Performance Management and Trust in Supervisors

Abstract

This article investigates the effect of performance management and explores one factor—trust in one’s supervisor—as a critical facilitator of it. Although improving performance is a prominent issue in public management and the term, performance management, is in heavy use in the public sector, systematic investigations on the subject are relatively rare. Defining performance management as a human resource management tool, this research empirically tests the associations between performance management and two outcomes: perceived work-unit performance and perceived agency performance. In addition, it examines the role of trust as a facilitator of the successful implementation of performance management. Using the Merit Principles Survey 2005, we test the ideas. Ordered logit regression analyses confirm that performance management promotes perceived performance of both work-unit and agency, and the leverages are further increased under a high level of trust in supervisors.

Keywords

performance and performance appraisal organizational behavior/development employee attitudes behavior and motivation federal government HRM new public management

Introduction

Improving performance has been a prominent topic of public management for decades, and its importance has been further emphasized with the emergence of New Public Management (NPM) as a dominant paradigm for public sector reform (Kelman, 2007; Kettl, 2000; Pollitt & Bouckaert, 2000). Radin (2000, p. 168) even singles out the demand for performance as the theme that best characterizes the public sector in the late 20th century. As a “standard component” of NPM reform packages (Peters, 2008, p. 195), performance management has obtained the attention of reformers throughout the world (Talbot, 2005). Many modern democracies are currently engaged in various forms of performance management (Lah & Perry, 2008; Organisation for Economic Co-operation and Development [OECD], 2005; Perrin, 2006), and the United States is no exception (Moynihan, 2008; Radin, 2006).

Underlying various performance management arrangements is the fundamental assumption that they will contribute to better performance. Performance budgeting, for example, improves performance by enhancing the efficiency of resource allocation, whereas performance-based human resource management does so by motivating people to work harder toward the achievement of organizational goals. This very assumption has strongly influenced the adoption and implementation of performance management by reformers at all levels of government (Behn, 2002; Forsythe, 2001), yet its validity has rarely been subject to rigorous evaluation (Sanger, 2008). Is performance management an effective performance enhancement tool? What are the external and internal factors that affect the implementation of performance management? Given the enormous time and resources devoted to performance management at all levels of government, and the continuing interest in and commitment to this managerial initiative, these questions warrant serious consideration. With few exceptions (e.g., Lee, Cho, & Kim, 2009; Yang & Hsieh, 2007), however, scant research addresses these critical questions by analyzing data from large, representative samples. There is a growing body of studies that examine the effectiveness of performance management and the conditions for its successful implementation (e.g., Heinrich, 2007; Moynihan, 2008; Perry, Engbers, & Jun, 2009; Radin, 2006), but these are largely qualitative or descriptive in nature, which makes it difficult to test the relative validity of competing perspectives on performance management.

In an attempt to fill the gap between normative discussion and empirical evidence, this research tests the effects of performance management and explores one condition required for its successful implementation. Of the many types of performance management, this study focuses on performance management as a tool for human resource management within organizations. Two outcomes of performance management are investigated: perceived work-unit performance and perceived agency performance. Achieving a high level of performance is an ultimate outcome that performance management pursues. We consider the factor in two different levels: work-unit and agency.¹ Using perceived measures of performance is one obvious limitation of this research. However, perceived measures are often employed in public management research and some studies demonstrate that employee perception is substantially correlated with an objective outcome (Boyd, Dess, & Raheed, 1993; Yang & Pandey, 2009).

The research explores the role of trust for the successful implementation of performance management. Trust increases organizational performance by working as a lubricant of various functions of organizations (Dirks & Ferrin, 2001; Kramer, 1999). Accordingly, scholars call for more research for the indirect effect of trust as well as its direct effect (Dirks & Ferrin, 2001). To this end, we test whether trust in supervisors, who evaluate employee performance, conditions the effective implementation of performance management.

This research therefore seeks to answer two questions:

Research Question 1: Are employees’ evaluations of performance management in their organizations positively associated with better perceptions of work-unit and agency performance?

Research Question 2: Does trust in supervisors strengthen the relationships between performance management and perceived performances of work-unit and agency?

Using the data of the 2005 Merit Principles Survey, we test these ideas by ordered logit regression analyses.

Performance Management: Current Trend, Definition, and Core Processes

In this section, we discuss several facets of performance management. We first explain several performance initiatives in the United States under the NPM movement. Then, the research defines the term performance management to be suitable for the current research. The current usage of the term is quite broad, and scholars often indicate different things by the term. Narrowing down the definition might be useful for more accurate discussion. We finally review four core processes of performance management and several motivation theories supporting their importance.

Performance Management Movement in the United States

Over the last several decades, governments at all levels in the United States have instituted a variety of performance management arrangements, such as performance budgeting, performance contracting, and performance-based human resource management (Abramson, Breul, & Kamensky, 2006). The Government Performance and Results Act of 1993 (GPRA), for example, mandates federal agency involvement in performance management activities: It requires federal agencies to develop annual and long-term plans for what they intend to achieve, measure how well they are doing, generate information for performance budgeting, and communicate performance information to Congress and the general public (Radin, 2006). The Program Assessment Rating Tool (PART), a federal program management tool developed by the George W. Bush administration, also demands federal agency participation in similar performance management activities at the program level (Moynihan, 2008). The Bush administration also sought to make federal personnel practices more flexible and performance-oriented by two major personnel reforms: the enactment of the Homeland Security Act of 2002 and the creation of the National Security Personnel System for civilian employees in the Department of Defense (Brook & King, 2008).

