and behavior. Chicago: The Dorsey Press.
Baier, A. 1986. Trust and antitrust. Ethics, 96: 231-260.
Baldwin, M. W. 1992. Relational schemas and the processing of social information.
Psychological Bulletin, 112: 461-484.
Barber, B. 1983. The logic and limits of trust. New Brunswick, NJ: Rutgers University Press.
Baxter, G. W. 1972. Personality and attitudinal characteristics and cooperation in two-person
games: A review. In L. S. Wrightsman, J. O’Connor & N. J. Baker (Eds.) Cooperation
and competition: Readings on mixed-motive games: 97-103. Belmont, CA:
Berg, J., Dickhaut, J. & McCabe, K. 1995. Trust, reciprocity, and social history. Unpublished
working paper, Department of Accounting, Carlson School of Management, University
Blau, P. M. 1964. Exchange and power in social life. New York: John Wiley & Sons.
Bowlby, J. 1982. Attachment and loss, Volume 1: Attachment. New York: Basic Books.
Brewer, M. B. 1981. Ethnocentrism and its role in interpersonal trust. In M. B. Brewer & B. E.
Collins (Eds.), Scientific inquiry and the social sciences: 214-231. San Francisco:
Brewer, M. B. & Silver, M. 1978. Ingroup bias as a function of task characteristics. European
Journal of Social Psychology, 8: 393-400.
Bromiley, P. & Cummings, L. L. 1995. Transactions costs in organizations with trust. In R.
Bies, B. Sheppard & R. Lewicki (Eds.), Research on negotiations in organizations:
5: 219-247. Greenwich, CT: JAI.
Burt, R. S. & Knez, M. 1996. Trust and third-party gossip. In R. M. Kramer & T. R. Tyler
(Eds.), Trust in organizations: Frontiers of theory and research: 68-89. Thousand
Oaks, CA: Sage.
Coleman, J. S. 1990. Foundations of social theory. Cambridge, MA and London: Harvard
Currall, S. C. & Judge, T. A. 1995. Measuring trust between organizational boundary role
persons. Organizational Behavior and Human Decision Processes, 64: 151-170.
Darley, J. M. & Berscheid, E. 1967. Increased liking as a result of anticipation of personal
contact. Human Relations, 20: 29-40.
Darley, J. M. & Fazio, R. H. 1980. Expectancy confirmation processes arising in the social
interaction sequence. American Psychologist, 35: 867-881.
Dasgupta, P. 1988. Trust as a commodity. In D. Gambetta (Ed.), Trust: Making and breaking
cooperative relations: 47-72. New York: Blackwell.
Davis, F. D. 1989. Perceived usefulness, perceived ease of use, and user acceptance of
information technology. MIS Quarterly, 13: 319-340.
Davis, F. D. & Kotteman, J. E. 1994. User perceptions of decision support effectiveness: Two
production planning experiments. Decision Sciences, 25: 57-77.
Davis-Blake, A. & Pfeffer, J. 1989. Just a mirage: The search for dispositional effects in
organizational research. Academy of Management Review, 14: 385-400.
Dion, K. K., Berscheid, E. & Walster, E. 1972. What is beautiful is good. Journal of
Personality and Social Psychology, 24: 285-290.
Dobing, B. 1993. Building trust in user-analyst relationships. Unpublished doctoral
dissertation, University of Minnesota, Minneapolis.
Erez, M. 1992. Reflections on the Latham/Erez/Locke Study. In P. J. Frost & R. E. Stablein
(Eds.), Doing exemplary research: 155-164. Newbury Park, CA: SAGE.
Erikson, E. H. 1968. Identity: Youth and crisis. New York: W. W. Norton.
Evans, M. G., Gunz, H. P. & Jalland, R. M. 1996. The aftermath of downsizing: A cautionary tale
about restructuring and careers. Business Horizons, 39: 61-66.
Fazio, R. H. & Zanna, M. P. 1981. Direct experience and attitude-behavior consistency. In L.
