TRUST IN MANAGEMENT AND PERFORMANCE: WHO MINDS THE SHOP WHILE THE EMPLOYEES WATCH THE BOSS?

Roger C. Mayer Management Department

Hankamer School of Business PO Box 98006

Baylor University Waco, TX 76798-8006

(254) 710-6203 Roger_Mayer@Baylor.edu

Mark B. Gavin Department of Management

College of Business Administration Oklahoma State University

Stillwater, OK 74078 (405) 744-8614

mgavin@okway.okstate.edu

MANUSCRIPT UNDER REVISION PLEASE DO NOT CITE OR DISTRIBUTE WITHOUT PERMISSION OF AUTHORS

Abstract

Researchers for decades have believed that trust increases performance, but empirical evidence of this has been sparse. This study investigates the relationship between an employee’s trust in the plant manager and in the top management team with the employee’s in-role performance and organizational citizenship behaviors (OCB). Results support a fully mediated model in which trust in both management referents was positively related to focus of attention, which, in turn, was positively related to performance. The results raise questions about appropriate levels of analysis for outcome variables.

2

Trust is mandatory for optimization of a system.... Without trust, each component will protect its

own immediate interests to its own long-term detriment, and to the detriment of the entire system.

- W. Edwards Deming (1994)

Over three decades ago, Argyris (1964) proposed that trust in management is important

for organizational performance. Recognition of the importance of trust in organizational

relationships has grown rapidly in recent years, evidenced by a large number of publications on

the topic addressing both academic and practitioner audiences (e.g., Annison & Wilford, 1998;

Fukuyama, 1995; Mishra, 1996; Shaw, 1997). In spite of this interest, difficulties in defining and

operationalizing trust have hampered the empirical study of its relationship with performance.

Argyris' theorizing and the quote by Deming above are intuitively appealing. However, little

empirical evidence in support of these assertions has been published.

Although some research has investigated how trust affects group performance (e.g.,

Dirks, 1999; Dirks, in press) and the performance of interorganizational relationships (e.g.,

Zaheer, McEvily, & Perrone, 1998), little has been done to establish how trust affects individual

employee job performance. One notable exception is the work of Robinson (1996), who found

that trust in the employer mediated the relationship between psychological contract breach and

self-reported performance. Examination of the process through which trust affected performance

was beyond the scope of that research. Thus, the main purpose of the present study is to consider

the relationship between trust in management and individual performance.

Defining Trust

There have been a number of attempts to define trust in the organizational literature. In a

recent attempt to integrate the important components of various approaches to trust, Mayer,

Davis, and Schoorman (1995) proposed that trust be defined as the willingness to be vulnerable

3

to another party when that party cannot be controlled or monitored. A key component of trust is

its relationship with risk (Deutsch, 1958; Giffin, 1967; Riker, 1974). This definition explicitly

recognizes the relationship between trust and risk, since risk is inherent in vulnerability.

Mayer et al.’s (1995) model explicitly separates trust from both its antecedents and its

outcomes. The theory explains that the perceptions of a trustee’s characteristics that make that

party trustworthy affect the actual level of trust in that trustee. Their theory proposes that

perceived trustworthiness is comprised of three factors: ability, benevolence, and integrity.

Ability is the perception that the trustee has skills and competencies in the domain of interest.

Benevolence is the trustor’s (i.e., trusting party’s) perception that the trustee cares about the

trustor’s well being. Integrity is the perception that the trustee adheres to a set of principles that

the trustor finds acceptable. In related research, Currall (1992) found that semantic differential

measures of benevolence and ability were predictive of trust and of expectations of trustworthy

behavior. Thus, we offer the following hypothesis:

H1: The level of trust in a trustee will be affected by the trustor’s perceptions of the

trustee’s ability, benevolence, and integrity.

Mayer et al. (1995) proposed that the outcome of trust is the trustor’s risk-taking in the

relationship with the trustee. A greater level of trust will lead to more actual risk-taking

behaviors on the part of the trustor. Here, rather than being willing to be vulnerable, the trustor's

behaviors actually allow vulnerability to the trustee.

Using this conceptualization of trust in a three-year study of a restaurant chain, Davis,

Schoorman, Mayer, and Tan (2000) investigated the relationship between employees’ level of

trust in the facility's general manager and several measures of organizational effectiveness. They

found that the restaurants in which the general manager garnered a higher level of trust from the

4

workforce had significantly higher sales and higher net profits. They also found some evidence

that units with more trusted managers had lower rates of employee turnover, a variable important

to long-term profitability.

While the Davis et al. (2000) study suggests that trust in management contributes to

higher levels of organizational performance, it did not delve into the mechanisms by which it

does so. Davis et al. suggested that trust might lead employees to engage in a higher level of

organizational citizenship behaviors (OCB). They suggested that when employees engage in

more helping behaviors, customers' needs are met more effectively. They reasoned that this

would contribute to higher restaurant performance. However, they did not empirically test this

proposed linkage. Thus, the purpose of this study is to examine both theoretically and

empirically the relationship between trust in management and employee performance-related

behaviors.