One can find similar initiatives at the lower levels of government. As Hays (2004) identifies performance management as one of the most significant trends in state and local human resource management in recent years, many state and local governments have implemented diverse performance management initiatives, including strategic planning and management, managing for results, and performance budgeting (Melkers & Willoughby, 1998, 2004; Poister & Streib, 1999, 2005).

Definition of Performance Management

Although diverse initiatives of performance management are currently employed, defining performance management, at least in the public management field, is not as straightforward as it may seem. Performance management has been used as an all-encompassing term to describe any management process that involves collecting, interpreting, and/or utilizing performance information for performance improvement—whether for budgeting, program management, performance contracting, or personnel management (Behn, 2002; Poister, 2003; Pollitt, 2001). This murky definition of performance management is not surprising, given the scope and breadth of performance management practices.

Behn (2002, p. 20), however, notes that the various definitions used by practitioners and scholars basically take two forms: referring either to “efforts to improve the performance of individual employees through personal performance plans, performance appraisals, and the usual collection of carrots and sticks” or to “the collection of organizational, managerial, and leadership strategies that are designed to get the people within a public agency—and their essential collaborators—to achieve specific public purposes.” Although large-scale performance management systems indicate that these two notions of performance management can and should be integrated, they are distinguished. The former, reflecting the perspective of human resource professionals (Behn, 2002), focuses primarily on the processes of motivating individuals to perform at higher levels. By enhancing individual performance—and thereby enhancing individual accountability as well—within an organization, performance management is expected to contribute to organizational performance. The enactment of the Civil Service Reform Act (CSRA) of 1978 is a good example of this first type of performance management. The CSRA, one of the earlier attempts at tightening the links between pay, firing, and performance, led to the development of more modern performance appraisal and performance pay in federal agencies (Haga, Richman, & Leavitt, 2010). On the other hand, employing a macro perspective, the latter concerns the achievement of performance targets and organizational mission by setting and implementing strategies to ensure support and commitment from internal and external stakeholders (Behn, 2002). The GPRA of 1993 is an example of this second type of performance management.

This study examines the impact of performance management as a human resource management tool by employing the former perspective. With this micro approach, it defines performance management by using human resource terms: Performance management is a process of managing employee performance by planning, monitoring, evaluating, and rewarding individual contributions.

Core Processes of Performance Management

Performance management as a human resource management process involves a sequence of four core steps: “identifying and setting clear and measurable performance goals; measuring performance to monitor progress toward the achievement of goals; providing feedback on performance results; and utilizing performance appraisals to inform major personnel management decisions such as rewards and accountability” (Schermerhorn, Hunt, & Osborn, 2008, p. 155).

As an integrated and continuous process, effective performance management depends on proper implementation of each of these steps. Concepts and theories of work motivation help to explain how these steps can work to motivate people to perform at higher levels. Goal-setting theory, first of all, stresses the importance of clear goals to individual motivation and performance (Locke & Latham, 2002). Clear and specific goals are more motivational than vague, general goals since they provide directions and guidance, serve as criteria for objective assessment of individual contributions, and give a sense of which tasks should take precedence. Employee participation is also emphasized by goal-setting theory in that the sense of ownership created through this process helps to build employee acceptance and commitment to goals (Locke & Latham, 2002). In addition, task feedback through performance appraisal helps employees understand where they stand in relation to achieving their goals and where they can improve.

Other theories of motivation also support the idea of using performance management as a motivational tool. Expectancy theory stresses the importance of establishing clear links between performance and rewards: the idea being that good performance will bring rewards—either extrinsic or intrinsic—as outlined in an agreed-upon plan that will motivate employees to perform at a higher level (Vroom, 1964). Equity theory highlights the centrality of fairness in the design and implementation of rewards and punishments; employees’ work efforts are affected by their perceptions of how others are rewarded (Adams, 1965). The effectiveness of any performance management scheme thus depends on proper management of employee perceptions of equity.

These theoretical underpinnings imply that performance management is indeed a management-intensive process and that successful implementation calls for careful managerial attention to each of the steps in the process. In other words, performance management cannot be successful unless all four steps are well managed and integrated as a whole system. Such a full-scale performance management requires managers at all levels to develop managerial skills and competencies relevant to each step (e.g., Bilgin, 2007; Grote, 2000; Latham, Almost, Mann, & Moore, 2005; Risher & Fay, 2007). In addition to these managerial skills and competencies, an intangible managerial resource is required of everyone in the organization—trust. The next section discusses this managerial resource.

Trust: A Facilitator of Performance Management

Definition and Expected Effects

Although trust has been defined in various ways, Rousseau, Sitkin, Burt, and Camerer (1998) find that most scholars’ definitions share a common element: trust as a psychological state. Accordingly, Rousseau et al. (1998, p. 395) define trust as “a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another.”

Interpersonal trust within organizations has been dealt as a critical factor in maintaining a highly motivated workforce. Scholars from various fields, including leadership theory (e.g., Dirks, 2000; Fairholm, 1994; Zand, 1997), transaction cost theory (e.g., Williamson, 1993; Zaheer, McEvily, & Perrone, 1997), and rational choice theory (e.g., Ostrom, 1998), insist that trust is a valuable managerial resource. In summary, Kramer (1999) provides three ways in which trust conveys benefits to organizations: by decreasing transaction costs, by increasing spontaneous sociability, and by facilitating voluntary deference. Interpersonal trust reduces the necessity of monitoring and control, which leads to decreased transaction costs (Williamson, 1993). It increases altruistic behavior within organizations. Organizational citizenship behavior (OCB), defined as “those organizationally beneficial behaviors and gestures that can neither be enforced on the basis of formal role obligations nor elicited by contractual guarantees or recompense” (Organ, 1990, p. 46), is representative of such behaviors. Dirks (2000) indicates that trust in leaders facilitates employee acceptance of organizational rules and goals. Empirical research based on these theories finds several outcomes of trust—among them high levels of job satisfaction, organizational commitment, and individual and organizational performance; and low levels of turnover (Dirks & Ferrin, 2001).