Berkowitz (Ed.), Advances in experimental social psychology, 14: 162-202. New
York: Academic Press.
Fishbein, M. & Ajzen, I. 1975. Belief, attitude, intention and behavior: An introduction to
theory and research. Reading, MA: Addison-Wesley.
Fiske, S. T. & Taylor, S. E. 1984. Social cognition. Reading, MA: Addison-Wesley.
Gabarro, J. J. 1978. The development of trust, influence, and expectations. In A. G. Athos & J. J.
Gabarro (Eds.), Interpersonal behavior: Communication and understanding in
relationships: 290-303. Englewood Cliffs, NJ: Prentice-Hall.
Gambetta, D. 1988. Can we trust trust? In D. Gambetta (Ed.), Trust: Making and breaking
cooperative relations: 213-237. New York: Blackwell.
Garfinkel, H. 1963. A conception of, and experiments with, “trust” as a condition of stable
concerted actions. In O. J. Harvey (Ed.), Motivation and social interaction: 187-238.
New York: Ronald Press.
Goldsteen, R., Schorr, J. K. & Goldsteen, K. S. 1989. Longitudinal study of appraisal at Three
Mile Island: Implications for life event research. Social Science and Medicine, 28:
Good, D. 1988. Individuals, interpersonal relations, and trust. In Gambetta, D. (Ed.), Trust:
Making and breaking cooperative relations: 31-48. New York: Blackwell.
Greenwald, A. G. 1980. The totalitarian ego: Fabrication and revision of personal history.
American Psychologist, 35: 603-618.
Henry, J. E. and Hartzler, M. 1997. Virtual teams: Today’s reality, today’s challenge. Quality
Progress, 30(5): 108-109.
Holmes, J. G. 1991. Trust and the appraisal process in close relationships. In W. H. Jones & D.
Perlman (Eds.), Advances in personal relationships, 2: 57-104. London: Jessica
Howard, J. W. & Rothbart, M. 1980. Social categorization and memory for ingroup and
outgroup behavior. Journal of Personality and Social Psychology, 38: 301-310.
Hunter, J. E. & Gerbing, D. W. 1982. Unidimensional measurement, second order factor
analysis, and causal models. In B. M. Staw & L. L. Cummings (Eds.), Research in
Organizational Behavior, 4: 267-320. Greenwich, CN: JAI Press.
Johnson-George, C. & Swap, W. C. 1982. Measurement of specific interpersonal trust:
Construction and validation of a scale to assess trust in a specific other. Journal of
Personality and Social Psychology, 43: 1306-1317.
Kahneman, D. & Tversky, A. 1973. On the psychology of prediction. Psychological Review,
Kee, H. W. & Knox, R. E. 1970. Conceptual and methodological considerations in the study of
trust and suspicion. Journal of Conflict Resolution, 14: 357-366.
Kramer, R. M. 1994. The sinister attribution error: Paranoid cognition and collective distrust in
organizations. Motivation and Emotion, 18: 199-230.
Kramer, R. M. 1996. Divergent realities and convergent disappointments in the hierarchic
relation: Trust and the intuitive auditor at work. In R. M. Kramer & T. R. Tyler (Eds.),
Trust in organizations: Frontiers of theory and research: 216-245. Thousand Oaks,
Kramer, R. M., Brewer, M. B. & Hanna, B. A. 1996. Collective trust and collective action: The
decision to trust as a social decision. In R. M. Kramer & T. R. Tyler (Eds.), Trust in
organizations: Frontiers of theory and research: 357-389. Thousand Oaks, CA: Sage.
Langer, E. J. 1975. The illusion of control. Journal of Personality and Social Psychology, 32:
Lewicki, R. J. & Bunker, B. B. 1995. Trust in relationships: A model of trust development and
decline. In B. B. Bunker & J. Z. Rubin (Eds.), Conflict, cooperation and justice: 133-
173. San Francisco: Jossey-Bass.
Lewis, J. D. & Weigert, A. J. 1985a. Trust as a social reality. Social Forces, 63: 967-985.