Sources of Vulnerability

The preceding section described the definition of trust adopted in this paper as a

willingness to be vulnerable to another party. In order to understand the relationship between

trust in management and employee performance, it is important to consider how employees can

be vulnerable to management, and how vulnerability relates to performance. Vulnerability can

derive from a number of sources. It is important to recognize that an employee can become

vulnerable both through active behavior and through passive behavior, or by opting not to engage

in self-protective behavior. For instance, sharing information with a manager that is potentially

damaging to the employee is an example of an active behavior that actually puts the employee at

risk. If the manager uses the information, either intentionally or unintentionally, in a way that

damages the employee’s interests, the outcome for the employee is negative. Mayer et al. (1995)

5

propose that such negative outcomes will lead the employee to reevaluate the manager’s

trustworthiness, and subsequently be less willing to be vulnerable to the manager at a later point

in time.

Sharing sensitive information is one example of an active behavior that puts the employee

at risk. However, not all behavior that puts the employee at risk is active. For example,

monitoring a manager who has a significant impact on an employee’s important outcomes is a

means of reducing the risk associated with the manager’s influence over the outcomes. If the

manager begins to take actions that could damage the employee’s interests, by closely

monitoring, the employee can more quickly take action to lessen the negative effects of the

manager’s influence. Another related action which is more active and which reduces the

employee’s risk is taking actions to "cover one's back." If an employee is unwilling to be

vulnerable to a given member of management (i.e., by definition lacks trust in the manager), the

employee will proactively attempt to gather information and present an image to influential

others in the organization that the employee's actions are justified and that his or her performance

is satisfactory. By doing so, the individual reduces the negative impact the particular manager

can have on the employee's interests. Conversely, trust (for the manager) in this case would lead

to vulnerability from the employee’s choice not to engage in such self-protective actions. Thus,

trust can be manifested in both active behaviors such as sharing sensitive information and in

passive behaviors, or the lack of engaging in self-protective behaviors.

Both the active and passive types of vulnerability described above can be expected to

impact the contribution the employee makes to the functioning of the organization. For example,

when an employee shares sensitive information about work-related difficulties with a manager

which allows the manager to give the employee specific developmental feedback, the result is

6

that the manager can aid the employee’s performance improvement. The quality and the

potential impact of the manager’s feedback are directly related to how specific the information is

which the employee shares. Alternatively, passive vulnerability can be expected to impact

performance. If the employee chooses not to engage in actively monitoring the manager, the

employee has more time available for work performance.

This section has described the relationship between an employee’s willingness to be

vulnerable and performance. In sum, through both active and passive behaviors, employees who

are willing to be vulnerable to management tend to contribute more to the functioning of the

organization. We turn nexty to a more precise consideratio n of the trust-performance linkage.

In the following section, the effect of the employee’s focus of attention in this linkage is

considered.

Focus of Attention

Employees who trust management should be able to focus greater attention toward adding

value to the organization. If an employee is unconcerned about being vulnerable to management,

the employee’s active and passive behaviors will be more likely to promote organizational

performance. As illustrated below, several earlier authors have suggested the importance of

focus of attention on various aspects of performance.

The work of Locke and others (e.g., Locke and Latham, 1990) on goal-setting theory

underscores the importance of focus. Goal-setting theory suggests that goals improve

performance in part because they increase persistence at the task, and help the individual to select

the issues to which attention should be paid. The theory suggests that this improves performance

because the individual is able to focus attention and energy toward the issues that matter most to

the task. The same general logic applies to the relationship between trust and performance. The

7

employee's level of trust in the manager will enable the employee to outperform a less-trusting

employee. While goals tend to direct the focus of attention to performance-relevant activities,

trust facilitates the focus of attention on such activities by helping liberate the employee from

distracting forces.

Similarly, George and Brief (1996) recently pointed out the importance of focus of

attention for an employee’s motivation. They suggested that a person is involved in multiple

roles in his or her life. The person is only performing at his or her full work-related capacity

when his/her attention is focused on the specific work role rather than splitting attention with

other roles, such as being a parent.

Katz, Kochan, and Weber (1985) came to a similar conclusion about focus of attention in

their consideration of industrial relations. Dealing with formal conflict resolution processes such

as grievances and disciplinary actions takes a considerable amount of time and resources. “To the

extent that management and unions devote time and effort to these formal adversarial procedures,

they limit resources available for training, problem solving, communications, and other activities

linked to productivity, human resource management, or organizational development” (p. 511).

They call this a displacement effect, referring to the fact that adversarial procedures displace the

focus of attention from activities that improve the organization’s effectiveness.