Of the many kinds of trust within organizations, we focus on trust in supervisors. The power of trust increases when a high risk or vulnerability exists (Dirks, 2000), and the relationship between subordinates and supervisors presents that kind of relationship—at least from the subordinates’ perspective. Employees’ trust in their supervisors is especially critical for performance management in that supervisors play substantial roles throughout the process, including evaluation of employee performance and provision of feedback.

Performance Management and Trust in Supervisors

The research assumes that trust has an indirect effect as well as a direct effect on the given outcomes: perceived work-unit and agency performance. Although most existing studies focus on the direct effect of trust, one can reasonably assume that trust also has an indirect effect in that it nurtures a condition under which management strategies work more effectively (Dirks & Ferrin, 2001). Several studies demonstrate such moderating effect of trust between motivation and group performance (Dirks, 1999), between performance appraisal feedback and individual performance (O’Reilly & Anderson, 1980), and between bargaining toughness and negotiation outcomes (Schurr & Ozanne, 1985).

From a theoretical perspective, expectancy theory provides a good explanation for the role of trust in the process of performance management. Under a high level of trust, two major elements of expectancy theory, expectancy and instrumentality, are enhanced. Expectancy, the perceived probability that employees’ efforts will lead to good performance, will be raised when employees trust that their supervisors fairly evaluate them and provide constructive feedback. On the other hand, instrumentality, which is the perceived probability that employees’ good performance will lead to rewards, will also be increased when employees believe their supervisors will give them rewards faithfully based on performance appraisal. Condrey’s (1995) study provides empirical support for this theoretical perspective. According to the study, federal managers with a high level of trust in the organization showed more positive evaluation of performance appraisal and pay–performance linkage as well as better acceptance of human resource management reform. By managing employees’ expectations, trust in supervisors will increase employees’ acceptance of the performance management system, which in turn will further enhance the effectiveness of performance management.

The positive role of trust is also explained by Kramer’s suggested routes, decreasing transaction costs and facilitating voluntary deference. A high level of trust enables managers to run the performance management system with less monitoring and control of their subordinates. It also elevates employees’ acceptance of the initiative and lets them follow the given rules and procedures (Condrey, 1995; Kramer, 1999). Finally, employees’ trust in their supervisors lays the fundamental groundwork for successful implementation of performance management by nurturing a supportive organizational climate (Condrey, 1995; Gabris & Ihrke, 2000; Reinke, 2003).

Method

This section discusses the data, measurements, and analytic method employed. In addition, we test the seriousness of common method bias, one critical limitation of this research. Harman’s single-factor test demonstrates that the bias is not very serious, which allows us to proceed to the analysis.

Data, Measurements, and Analytic Method

This research uses the Merit Principles Survey 2005 (MPS) to test the effects of performance management and trust. The U.S. Merit Systems Protection Board (MSPB) examines the health of federal merit systems by periodically conducting a governmentwide survey (U.S. Merit Systems Protection Board, 2007). The MSPB conducted its online survey during the summer and fall of 2005. Stratified random sampling was employed to represent all agencies and all levels of employees. Of the 74,000 federal employees who were invited to participate, 36,926 employees completed the survey (U.S. Merit Systems Protection Board, 2007). The survey is useful for this research in that it provides specified questions concerning the four core steps of performance management. The survey also contains several questions asking about trust in supervisors for various aspects.

Like most surveys, MPS 2005 uses a 5-point Likert-type scale for most questions. To minimize measurement error, major independent variables, including performance management and trust in supervisors, are measured by multiple measurements. Mean values are used to integrate multiple measures into single indicators. Table 1 explains measurements of variables.

Table 1.

Measures of Variables

Variable name	Survey items
Work-unit performance	My work unit produces high quality products and services
Agency performance	My agency is successful in accomplishing its mission.
Performance management	I know what is expected of me on the job. I understand the basis for my most recent performance rating. The standards used to appraise my performance are appropriate. I participate in setting standards and goals used to evaluate my job performance. I understand what I must do to receive a high performance rating. I have sufficient opportunities (such as challenging assignments or projects) to earn a high performance rating. I know how my performance rating compares to others in my organization who have similar jobs. In my work unit, performance ratings accurately reflect job performance. Recognition and rewards are based on performance in my work unit.
Trust in supervisors	I trust my supervisor to . . . Listen fairly to my concerns Apply discipline fairly and only when justified Clearly communicate conduct expectations Act with integrity Refrain from favoritism Keep me informed
Self-efficacy	I am treated with respect at work My opinions count at work My job makes good use of my skills and abilities
Individual resources	I have the resources to do my job well.
Organizational resources	The workforce has the job-relevant knowledge and skills necessary to accomplish organizational goals.
Agency experience	How many years have you been with your current Federal Agency?
Supervisory status	What is your supervisory status?
	(a) Nonsupervisor, (b) Team leader, (c) Supervisor, (d) Manager, (e) Executive
	Note: Coded as a binary variable. Nonsupervisor and Team leader were coded as 0.
Female	Are you (a) Male (b) Female
	Note: Coded as a binary variable. Females were coded as 1.
Minority	What race or ethnic category do you consider yourself to be?
	(a) American Indian/Alaskan Native, (b) Asian, (c) Black/African American, (d) Native Hawaiian/Pacific Islander, (e) White, (f) Hispanic or Latino
	Note: Coded as a binary variable. Whites were coded as 0, and others were coded as 1.
Education	What is your current education level?
	(a) High school, GED, or equivalent, (b) Associate’s degree, (c) Bachelor’s degree, (d) Master’s degree, (e) Doctorate or equivalent
	Note: The variable was treated as an ordinal variable.