Lewis, J. D. & Weigert, A. J. 1985b. Social atomism, holism, and trust. The Sociological
Quarterly, 26: 455-471.
Lipnack, J. & Stamps, J. 1997. Virtual teams: Reaching across space, time and organization
with technology. New York: John Wiley & Sons.
Luhmann, N. 1979. Trust and power. New York: John Wiley.
Markus, H. 1977. Self-schemata and processing information about the self. Journal of
Personality and Social Psychology, 35: 63-78.
Mayer, R. C., Davis, J. H. & Schoorman, F. D. 1995. An integrative model of organizational
trust. Academy of Management Review, 20: 709-734.
Meyerson, D., Weick, K. E. & Kramer, R. M. 1996. Swift trust and temporary groups. In R. M.
Kramer & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and research:
166-195. Thousand Oaks, CA: Sage.
Mitroff, I. I. 1974. The subjective side of science. Amsterdam: Elsevier.
Mullins, J. W. & Cummings, L. L. In press. Situational strength: A framework for
understanding the role of individuals in bringing about proactive strategic change.
Journal of Organizational Change Management.
Nisbett, R. E. & Ross, L. 1980. Human inference: Strategies and shortcomings of social
judgments. Englewood Cliffs, NJ: Prentice-Hall.
Orbell, J., Dawes, R. & Schwartz-Shea, P. 1994. Trust, social categories, and individuals: The
case of gender. Motivation and Emotion, 18: 109-128.
Paese, P. W. & Sniezek, J. A. 1991. Influences on the appropriateness of confidence in
judgment: Practice, effort, information, and decision-making. Organizational Behavior
and Human Decision Processes, 48: 100-130.
Poole, M. S. & Van de Ven, A. H. 1989. Using paradox to build management and organization
theories. Academy of Management Review, 14: 562-578.
Powell, W. W. 1996. Trust-based forms of governance. In R. M. Kramer & T. R. Tyler (Eds.),
Trust in organizations: Frontiers of theory and research: 51-67. Thousand Oaks, CA:
Rempel, J. K., Holmes, J. G., & Zanna, M. P. 1985. Trust in close relationships. Journal of
Personality and Social Psychology, 49: 95-112.
Riker, W. H. 1971. The nature of trust. In J. T. Tedeschi (Ed.), Perspectives on social power:
63-81. Chicago: Aldine Publishing Company.
Robinson, S. L. 1996. Trust and breach of the psychological contract. Administrative Science
Quarterly, 41: 574-599.
Rosenberg, M. 1957. Occupations and values. Glencoe, IL: Free Press.
Ross, L. & Anderson, C. A. 1982. Shortcomings in the attribution process: On the origins and
maintenance of erroneous social assessments. In D. Kahneman, P. Slovic & A. Tversky
(Eds.), Judgment under uncertainty: Heuristics and biases: 129-152. Cambridge:
Cambridge University Press.
Rotter, J. B. 1967. A new scale for the measurement of interpersonal trust. Journal of
Personality, 35: 651-665.
Rotter, J. B. 1971. Generalized expectancies for interpersonal trust. American Psychologist,
Rotter, J. B. 1980. Interpersonal trust, trustworthiness, and gullibility. American Psychologist,
Shapiro, S P. 1987a. The social control of impersonal trust. American Journal of Sociology,
Shapiro, S. P. 1987b. Policing trust. In C. D. Shearing & P. C. Stenning (Eds.), Private
policing: 194-220. Newbury Park, CA: Sage.
Shapiro, D. L., Sheppard, B. H., & Cheraskin, L. 1992. Business on a handshake. Negotiation
Journal, 3: 365-377.
Sitkin, S. B. 1992. Learning through failure: The strategy of small losses. In B. Staw & L. L.
Cummings (Eds.), Research in Organizational Behavior, 14: 231-266. Greenwich, CN:
Sitkin, S. B. 1995. On the positive effect of legalization on trust. In R. Bies, B. Sheppard & R.