It seems reasonable to assume that multiple levels of management ought to affect an

employee’s ability to focus on getting the work done. Managers who are in direct contact with

the employee are important to their focus because their operational and tactical decisions greatly

affect the employee’s daily life. Time and energy spent by an employee either attempting to

monitor the manager’s actions or worrying about their actions distract the employee’s attention

from the work that needs done. In addition to managers in direct contact with the employees,

8

those in top management make strategic decisions that affect the overall culture and success of

the organization. They are also responsible for decisions about what to communicate to

employees concerning the company’s financial position, impending strategic decisions and

moves, and policy changes about such things as monetary raises, changes in benefit packages and

personnel policies. Thus, a lack of trust in top management would likely leave employees

spending time and mental energy speculating about their futures in the company or the future of

the company itself. Conversely, trust in top management should help allow employees to focus

instead on the work that needs done, rather than worrying about such issues as the viability of

their future employment with the company.

This line of reasoning suggests that trust in various levels of management affects an

employee’s ability to focus attention on activities that add value to the organization. Some of

these activities would be reflected in formal performance evaluations, and some would contribute

to the organization in ways that go beyond specified in-role performance expectations. For

example, Podsakoff, MacKenzie, Moorman, and Fetter (1990) provide some evidence that trust

is related to OCB.

Taken as a whole, we believe that the literature reviewed above supports the following

three hypotheses concerning the relationship among trust, focus, and performance:

H2: The level of trust in salient parties within the management hierarchy will be

positively related to employee ability to focus.

H3: Focus will be positively related to individual performance- both in-role and OCB.

H4: Focus will mediate the relationship between trust in salient parties and individual

performance.

Taken together, this study’s hypotheses lead to the theoretical model presented in Figure 1.

9

---------------------------------

Insert Figure 1 about here

---------------------------------

METHOD

Sample and Procedure

This study was conducted in a small nonunion manufacturing firm headquartered

in the Midwestern U.S. that produces tools used by other manufacturers. The organization

consists of eight manufacturing plants and a corporate headquarters, with 333 total employees.

All of the employees in the company were offered the opportunity to participate in the study by

completing a survey.

In three plants that were located in close proximity to corporate headquarters, employees

were given time away from their regular duties to complete the survey in the presence of one of

the researchers. The remaining five production plants were geographically dispersed across the

US to be in proximity to the relevant customer bases. In these plants, the employees were offered

the opportunity to participate in the study via a survey that was mailed directly to their homes.

The cover letter provided a description of the amount of time likely needed to fill out the survey

as well as a description of the feedback procedure. Employees were also given the address and

phone number of the research team in the cover letter and were encouraged to make contact if

they had concerns about confidentiality or the process. A stamped, pre-addressed envelope was

also included for returning the survey directly to the research team. Two weeks after the

employees received the survey in the mail, a reminder notice was sent to those employees who

had not returned their surveys urging them to do so. 288 surveys were returned, for an overall

10

company response rate of 86.2%. 100 of the 138 employees in the remote plants returned

surveys for a response rate of 72.5%. The mean age of respondents was 37.3 years, and their

average organizational tenure was 7.4 years.

To reduce concerns with single-source method bias, the outcome measures for this study

were collected from the employees’ supervisors. Three months after the employee survey was

conducted, supervisors received a brief survey for each employee as described below. Inclusion

of a respondent in the final data set thus required that an employee returned his/her completed

survey and that the employee’s supervisor provided ratings on the outcome variables for the

employee. Of the 288 initial respondents, matched data were available for 247 employees who

subsequently comprised the final sample (85.1% of initial respondents).

Measures

Based on the model of organizational trust described earlier in this paper, we measured

trust and the three factors of trustworthiness. Each employee responded twice to versions of

these four scales: once in reference to trust in the employee’s plant manager (PM), and once in

reference to trust in the top management team (TMT). The items were altered slightly to reflect

the two trust referents.

The three scales measuring the factors of trustworthiness from Mayer and Davis (1999)

were utilized in this study. Thus, six-item scales measured the perceived ability and integrity of

the trustee. The perception of the trustee’s benevolence was measured with a five-item scale.

Cronbach’s alphas for these three scales were .91, .89, and .92 respectively for the PM and .89,

.85, and .87 for the TMT.

The four-item measure of trust utilized by Schoorman, Mayer, and Davis (1995) and

Mayer and Davis (1999) was used as a foundation for the current measure of trust. In the

11

Schoorman et al. study, Cronbach’s alpha for the measure was .82. Using the same measure with

stems modified to reflect the TMT, Mayer and Davis found Cronbach’s alphas of .59 and .60 in

different waves of data collection. Despite these alphas, they found the measure to have good

test-retest reliabilities of .75 in a five-month time lag and .66 over nine months. These authors

encouraged further developmental work with the measure, pointing to the possibility that the

measure may include multiple dimensions.

In response to concerns with the level of alpha and the possibility of dimensionality of the

construct, six additional items were generated to measure trust. Each of the items reflected

statements from participants in employee focus groups that related to various ways of being

vulnerable to a given managerial referent. The final trust measure thus consisted of ten items,

each of which queried the respondent’s willingness to be vulnerable to the particular trustee. An

example item is “I would be willing to let my supervisor have complete control over my future in

this company.” The trust and trustworthiness scales used 5-point Likert-type items with

agree/disagree anchors for each scale point. Higher scores indicate greater amounts of the

construct.