Both dependent variables, perceived work-unit performance and perceived agency performance, are ordinal in that they are measured by a single indicator. When considering that performance is a multidimensional concept, using a single indicator is a limitation of this research. Focusing on service quality, for example, the work-unit performance measure cannot capture efficiency of work. Likewise, by the current agency performance measure, one cannot examine managerial quality. This limitation requires caution in interpreting the analysis results.

We measure performance management by integrating nine questions that cover a variety of performance management activities in the human resource management process; these include employees’ understanding of the appraisal system, their participation in the system’s design, the transparency of criteria, the proper implementation of the system, and so on. The measurements reflect all four core steps we discussed above. For example, the first four questions reflect the first step, identifying and setting clear goals, whereas the final question deals with the fourth step, utilizing performance appraisal for rewards. Integrating these measures is supported from a statistical perspective, as the Cronbach’s alpha value is .88 and only one factor is retained by the principal component factor analysis.

Trust in supervisors is measured by six items, which concern various aspects of supervisors’ personality and behaviors. The MPS 2005 has a few more measures of trust in supervisors, but we do not include them to clearly distinguish the trust measure from the performance management measure. For example, the item I trust my supervisor to fairly assess my performance and contributions is not included because it is related to the both variables. The principal component factor analysis differentiates trust measures from those of performance measurement (see appendix for the result). Like the performance management measures, the six measures of trust in supervisors are integrated by using a mean value, and the Cronbach’s alpha value of the trust measures is .95.

The control variables, except self-efficacy, are measured by single indicators such as the 5-point Likert-type scale and dummy variables.² Self-efficacy captures individuals’ own perceived impact in the workplace, which may affect their perception of their organizational performance. The measure of individual resources focuses on resources necessary to each employee’s individual job, whereas the measure of organizational resources indicates whether the workforce has the job-specific knowledge and skills to accomplish organizational missions. Three demographic variables—including female, race, and supervisory status—are measured as binary variables. Agency experience is used as an interval variable, whereas education is used as an ordinal variable.

For an analytic method, ordered logit regression is employed because both dependent variables are ordinal. The moderating effect of trust is tested by using the interaction term of performance management and trust. The multicollinearity issue is of particular serious concern when one uses an interaction term in the regression analysis (Friedrich, 1982). To avoid this problem, both performance management and trust measures were standardized in calculating the interaction term.

Test of Common Method Bias

One explicit limitation of this research is using perceived, rather than objective, measures of performance. A related limitation, from a statistical perspective, is the reliance on a single data source, the Merit Principles Survey 2005. Survey respondents with overall positive attitudes might respond to all survey items in a positive manner because they have a tendency to maintain consistency in their responses (Podsakoff & Organ, 1986). Thus, it becomes problematic when both our dependent and independent variables are derived from the perceptions of survey respondents. This problem is called common method bias, a limitation of most empirical studies using self-administered surveys. This bias makes it difficult to discern between respondents’ overall attitudes and the real functional relationships among the variables (Podsakoff & Organ, 1986), which lessens the validity of the analysis.

Although scholars should pay attention to the bias, it does not make survey research useless. The current consensus seems to be that although common source bias does exist, it somewhat inflates investigated relationships rather than completely invalidating the results (Crampton & Wagner, 1994). But, it is necessary to test the seriousness of the bias. To test the degree of the bias, we ran Harman’s single-factor test that implements principal component factor analysis of all the variables in the model. One concludes that the bias is serious when the analysis retains only one factor or a single factor explains most covariance (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). The analysis revealed that eight factors were retained, and the first factor explained only 18% of the variance. To be stricter, we performed the analysis without demographic variables; only two factors were retained and the first factor explained 41% of the entire variance. Although the explained variance increased, it was still less than 50%. The result shows that the bias is not so serious as to make our analysis invalid even though it may still inflate regression coefficients.

Results

Descriptive Statistics

Table 2 explains descriptive statistics, including mean, standard deviation, and zero-order correlations among the variables. Regarding the two dependent variables, work-unit performance received a higher rating than agency performance; the mean value of the former is 4.13, whereas that of the latter is 3.93. Regarding the demographic variables, when compared with the entire population of federal employees, the current sample seems representative although experienced employees are slightly overrepresented.³ In the MPS 2005, 41% of the respondents are females, while 44% are females in the entire population of federal employees in the year of 2005. Minorities are 26% in the sample, while the percentage is 32% in the population in 2006.⁴ Although employees, whose federal service experience is above 30, are 18.4% in the 2005 MPS sample, 10.4% have such length of experience in the entire population. The overrepresentation of experienced employees might be related to underrepresentation of minorities in that minorities are negatively associated with agency experience (–0.07).

Table 2.

Descriptive Statistics

Variables	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12
1. Work-unit performance	4.13	0.81
2. Agency performance	3.93	0.87	0.45**
3. Performance management	3.61	0.78	0.48**	0.45**
4. Trust in supervisors	3.80	1.01	0.39**	0.35**	0.63**
5. Performance Management × Trust	0.63	1.24	−0.17**	−0.13**	−0.29**	−0.36**
6. Self-efficacy	3.76	0.90	0.48**	0.50**	0.69**	0.63**	−0.29**
7. Individual resources	3.53	1.03	0.31**	0.41**	0.42**	0.32**	−0.12**	0.44**
8. Organizational resources	3.89	0.91	0.38**	0.51**	0.39**	0.28**	−0.11**	0.41**	0.39**
9. Agency experience	16.18	9.58	0.08**	0.03**	0.06**	0.04**	−0.03**	0.08**	−0.01	0.03**
10. Supervisory status	0.46	0.49	0.17**	0.10**	0.19**	0.13**	−0.05**	0.20**	−0.06**	0.06**	0.23**
11. Female	0.41	0.49	−0.02**	0.01	−0.00	−0.04**	0.01	−0.04**	0.02**	−0.01*	−0.06**	−0.14**
12. Minority	0.26	0.44	−0.07**	−0.01*	−0.01*	−0.07**	0.02**	−0.07**	0.03**	0.01	−0.07**	−0.08**	0.14**
13. Education	2.89	1.23	0.07**	−0.01	0.02**	0.05**	0.01	0.05**	−0.08**	0.01*	−0.04*	0.18**	−0.17**	−0.11**

Note. N = 30,367.

p < .05. **p < .01.