Lewicki (Eds.), Research on negotiations in organizations: 5: 185-217. Greenwich,
Sitkin, S. B. & Pablo, A. L. 1992. Reconceptualizing the determinants of risk behavior.
Academy of Management Review, 17: 9-38.
Sitkin, S. B. & Roth, N. L. 1993. Explaining the limited effectiveness of legalistic "remedies"
for trust/distrust. Organization Science, 4: 367-392.
Smith, G. F., Benson, P. G. & Curley, S. P. 1991. Belief, knowledge, and uncertainty: A
cognitive perspective on subjective probability. Organizational Behavior and Human
Decision Processes, 48: 291-321.
Starbuck, W. H. & Milliken, F. J. 1988. Challenger: Fine-tuning the odds until something
breaks. Journal of Management Studies, 25: 319-340
Swann, W. B. Jr. 1983. Self-verification: Bringing social reality into harmony with the self. In
J. Suls & A. G. Greenwald (Eds.), Social psychology perspectives, 2: 33-66. Hillsdale,
Swann, W. B. Jr. & Read, S. J. 1981. Self-verification processes: How we sustain our self-
conceptions. Journal of Experimental Social Psychology, 17: 351-370.
Tajfel, H. 1969. Cognitive aspects of prejudice. Journal of Social Issues, 25: 79-97.
Taylor, S. E. & Brown, J. D. 1988. Illusion and well-being: A social psychological perspective
on mental health. Psychological Bulletin, 103: 193-210.
Taylor, S. E. & Crocker, J. 1981. Schematic-bases of social information processing. In E. T.
Higgins, C. P. Herman, & M. P. Zanna (Eds.), Social cognition: The Ontario
symposium, 1: 89-134. Hillsdale, NJ: Erlbaum.
Tetlock, P. 1985. Accountability: The neglected social context of judgment and choice. In L. L.
Cummings and B. M. Staw (Eds.), Research in organizational behavior, 7: 297-332.
Greenwich, CN: JAI Press.
Wason, P. C. 1960. On the failure to eliminate hypotheses in a conceptual task. Quarterly
Journal of Experimental Psychology, 12: 129-140.
Williamson, O. E. 1993. Calculativeness, trust, and economic organization. Journal of Law and
Economics, 34: 453-502.
Worchel, P. 1979. Trust and distrust. In W. G. Austin & S. Worchel (Eds.), The social
psychology of intergroup relations: 174-187. Monterrey, CA: Brooks/Cole Publishing.
Wrightsman, L. S. 1991. Interpersonal trust and attitudes toward human nature. In J. P.
Robinson, P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of personality and
social psychological attitudes: Vol. 1: Measures of social psychological attitudes:
373-412. San Diego, CA: Academic Press.
Zaheer, A., McEvily, B. & Perrone, V. In press. Does trust matter? Exploring the effects of
interorganizational and interpersonal trust on performance. Organizational Science.
Zand, D. E. 1972. Trust and managerial problem solving. Administrative Science Quarterly,
Zucker, L. G. 1986. Production of trust: Institutional sources of economic structure, 1840-1920.
In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behavior, 8: 53-
111. Greenwich, CN: JAI Press.
Zucker, L. G., Darby, M. R., Brewer, M. B. & Peng, Y. 1996. Collaboration structure and
information dilemmas in biotechnology: Organizational boundaries as trust production.
In R. M. Kramer & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and
research: 90-113. Thousand Oaks, CA: Sage.
High-Level Model of Initial Formation of Trust
Detailed Model of Initial Formation of Trust
Boxes containing other boxes are not measurable constructs but are categories of constructs or “second-order”
constructs (see Hunter & Gerbing, 1982). Arrows directed into/out of the trusting beliefs category box represent
individual relationships with each of the four trusting beliefs. Some combination of the four trusting beliefs could
be used in empirical work, or a combination of them could be used as a second-order construct. Also, each could
be used on its own. Processes are pictured in oval shape to distinguish them from constructs.
control process P6
Disposition to trust