Six items that comprised the measure of focus were generated based on comments made

in the employee focus groups. The items measure the extent to which the respondent feels able

to focus attention on the work that needs to be done. Examples of the items are “The work

climate here allows me to focus on doing my job” and “In this company, you need to make sure

you ‘cover your backside’” (reverse scored). All of the trust and focus items were in five-point

Likert-type format, with scale anchors from “strongly disagree” to “strongly agree”

accompanying each scale point

12

As mentioned previously, data were collected from the supervisors to measure the

performance-related outcomes. Supervisors rated employees on three performance dimensions.

First, employees were rated on a seven-item measure of in-role performance taken from Williams

and Anderson (1991). One item from the in-role performance measure (i.e., “Engages in

activities that will directly affect his/her performance evaluation”) was dropped because it

conceptually overlapped with focus, thereby resulting in a six-item measure of in-role

performance. Also, Williams and Anderson’s (1991) scales were utilized to measure OCB

directed toward specific individuals such as coworkers and supervisors (OCBI; 7 items) and

OCB directed toward the organization as a whole (OCBO; 6 items). Each of the in-role, OCBI,

and OCBO items used response categories ranging from “never” (1) to “very often” (5), with

higher scores representing more frequent displays of the performance behaviors. Cronbach’s

alpha was .91, .82, and .67 for the in-role performance, OCBI, and OCBO scales, respectively.

RESULTS

Means, standard deviations, Cronbach alphas, and intercorrelations of the variables can

be found in Table 1. All of the significant correlations were in the expected directions.

Consistent with the hypotheses, the trustworthiness factors were all significantly correlated with

trust in the corresponding referent. Using the five-item measures of trust described below, the

correlations between trust and the three antecedents ranged from .72 to .76 for the PM and .62 to

.71 for the TMT. These results support Hypothesis 1. Also, trust in each referent, specifically

the PM and the TMT, was significantly correlated with focus (r=.35 and .24 respectively). These

results support Hypothesis 2. Lastly, consistent with Hypothesis 3, focus was significantly

correlated with both the employee’s OCBI and OCBO (r=.15 and .19 respectively), although not

with in-role performance.

13

---------------------------------

Insert Table 1 about here

---------------------------------

Scale Development

Exploratory factor analysis of the six focus items using principal factors extraction

yielded a single-factor solution. All factor loadings exceeded .55 and the single factor accounted

for 47% of the variance. Cronbach’s alpha for the focus scale was .77.

An exploratory factor analysis was conducted on the items relating to trust in the TMT.

The factor analysis used an oblique rotation and extracted two factors. Five of the items loaded

on each factor. Examination of the pattern of loadings revealed that the first factor was

comprised of the four original items from Schoorman et al. (1996) and Mayer and Davis (1999)

plus one additional item. These items tended to reflect a “generic” willingness to take risk in that

they described a willingness to be vulnerable without describing specific behavi (e.g., I would be

willing to let my PM have complete control over my future in this company). The second factor

tended to include items that proposed a willingness to engage in more specific risk-taking

behaviors, often related to communication with the trust referent (e.g, I would tell my PM about

mistakes I’ve made on the job, even if they could damage my reputation). The generic trust

items demonstrated a pattern of consistently higher correlations with other variables in the study.

This procedure was repeated on the PM data with very similar results (the final item had small

loadings on both factors, and was assigned to the “specific” factor based on both consistency

with the TMT analyses and the specificity of its behavioral referent). Based on this external

validity evidence with the other variables in the nomological net, we decided to initially use the

five-item measure for the analyses. Cronbach’s alphas for the ten item scales were .82 and .76

14

for PM trust and TMT trust respectively. Alphas for the five-item generic trust scales were .81

and .72, respectively.

Model Testing

The hypothesized model shown in Figure 1 was tested using structural equation modeling

with the LISREL 8.30 software package (Joreskog & Sorbom, 1993). Each of the constructs was

modeled as a latent variable with a single indicator. The single indicators were formed as scale

scores and reliability information was used to fix the measurement model. Thus, the specified

model consisted of twelve latent variables and twelve indicators. Also, while not depicted in

Figure 1, the six exogenous latent variables (i.e., the trustworthiness factors for the PM and

TMT) were allowed to intercorrelate.

Overall model fit was assessed using the Comparative Fit Index (CFI: Bentler, 1990) and

the root mean square error of approximation (RMSEA: Browne & Cudek, 1993; Steiger & Lind,

1980). The CFI ranges from 0 to 1.0, with values of .90 and higher indicating good overall fit.