As expected, both performance management and trust in supervisors are positively correlated with the dependent variables. One can observe that performance management and trust in supervisors are highly correlated; they have the second highest correlation, .63, among the variables in the model. Supervisory status is more strongly correlated with the dependent variables than other demographic factors. Variables such as female, minority, and education level are barely correlated with both dependent variables and performance management.

Ordered Logit Analysis of Perceived Work-Unit Performance

Table 3 explains the ordered logit regression analysis for perceived work-unit performance. Because a large sample size makes all independent variables statistically significant, we focus on leverages of variables by using odds ratios. The final column of the table indicates percent change in odds by one standard deviation change. Performance management has the largest leverage; by one standard deviation, the odds of having a more positive evaluation of work-unit performance increases by 75%, holding all other variables constant. Trust in supervisors is also significantly associated with work-unit performance, but its effect is much smaller than that of performance management; by one standard deviation increase, the odds of having higher work-unit performance is enhanced by 17%.

Table 3.

Ordered Logit Regression of Perceived Work-Unit Performance

	Coefficient	SE	% Change in odds by one SD
Performance management	0.723***	0.023	75.1
Trust in supervisors	0.156***	0.016	17.1
Performance Management × Trust	0.125***	0.011	16.7
Self-efficacy	0.488***	0.021	54.7
Individual resources	0.156***	0.014	17.5
Organizational resources	0.482***	0.015	55.2
Agency experience	0.007***	0.001	7.0
Supervisory status	0.309***	0.025	16.7
Female	0.061*	0.024	3.0
Minority	−0.246***	0.027	−10.2
Education	0.095***	0.010	12.3
Cut 1	2.360
Cut 2	4.093
Cut 3	5.893
Cut 4	8.978
N	30,392
LR chi-square	11,496.83
Pseudo R²	0.1692

p < .05. **p < .01. ***p < .001.

However, trust in supervisors also affects work-unit performance through its indirect effect. One can observe that the interaction term between performance management and trust is positive and significant. Figures 1 and 2 show how levels of trust influence the effect of performance management. From Figure 1, the predicted probability of choosing the “strongly agree” category for work-unit performance is dramatically enhanced as the evaluation of performance management increases. Moreover, the effect of performance management is further elevated under a high level of trust. In contrast, as shown in Figure 2, the predicted probability of “strongly disagree” with excellence of work-unit performance decreases as performance management increases, and the probability is further suppressed with a high level of trust. The difference is largest in the high level of performance management for the “strongly agree” category (roughly a 20 percentage point difference), whereas the largest difference is observed in the lowest level of performance management for “strongly disagree” category (1.5 percentage point difference). Altogether, these figures indicate that trust in supervisors strengthens the relationship between performance management and perceived work-unit performance.

Figure 1.

Predicted probability of “strongly agree” with work-unit performance

Figure 2.

Predicted probability of “strongly disagree” with work-unit performance

Most control variables show expected directions. As a motivational factor, self-efficacy is strongly related to the positive evaluation of work-unit performance. Both individual and organizational resources matter for work-unit performance, but the leverage of organizational resources is three times larger than that of individual resources. Among the demographic factors, supervisory status shows the largest influence. Supervisors, who commonly have a higher stake in organizations than do nonsupervisors, may perceive themselves more integral to their organizations (Whitener, Brodt, Korsgaard, & Werner, 1998) and evaluate their organizations more positively. Employees with longer agency experience, employees who are highly educated, and female employees give a more positive evaluation of work-unit performance than do others. On the other hand, minorities show more negative evaluations than do Whites. This might come from their negative experience in work groups caused by cultural differences and perceived discrimination (Jackson & Alvarez, 1992), but more investigation is required to confirm the argument.

Ordered Logit Analysis of Perceived Agency Performance

As shown in Table 4, performance management is substantially associated with perceived agency performance. Compared with the previous analysis, however, the leverage is decreased by half; by a one standard deviation increase of performance management, the odds of having higher agency performance is enhanced by 35%, holding all other variables constant.

Table 4.

Ordered Logit Regression of Perceived Agency Performance

	Coefficient	SE	% Change in odds by one SD
Performance management	0.389***	0.023	35.2
Trust in supervisors	0.054***	0.016	5.6
Performance Management × Trust	0.138***	0.010	18.6
Self-efficacy	0.665***	0.021	81.2
Individual resources	0.348***	0.014	43.2
Organizational resources	0.876***	0.015	122.3
Agency experience	−0.004***	0.001	−4.0
Supervisory status	0.181***	0.025	9.4
Female	0.086***	0.024	4.3
Minority	−0.051	0.027	−2.2
Education	−0.039***	0.010	−4.6
Cut 1	3.227
Cut 2	5.188
Cut 3	7.073
Cut 4	10.400
N	30,367
LR chi-square	14,888.72
Pseudo R²	0.2046

p < .05. **p < .01. ***p < .001.