For the RMSEA, values of .08 and lower are viewed as indicating acceptable fit. With regard to

conventional criteria, the overall fit of the hypothesized model was below acceptable levels

although the CFI was reasonably close (χ2 = 280.39 with 40 df, p < .01; CFI = .86; RMSEA =

.16). However, it is worth noting the nearly all of the paths were significant and in the

hypothesized direction. Specifically, consistent with Hypothesis 1, the paths relating perceived

ability and integrity of both the PM and of the TMT to trust in the respective referents were

positive and significant. Also, the paths relating trust in both the PM and in the TMT to focus

were both positive and significant. This provides further support for Hypothesis 2. The paths

from focus to each of in-role performance, OCBI and OCBO were also significant, as predicted

in Hypothesis 3. In fact, the only hypothesized paths that were not significant in this model were

15

the two relating perceived benevolence of the PM and of the TMT to trust in the respective

referent. The fact that the perceived benevolence of both the PM and the TMT were unrelated to

trust in the respective referent is very likely due to multicollinearity among the trustworthiness

factors, which ranged from .65 to .80 for the PM and .55 to .78 for the TMT.

In assessing potential sources of misfit for the above model, we first examined the zero-

order correlations to see if there were any potential relationships for which we were not

accounting in our specified model. One concern arose with the interrelationships among the

performance outcomes, which have substantial zero-order correlations ranging from .42 to .62.

The magnitudes of these correlations suggest that not accounting for the relationship among the

performance outcomes might be causing misfit in the model. Subsequent examination of the

modification indices for the hypothesized model confirmed this, with large modification indices

generated for all interrelationships among the performance outcomes.

We took this evidence to suggest that performance might be better conceptualized in this

study as a single construct, with each of the three components representing different facets of

overall performance. Several arguments support this approach. First, as mentioned above, the

zero-order correlations and modification indices suggest that, at least as far as supervisors in this

sample are concerned, the different performance components are not entirely separable. That is,

it appears that supervisors might have had some difficulty in differentiating among the three

performance constructs. This line of reasoning is consistent with the findings of Morrison

(1994), who found that the line between in-role and extra-role performance (i.e., OCBs) is not

necessarily clear. Supervisors in this study may well have experienced such ambiguity, since

they are ultimately interested only in the performance necessary for their units and for the

company—and that performance can take many forms in different situations. Second, our

16

primary goal in this study was linking trust to performance and investigating the role of focus in

the process. By viewing the different performance components simply as indicators of a more

global performance construct, we could still accomplish this goal without altering the conceptual

underpinnings of the original model. Thus, based upon the above empirical evidence and

conceptual arguments, we chose to collapse the three performance outcomes into a single

performance construct for subsequent model testing.

The respecified hypothesized model is shown in Figure 2. It is identical to the original

model except that the three performance latent variables have been replaced by a single

performance latent variable, with each of the three performance components (in-role

performance, OCBI, and OCBO) serving as an indicator. The rest of the measurement model

remains identical to the original model. Thus, the new model consists of ten latent variables and

twelve indicators. The overall fit of this model is substantially better than the original, achieving

acceptable fit with regard to conventional criteria (χ2 = 105.76 with 39 df, p < .01; CFI = .96;

RMSEA = .08). Having achieved acceptable fit, we next compared this model to one other

before examining individual parameter estimates. Specifically, since the hypothesized model

posits full mediation of the effects of trust on performance through focus, we specified a partially

mediated model that adds the two direct effects of trust in the PM and in the TMT on

performance. As expected, this model also achieved an acceptable fit (χ2 = 103.12 with 37 df, p

< .01; CFI = .96; RMSEA = .08). However, the difference between the two models was not

significant (χ2 difference = 2.64 with 2 df). These results favor the more parsimonious fully

mediated model as originally specified. Additionally, neither of the added paths for the effects of

trust in the PM and in the TMT on overall performance was significant. Subsequently, the fully

mediated model was retained, providing omnibus support for Hypothesis 4.

17

---------------------------------

Insert Figure 2 about here

---------------------------------

The estimates for the individual parameters of the fully mediated model can be found in

Figure 2. The parameter estimates are supportive of nearly all of the hypotheses. Specifically,

the paths relating perceived ability and integrity of the PM and of the TMT to trust in the

respective referent were significant and positive (i.e., Hypothesis 1). Also, the two paths relating

trust in the PM and in the TMT to focus were significant and positive (i.e., Hypothesis 2), as was

the path from focus to overall performance (i.e., Hypothesis 3). Along with the non-significant

chi-square difference test between the fully and partially mediated models, the parameter

estimates supporting Hypotheses 2 and 3 provide support for Hypothesis 4, as does the fact that

the indirect effects of trust in both referents on performance were significant in the LISREL

output. As was the case in the original model, the only paths that were not significant are the two

relating perceived benevolence of the PM and the TMT to trust in the respective referent. Again,

this is likely due to a multicollinearity problem among the trustworthiness factors.

For reasons discussed previously, the above models were all run using the five-item

measure of generic trust. As part of the examination of the full ten-item measure, we replicated

the complete model testing sequence described above using the ten-item measures of trust in the

PM and in the TMT. While the specific values for the fit indices and parameter estimates were

slightly different, the results were identical in terms of model comparisons and patterns of

significance for the individual parameter estimates.1

18

DISCUSSION

While trust has long been assumed to relate to performance in organizations, the means

through which it has its effect have been less clear. This study provides empirical support for our

assertion that trust in management allows employees to focus on the tasks which need done to

add value to the organization. The modeling results in this study indicate that the relationship

between trust in the two levels of management and a multifaceted treatment of performance was

mediated by employees’ perceived ability to focus attention on value-adding activities.

One of the strengths of this study was that the data were collected from both the

employees and their supervisors, thereby eliminating any concern with common method

variance. The longitudinal nature of this study was also a methodological strength. The outcome

data were collected three months after the employee data, which allowed their attitudes toward

management and their work environment to be reflected in the measures of their performance.

Despite its strengths, no study is without limitations. The setting for this study was a

small, nonunion company. Employees may make attributions of trust and trustworthiness

somewhat differently than in a large corporation. The focal company in this study lacked

objective measures of individual performance, which would have been valuable for the present

purposes.

While trust in management was not directly correlated with in-role performance, this

study offers evidence that trust does affect performance. Trust in specific managerial referents

appears to contribute to a more generalized construct in focus of attention. A number of factors

are likely to affect an employee’s ability to focus attention on work that needs to be done, as

evidenced by the work on goal setting theory and other more recent work on motivation

discussed earlier in this paper. On a more macro level, organizational mission statements and

19

other such documents are intended to motivate employees by getting them to see the “big

picture” and to contribute to the organization in ways which may or may not be deemed part of

their role. The results of this study suggest that a lack of trust in management may undermine

management’s attempts to direct employees’ attention at accomplishing outcomes important to

the organization, particularly when the necessary behaviors fall outside the employees’ specified

roles. Job myopia seems to be a greater threat among employees who do not trust management.

The modeling results from the current study provide evidence that focus of attention may

play an important role in a process through which trust in management influences performance.

The current results suggest that increments in performance may be at both the individual and

organizational levels. The current study provides evidence that employees who are more willing

to be vulnerable to management are more able to focus their attention on activities that add value

for the organization, and that this focus of attention in turn contributes to both their in-role and

extra-role contributions to the organization. Morrison’s (1994) work suggests that the

demarcation between in-role and extra-role performance may not be as clear as is commonly

assumed. It is possible that any ambiguity between the two may have attenuated the trust to in-

role performance relationship found in the current study. It is important to bear in mind,

however, that both in-role performance and OCB affect the performance of the business unit and

the organization. Whether contributions are attributed to in-role performance or to OCB by the

supervisors, trust in management appears to have positively affected employee behaviors that

contribute to organizational performance.

The current study provides evidence that trust in management allows employees to focus

more attention on value-producing activities, which contributes to performance. Since both in-

role performance and OCB affect business unit and organizational performance, the results of the

20

current study raise the possibility that in order to reliably detect a relationship between trust in

management and performance, there is a need to also assess performance at a higher level of

analysis such as a department or organization level. It is possible that the dearth of published

evidence that trust in management affects individual performance might be due in part to

problems with measuring in-role performance, and also to the fact that trust in management

apparently allows employees to focus on the work that needs done irrespective of whether that

work falls within the strict definition of the employee’s role. Such contributions increase

organizational performance, irrespective of whether or not measures of individual in-role

performance are sensitive enough or inclusive enough to detect them.

In this study, trust in the PM had a greater effect on focus than did trust in the TMT. This

is apparent in both the zero-order correlations and in the modeling parameter estimates in Figure

2. This makes sense given that an employee has much greater exposure to the PM, particularly in

this research site wherein five of the eight production plants were geographically dispersed

around the U.S. This dispersion further reduced contact between employees and the TMT.

Perhaps in organizations where facilities are less dispersed and employees have greater exposure

to the TMT, the effects of trust in the TMT may be more important to focus and subsequently to

performance.

In the current study, a longer measure of trust was developed that is consistent with prior

conceptual and empirical work that views trust as a willingness to be vulnerable. The current

results appear promising in that this might provide a more reliable means of measuring a given

party’s willingness to be vulnerable to a specific other. Preliminary evidence suggests that the

generic and specific types of vulnerabilities merit further attention as dimensions of trust.

21

A second methodological contribution of this study is the use of the videotape in remote

data collections. In this study, the response rate for employees surveyed by mail was 72.5%.

This is considerably higher than the rate normally expected. Discussions with PMs suggested

that the videotape procedure used in this study increased the response rate. It appears that the use

of this procedure may have utility for field researchers who attempt to collect data from research

sites where they lack physical access. Further empirical research that directly tests the utility of

this procedure on mail survey response rates is clearly warranted.

Managerial Implications

The relationship between employee trust in management and performance is to date not

well understood. However, managers commonly understand the need to keep employees focused

on pursuing the organization’s goals. This study provides evidence that as various levels of

management garner greater trust, employees’ ability to focus attention on the work that the

organization needs done is enhanced. Thus, the results of this study can help make clear to

managers that it is in their own best interest to build the level of trust they garner from their

employees. Furthermore, it makes clear that trust in multiple levels of management referents is

important to employees’ ability to focus attention on the work. These results can be used to

make salient to managers at all levels why it is important to behave in a trustworthy fashion.