The percent change in odds ratio shows that trust in supervisors has a minimal direct effect on perceived agency performance. However, it also has an indirect effect; from the significant interaction term, it affects agency performance through the process of performance management. Figure 3 indicates that the predicted probability of “strongly agree” with agency performance increases as performance management is elevated, and the probability is further promoted when employees have a high level of trust in supervisors. On the other hand, the probability of “strongly disagree” with agency performance decreases as the level of performance management is enhanced (Figure 4). Under a high level of trust, the probability is further depressed. Again, the value of trust as a managerial resource is demonstrated in that it facilitates the effectiveness of performance management.

Figure 3.

Predicted probability of “strongly agree” with agency performance

Figure 4.

Predicted probability of “strongly disagree” with agency performance

However, we should not overestimate the effects of performance management and trust in supervisors on agency performance. Control variables, including self-efficacy, individual, and organizational resources, have much higher percentages of odds than performance management. Especially, organizational resources have the largest coefficient among the variables as its importance is already demonstrated in several studies (e.g., Boyne, 2003; Rainey & Steinbauer, 1999). Demographic variables except minority are statistically significant, but their leverages are negligible; the percentage of change in odds by one standard deviation is less than 10%.

Discussion

The results of our analysis enable us to answer both research questions in the affirmative. At least perceptually, performance management drives organizational performance—it positively affects both work-unit performance and agency performance. Although there continue to be challenges, performance management can bring positive results to federal workplaces if properly designed and implemented. This research also demonstrates the value of trust. Although its effect is not dominant in the model, trust in supervisors lays fertile ground for successful performance management; it strengthens the relationships between performance management and the considered outcomes.

We have already mentioned several limitations of this research, which project some future research directions. From the methodological perspective, although Harman’s single-factor test demonstrates that the common source bias is not serious, one may want to employ multiple data sources to completely remove the bias. If future research employing objective performance measures shows an identical result, that result will increase the validity of our finding. Using multiple measures incorporating diverse dimensions of performance is also desired. Related to that, using a cross-sectional data set prevents us from developing causal relationships. Even though we ground our discussion on theories and previous research, the current analysis reveals associations among variables rather than causal relationships. To address this limitation, a sophisticated framework such as time-series analysis is desired. This research employs a micro approach by defining performance management as a human resource management tool. However, investigating the issue in a comprehensive way is also required. Measuring the concept comprehensively by considering managerial practices as well as organizational strategies may reveal multidimensions of performance management and how those dimensions are connected each other. In addition, replication in different settings is necessary. We will be more confident in the effect of performance management and the value of trust when we observe similar results at the state and local level.

In spite of the above limitations, the current research still provides some implications. From a theoretical perspective, the result implies that Theory X and Theory Y should go together for the maximum effectiveness of management tools. Four core steps of performance management include both Theory X components and Theory Y components. The reinforcement mechanism of rewarding high performers and punishing low performers has Theory X components, whereas facilitating employee participation and providing feedback reflect Theory Y components. The system will be effective when both components are integrated, and the effectiveness will be further enhanced under a high level of trust, the proliferation of Theory Y. When federal agencies incorporate both components in their management systems and facilitate trusting relationships among the workforce, one can expect better organizational outcomes.

That theoretical implication naturally leads to some practical implications about building trust. Having trustworthy supervisors who exercise fairness in managerial activities and take care of their subordinates is critical to the successful implementation of performance management. Accordingly, top management should be attentive to cultivating a fair and benevolent culture within organizations. Developing training programs to make trustworthy supervisors deserves attention. Another practical implication comes from designing the performance management system. As discussed, the four core steps must be presented to be successful; management should clarify goals and objectives, fairly measure performance and provide feedback, and utilize performance information as major determinants of rewards and recognition (Schermerhorn et al., 2008). In addition, federal agencies may want to consider employee participation in the system design.

Nevertheless, we also need caution in interpreting and applying the current findings. Historically, the performance management movement has not been very successful. Although various reform initiatives were adopted since the CSRA of 1978, their success was limited. For example, the Performance Management and Recognition System that was introduced in 1984 sought to replace the previous Merit Pay System, but it was discontinued (Condrey, 1995). Recently, the National Security Personnel System was created for civilian employees in the Department of Defense to provide more flexibility in human resource management. However, the system was also repealed with much controversy and resistance from employees (Haga et al., 2010). Some scholarly findings also echoed those negative results. Managers did not believe that merit pay awards would be based on their performance (Pearce & Perry, 1983), and merit pay failed to increase employee performance (Gaertner & Gaertner, 1984). When considering the limitation of the current research using the perceived measure of performance, we need more solid evidence for achievement of performance management. Finally, although this research suggests some conditions for successful performance management, more research is desired.

This research begins to fill the research gap in areas of performance management and trust in the public management field. More empirical research is required to validate current arguments of scholars and practitioners. In addition, there is much knowledge to be obtained to build a better performance management system and develop and maintain trust within organizations. We hope this research will facilitate efforts pursuing new knowledge in these important areas.