Future Research

The results of this study indicate several potentially fruitful areas for future research.

Focus of attention should be examined in other settings to further clarify its role in the

relationship between trust and performance. Further work on the measurement of trust and of

focus would help to improve our understanding of these constructs and their relationships with

other variables. Further consideration of focus and how it mediates the role between trust and

22

performance would also be of value. Undoubtedly there are a number of additional factors that

influence employees’ ability to focus attention on the work that needs to be done. A better

understanding of these factors and their interrelationships with one another may be a fruitful

means of improving employee productivity and organizational performance. The dimensions of

the measure of trust found here provide a basis for further investigation of trust’s dimensionality.

As mentioned above, future research should investigate the relationship between trust and

performance by looking at both individual (i.e., in-role and OCB) and higher level (i.e.,

departmental and organizational) measures of performance. In the current study, it was

suggested that a lack of trust in management might contribute to job myopia, or an unwillingness

to engage in activities outside those which are formally defined and rewarded within the

employee’s job. This possibility merits direct investigation.

This study examined the relationship between employee trust in management and

performance. Whitney (1994) suggested that management’s trust in employees is also important

to performance. This is a related topic for future research that appears to merit attention.

23

REFERENCES Argyris, C. 1964. Integrating the individual and the organization. New York: John Wiley &

Sons, Inc. Annison, M. H., & Wilford, D. S. 1998. Trust matters: New directions in health care leadership.

San Francisco: Jossey-Bass Publishers. Bentler, P. M. 1990. Comparative fit indexes in structural models. Psychological Bulletin, 107:

238-246. Browne, M. W., & Cudek, R. 1993. Alternative ways of assessing model fit. In K. A. Bollen &

J. S. Long (Eds.), Testing structural equation models. Newbury Park, CA: Sage. Currall, S. C. 1992. Group representatives in educational institutions: An empirical study of

superintendents and teacher union presidents. Journal of Applied Behavioral Science, 28: 296-317.

Davis, J. H., Schoorman, F. D., Mayer, R. C., & Tan, H. H. 2000. The trusted general manager and business unit performance: Empirical evidence of a competitive advantage. Strategic Management Journal, 21: 563-576.

Deming, E. W. 1994. Foreword. In J. O. Whitney, The trust factor: Liberating profits and restoring corporate vitality. New York: McGraw-Hill.

Deutsch, M. 1958. Trust and suspicion. Journal of Conflict Resolution, 2: 265-279. Dirks, K. T. 1999. The effects of interpersonal trust on work group performance. Journal of

Applied Psychology, 84, 445-455. Dirks, K. T. In press. Trust in leadership and team performance: Evidence from NCAA

Basketball. Journal of Applied Psychology. Fukuyama, F. 1995. Trust: The social virtues & the creation of prosperity. New York: Free

Press. George, J. M., & Brief, A. P. 1996. Motivational agendas in the workplace: The effects of feelings

on focus of attention and work motivation. In B. M. Staw & L. L. Cumings (Eds.), Research in organizational behavior, vol. 18. Greenwich, CT: JAI Press, Inc.

Giffin, K. 1967. The contribution of studies of source credibility to a theory of interpersonal trust in the communication department. Psychological Bulletin, 68: 104-120.

Joreskog, K. G., & Sorbom, D. 1993. LISREL 8: User’s reference guide. Chicago: Scientific Software.

Katz, H. C., Kochan, T. A., & Weber, M. R. 1985. Assessing the effects of industrial relations systems and efforts to improve the quality of working life on organizational effectiveness. Academy of Management Journal, 28: 509-526.

Locke, E. A., & Latham, G. P. 1990. A theory of goal setting & task performance. Englewood Cliffs: Prentice Hall.

Mayer, R. C., & Davis, J. H. 1999. The effect of the performance appraisal system on trust for management: A field quasi-experiment. Journal of Applied Psychology, 84, 123-136.

Mayer, R. C., Davis, J. H., & Schoorman, F. D. 1995. An integrative model of organizational trust. Academy of Management Review, 20: 709-734.

Mishra, A. K. 1996. Organizational responses to crisis: The centrality of trust. In Trust in Organizations: Frontiers of theory and research, R. M. Kramer & T. R. Tyler (Eds.), Thousand Oaks, CA: Sage.

Morrison, E. W. 1994. Role definitions and organizational citizenship behaviors: The importance of the employee’s perspective. Academy of Management Journal, 37, 1543-1567.

24

Podsakoff, P. M., MacKenzie, S. B., Moorman, R. H., & Fetter, R. 1990. Transformational leader behaviors and their effects on followers’ trust in leader, satisfaction, and organizational citizenship behaviors. Leadership Quarterly, 1: 107-142.

Robinson, S. L. 1996. Trust and breach of the psychological contract. Administrative Science Quarterly, 41:574-599.

Riker, W. H. 1974. The nature of trust. In J. T. Tedeschi (Ed.) Perspectives on social power. Chicago: Aldine.