Footnotes

Appendix

Factor Analysis of Performance Management and Trust in Supervisors

Variables	Survey questions	Factor loadings
Performance management	I know what is expected of me on the job.	0.3445	0.5104
	I understand the basis for my most recent performance rating.	0.3580	0.6169
	The standards used to appraise my performance are appropriate.	0.4265	0.5762
	I participate in setting standards and goals used to evaluate my job performance.	0.3042	0.7523
	I understand what I must do to receive a high performance rating.	0.2736	0.7624
	I have sufficient opportunities (such as challenging assignments or projects) to earn a high performance rating.	0.2401	0.6862
	I know how my performance rating compares to others in my organization who have similar jobs.	0.2909	0.7657
	In my work unit, performance ratings accurately reflect job performance.	0.3528	0.6930
	Recognition and rewards are based on performance in my work unit.	0.0920	0.5308
Trust in supervisors	I trust my supervisor to . . .
	Listen fairly to my concerns	0.8781	0.2388
	Apply discipline fairly and only when justified	0.8721	0.2212
	Clearly communicate conduct expectations	0.8496	0.2502
	Act with integrity	0.8932	0.1947
	Refrain from favoritism	0.8786	0.2126
	Keep me informed	0.8446	0.2450

Note. Principal component factor analysis is used with Varimax rotation.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Bios

Yoon Jik Cho is an assistant professor of Public Administration Department at Yonsei University in Korea. His research focuses on role of trust in public management, human resources management, and leadership.

Jung Wook Lee is an assistant professor of public administration at Yonsei University, Seoul, Korea. Prior to joining Yonsei in 2009, he taught at the University of Illinois-Springfield, U.S.A. His research focuses on government performance and the political context of public management.

References

Abramson

M. A.

Breul

J. D.

Kamensky

J. M.

(2006). Six trends transforming government. Washington, DC: IBM Center for the Business of Government.

Adams

J. S.

(1965). Inequity in social exchange. In Berkowitz

(Ed.), Advances in experimental and social psychology (Vol. 2, pp. 267-299). Orlando, FL: Academic Press.

Behn

R. D.

(2002). The psychological barriers to performance management. Public Performance & Management Review, 26(1), 5-25.

Bilgin

K. U.

(2007). Performance management for public personnel: Multi-analysis approach toward personnel. Public Personnel Management, 36(2), 93-113.

Boyd

B. K.

Dess

G. G.

Raheed

A. M. A.

(1993). Divergence between archival and perceptual measures of the environment: Causes and consequences. Academy of Management Review, 18, 204-226.

Boyne

G. A.

(2003). Sources of public service improvement: A critical review and research agenda. Journal of Public Administration Research and Theory, 13, 367-394.

Brook

D. A.

King

C. L.

(2008). Federal personnel management reform: From Civil Service Reform Act to national security reforms. Review of Public Personnel Administration, 28, 205-221.

Condrey

S. E.

(1995). Reforming human resource management systems: Exploring the importance of organizational trust. American Review of Public Administration, 25, 341-354.

Crampton

S. M.

Wagner

J. A.

III. (1994). Percept-percept inflation in microorganizational research: An investigation of prevalence and effect. Journal of Applied Psychology, 79(1), 67-76.

10.

Dirks

K. T.

(1999). The effects of interpersonal trust on work group performance. Journal of Applied Psychology, 84, 445-455.

11.

Dirks

K. T.

(2000). Trust in leadership and team performance: Evidence from NCAA basketball. Journal of Applied Psychology, 85, 1004-1012.

12.

Dirks

K. T.

Ferrin

D. L.

(2001). The role of trust in organizational settings. Organization Science, 12, 450-467.

13.

Fairholm

G. W.

(1994). Leadership and the culture of trust. Westport, CT: Praeger.

14.

Forsythe

D. W.

(2001). Pitfalls in designing and implementing performance management systems. In Forsythe

D. W.

(Ed.), Quicker, better, cheaper: Managing performance in American government (pp. 519-551). Albany, NY: Rockefeller Institute Press.

15.

Friedrich

R. J.

(1982). In defense of multiplicative terms in multiple regression equations. American Journal of Political Science, 26, 797-833.

16.

Gabris

G. T.

Ihrke

D. M.

(2000). Improving employee acceptance toward performance appraisal and merit pay systems: The role of leadership credibility. Review of Public Personnel Administration, 20(1), 41-53.

17.

Gaertner

K. N.

Gaertner

G. H.

(1984). Performance evaluation and merit pay: Results in the Environmental Protection Agency and the Mine Safety and Health Administration. In Graham

P. W.

Ban

(Eds). Legislating bureaucratic change: The Civil Service Reform Act of 1978 (pp. 87-111). Albany: State University of New York Press.

18.

Grote

(2000). Public sector organizations: Today’s innovative leaders in performance management. Public Personnel Management, 29(1), 1-20.

19.

Haga

B. I.

Ruchman

Leavitt

(2010). System failure: Implementing pay for performance in the Department of Defense’s National Security Personnel System. Public Personnel Management, 39, 211-230.

20.

Hays

(2004). Trends and best practices in state and local human resources management: Lessons to be learned. Review of Public Personnel Administration, 24, 256-275.

21.

Heinrich

C. J.

(2007). Evidence-based policy and performance management. American Review of Public Administration, 37, 255-277.

22.

Jackson

Alvarez

(1992). Working through diversity as a strategic imperative. In Jackson

(Ed.), Diversity in the workplace (pp. 13-29). New York, NY: Guilford.

23.

Kettl

D. F.

(2000). The global public management reform: A report on the transformation of governance. Washington, DC: Brookings Institution Press.

24.

Kelman

(2007, December). Public administration and organizational studies. The Academy of Management Annals, 1, 225-267.

25.

Kramer

R. M.

(1999). Trust and distrust in organizations: Emerging perspectives, enduring questions. Annual Review of Psychology, 50, 569-598.

26.

Lah

T. J.

Perry

J. L.

(2008). The diffusion of the Civil Service Reform Act of 1978 in OECD countries: A tale of two paths to reform. Review of Public Personnel Administration, 28, 282-299.

27.

Latham

G. P.

Almost

Mann

Moore

(2005). New developments in performance management. Organizational Dynamics, 34(1), 77-87.

28.

Lee

J. W.

Cho

Y. J.

Kim

S. E.

(2009). Improving performance: Does performance-oriented management really matter? International Review of Public Administration, 13(3), 17-34.

29.

Locke

E. A.

Latham

G. P.

(2002). Building a practically useful theory of goal setting and task motivation: A 35-year odyssey. American Psychologist, 57, 705-717.

30.

Melkers

Willoughby

(1998). The state of the states: The performance-based budgeting requirements in 47 out of 50. Public Administration Review, 58(1), 66-73.

31.

Melkers

Willoughby

(2004). Staying the course: The use of performance measurement in state governments. Washington, DC: IBM Center for the Business of Government.

32.

Moynihan

D. P.

(2008). The dynamics of performance management. Washington, DC: Georgetown University Press.

33.

O’Reilly

III. Anderson

(1980). Trust and the communication of performance appraisal information: The effect of feedback on performance and job satisfaction. Human Communication Research, 6, 290-298.

34.

Organ

D. W.

(1990). The motivational basis of organizational citizenship behaviors. In Cummings

L. L.

Straw

B. M.

(Eds.), Research in organizational behavior (pp. 43-72). Greenwich, CT: JAI Press.

35.

Organisation for Economic Co-operation and Development. (2005). Modernising government: The way forward. Paris, France: Author.

36.

Ostrom

(1998). A behavioral approach to the rational choice theory of collective action: Presidential address, American Political Science Association, 1997. American Political Science Review, 92(1), 1-22.

37.

Pearce

J. L.

Perry

J. L.

(1983). Federal merit pay: A longitudinal analysis. Public Administration Review, 43, 315-325.

38.

Perrin

(2006). Moving from outputs to outcomes: Practical advice from governments around the world. Arlington, VA: IBM Center for the Business of Government.

39.

Perry

J. L.

Engbers

T. A.

Jun

S. Y.

(2009). Back to the future? Performance-related pay, empirical research, and the perils of persistence. Public Administration Review, 69(1), 39-51.

40.

Peters

(2008). Debate: The two futures of public administration. Public Money and Management, 28, 195-196.

41.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

Podsakoff

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879-903.

42.

Podsakoff

P. M.

Organ

D. W.

(1986). Self-reports in organizational research: Problems and prospects. Journal of Management, 12, 531-544.

43.

Poister

T. H.

(2003). Measuring performance in public and nonprofit organizations. San Francisco, CA: Jossey-Bass.

44.

Poister

T. H.

Streib

(1999). Performance measurement in municipal government: Assessing the state of the practice. Public Administration Review, 59, 325-335.

45.

Poister

T. H.

Streib

(2005). Elements of strategic planning and management in municipal government: Status after two decades. Public Administration Review, 65(1), 45-56.

46.

Pollitt

(2001). Integrating financial management and performance management. OECD Journal on Budgeting, 1(2), 7-37.

47.

Pollitt

Bouckaert

(2000). Public management reform. Oxford, UK: Oxford University Press.

48.

Radin

(2000). Beyond Machiavelli: Policy analysis comes of age. Washington, DC: Georgetown University Press.

49.

Radin

(2006). Challenging the performance movement: Accountability, complexity, and democratic values. Washington, DC: Georgetown University Press.

50.

Rainey

H. G.

Steinbauer

(1999). Galloping elephants: Developing elements of a theory of effective government organizations. Journal of Public Administration Research and Theory, 9(1), 1-32.

51.

Reinke

S. J.

(2003). Does the form really matter? Leadership, trust, and acceptance of the performance appraisal process. Review of Public Personnel Administration, 23(1), 23-37.

52.

Risher

Fay

C. H.

(2007). Managing for better performance: Enhancing federal performance management practices. Washington, DC: IBM Center for the Business of Government.

53.

Rousseau

D. M.

Sitkin

S. B.

Burt

R. S.

Camerer

(1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23, 393-404.

54.

Sanger

M. B.

(2008). From measurement to management: Breaking through the barriers to state and local performance [Special issue]. Public Administration Review, 68, S70-S85.

55.

Schermerhorn

J. R.

Hunt

J. G.

Osborn

R. N.

(2008). Organizational behavior (10th ed.). New York, NY: Wiley.

56.

Schurr

Ozanne

(1985). Influences on exchange processes: Buyers’ preconceptions of a seller’s trustworthiness and bargaining toughness. Journal of Consumer Research, 11, 939-953.

57.

Talbot

(2005). Performance management. In Ferlie

Lynn

L. E.

Pollitt

(Eds.), The Oxford handbook of public management (pp. 491-517). New York, NY: Oxford University Press.

58.

U.S. Merit Systems Protection Board. (2007). Accomplishing our mission: Results of the Merit Principles Survey 2005. Retrieved from http://www.mspb.gov/sites/mspb/default.aspx

59.

Vroom

V. H.

(1964). Work and motivation. New York, NY: Wiley.

60.

Whitener

E. M.

Brodt

S. E.

Korsgaard

M. A.

Werner

J. M.

(1998). Managers as initiators of trust: An exchange relationship framework for understanding managerial trustworthy behavior. Academy of Management Review, 23, 513-530.

61.

Williamson

(1993). Calculativeness, trust and economic organization. Journal of Law & Economics, 36, 453-486.

62.

Yang

Hsieh

J. Y.

(2007). Managerial effectiveness of government performance measurement: Testing a middle-range model. Public Administration Review, 67, 861-879.

63.

Yang

Pandey

S. K.

(2009). How do perceived political environment and administrative reform affect employee commitment? Journal of Public Administration Research and Theory, 19, 335-360.

64.

Zaheer

McEvily

Perrone

(1997). Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance. Organization Science, 9, 141-159.

65.

Zand

D. E.

(1997). The leadership triad: Knowledge, trust, and power. New York, NY: Oxford University Press.