Schoorman, F. D., Mayer, R. C., & Davis, J. H. 1996. Empowerment in Veterinary Clinics: The role of trust in delegation. Paper presented at Annual Meeting of Society for Industrial and Organizational Psychology, San Diego.

Shaw, R. B. 1997. Trust in the balance: Building successful organizations on results, integrity, and concern. San Francisco: Jossey-Bass Publishers.

Steiger, J. H., & Lind, J. C. 1980. Statistically-based tests for the number of common factors. Handout for a presentation delivered at the meeting of the Psychometric Society, Iowa City, IA.

Whitney, J. O. 1994. The trust factor: Liberating profits and restoring corporate vitality. New York: McGraw-Hill.

Williams, L. J., & Anderson, S. E. 1991. Job satisfaction and organizational commitment as predictors of organizational citizenship and in-role behaviors. Journal of Management, 17, 601-617.

Zaheer, A., McEvily, B., & Perrone, V. 1998. Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance. Organization Science, 9, 141-159.

ENDNOTE

1 The results from these models are available from the authors upon request.

TA

BLE

1

Des

crip

tive

stat

istic

s an

d in

terc

orre

latio

ns

M

ean

SD

1

2 3

4 5

6 7

8 9

10

11

12

13

14

1 A

bilit

yP

M

3.45

0.

82

.91

2 B

enev

olen

ce PM

3.

14

0.93

.6

5**

.92

3 In

tegr

ityP

M

3.23

0.

85

.73*

* .8

0**

.89

4 A

bilit

yT

MT

2.94

0.

70

.27*

* .1

5*

.14*

.8

9

5 B

enev

olen

ce TM

T

2.59

0.

75

.23*

* .3

5**

.23*

* .5

5**

.87

6 In

tegr

ityT

MT

2.79

0.

63

.20*

* .2

7**

.25*

* .6

6**

.78*

* .8

5

7 T

rust

5 PM

3.

21

0.77

.7

4**

.72*

* .7

6**

.13*

.2

0**

.19*

* .8

1

8 T

rust

10P

M

3.51

0.

59

.70*

* .6

9**

.73*

* .1

1 .1

9**

.16*

.9

0**

.82

9 T

rust

5 TM

T

2.72

0.

63

.22*

* .1

7**

.15*

.6

2**

.66*

* .7

1**

.26*

* .1

8**

.72

10 T

rust

10T

MT

3.08

0.

53

.14*

.1

1 .0

7 .5

3**

.60*

* .6

2**

.19*

* .2

6**

.80*

* .7

6

11 F

ocus

2.

66

0.88

.2

9**

.42*

* .4

0**

.23*

* .3

1**

.32*

* .3

5**

.33*

* .2

4**

.22*

* .7

7

12 In

-rol

e 4.

05

0.61

.0

3 .1

0 .1

1 -.

08

-.05

.0

2 .0

3 .0

2 -.

05

-.03

.0

9 .9

1

13 O

CB

I 2.

94

0.64

.1

2 .2

0**

.16*

-.

10

.05

.01

.19*

* .2

0**

-.04

.0

3 .1

5*

.50*

* .8

2

14 O

CB

O 3.

82

0.62

.1

9**

.22*

* .2

3**

.02

.10

.17*

* .1

6*

.16*

.1

2*

.10

.19*

* .6

2**

.42*

* .6

7

* p

< .0

5; *

* p

< .0

1; P

M =

pla

nt m

anag

er; T

MT

= to

p m

anagem

ent t

eam

; n =

247

; coe

ffici

ent a

lpha

s ar

e on

the

diag

onal

.

FIG

UR

E 1

Hyp

othe

size

d m

odel

PM

= p

lant

man

ager

, TM

T =

top

man

agem

ent t

eam

; cor

rela

tions

am

ong

the

exog

enou

s va

riabl

es o

mitt

ed fo

r cl

arity

A

bilit

y P

M

B

enev

olen

ce PM

In

tegr

ity PM

A

bilit

y T

MT

B

enev

olen

ce TM

T

In

tegr

ity TM

T

T

rust

PM

Tru

st TM

T

Foc

us

OC

BI

In-r

ole

Per

form

ance

OC

BO

27

F

IGU

RE

2

R

esul

ts fr

om r

espe

cifie

d hy

poth

esiz

ed m

odel

with

a s

ingl

e ov

eral

l per

form

ance

con

stru

ct

*p <

.05;

PM

= p

lant

man

ager

, TM

T =

top

man

agem

ent t

eam

; par

amet

er e

stim

ates

from

the

stan

dard

ized

sol

utio

n; c

orre

latio

ns a

mon

g

the

exog

enou

s va

riabl

es o

mitt

ed fo

r cl

arity

A

bilit

y P

M

B

enev

olen

ce PM

In

tegr

ity PM

A

bilit

y T

MT

B

enev

olen

ce TM

T

In

tegr

ity TM

T

T

rust

PM

Tru

st TM

T

Foc

us

Per

form

ance

.36*

.18

.19

.44*

.25*

.56*

.41*

.26*

.21*