The Leadership Quarterly 22 (2011) 92105

Contents lists available at ScienceDirect

The Leadership Quarterly

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / l e a q u a

CEO leadership behaviors, organizational performance, and

employees' attitudes
Hui Wang a,, Anne S. Tsui b,c, Katherine R. Xin d
a
Guanghua School of Management, Peking University, Beijing 100871, China
b
W. P. Carey School of Business, Arizona State University, USA
c
Peking University, China
d
China Europe International Business School (CEIBS), China

a r t i c l e i n f o a b s t r a c t

Available online 4 February 2011 This study explores the links between CEO leadership behaviors, firm performance and
employees' attitudes in a sample drawn from 125 firms in China. We first inductively identified
Keywords: categories of CEO leadership behaviors in the Chinese context. Through a factor analysis, we
CEO leadership behaviors developed a six-dimension measure of CEO leadership behaviors, with three dimensions focused
Firm performance on tasks and three dimensions focused on relationships. Our hypotheses were tested on a matched
Employees' attitudes data set including 739 middle managers and their supervisors (top managers) in the 125 firms.
Results from the structural equation modeling analysis show that the CEO's task-focused
behaviors are directly linked to firm performance. The CEO's relationship-focused behaviors are
related to employees' attitudes and, through these attitudes, to firm performance. Limitations and
implications for future research are discussed.
2010 Elsevier Inc. All rights reserved.

1. Introduction

Many studies have explored the link between executive leadership and organizational outcomes (see reviews by Boal &
Hooijberg, 2000; Cannella & Monroe, 1997; Carpenter, Geletkanycz, & Sanders, 2004; Hunt, 1991), but the results are inconsistent.
Some studies showed that chief executive ofcers (CEOs) are critically important for an organization to achieve a high level of
performance (e.g., Finkelstein & Hambrick, 1996; Katz & Kahn, 1978; Peterson, Smith, Martorana, & Owens, 2003; Thomas, 1988).
Other scholars have argued that CEOs are inconsequential to organizational effectiveness (e.g., Lieberson & O'Connor, 1972;
Meindl, Ehrlich, & Dukerich, 1985; Pfeffer, 1977). These mixed results were also found in recent studies (Agle, Nagarajan,
Sonnenfeld, & Srinivasan, 2006; Ling, Simsek, Lubatkin, & Veiga, 2008a; Ling, Simsek, Lubatkin, & Veiga, 2008b; Tosi, Misangyi,
Fanelli, Waldman, & Yammarino, 2004; Waldman, Javidan, & Varella, 2004). Even though there is a consensus in the literature that
the CEO matters to rm performance, there is no clear answer to the question of which aspects of CEO characteristics are important
to organizational effectiveness and how a CEO affects organizational performance. To resolve this puzzle, another group of scholars
argued that CEOs would have the most impact on rm outcomes during periods of crises (House, Spangler, & Woycke, 1991) or in
environments with high uncertainty (Waldman, Ramirez, House, & Puranam, 2001). The present study aims to contribute to this
discussion by exploring the question of how CEO's leadership behaviors may be related to rm outcomes in a context where some
CEOs should be able to exert great inuence. The context is the People's Republic of China where traditional rms are currently
experiencing major corporate transformations along with a corresponding rise in many new and young entrepreneurial rms
(Tsui & Lau, 2002). This condition creates uncertainty for rms and provides a context in which leaders should have high levels of
discretion to inuence rm outcomes (Boal & Hooijberg, 2000; Hambrick & Finkelstein, 1987). As argued by Boal (2004), the

Corresponding author. Tel.: +86 10 6275 3645; fax: +86 10 6275 1463.
E-mail address: wanghui@gsm.pku.edu.cn (H. Wang).

1048-9843/$ see front matter 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.leaqua.2010.12.009
 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105 93

greater the discretion a CEO has, the more inuence the CEO's choices, including decisions and their behavioral tendencies or
leadership style, will have on organizational outcomes.
Further, in traditional Chinese values, leaders have supreme authority over the organization and are the primary decision
makers, with employees looking to the top leader for guidance and approval. The CEO is an important (father) gure in the
organization whose behaviors are observed and internalized (Farh & Cheng, 2000). This context also provides a good setting to
investigate the mediating role of employee reactions to the CEO's leadership in the relationship between a CEO's behaviors and
rm performance.
Different theoretical perspectives focus on the relevance of different CEO characteristics that may inuence rm outcomes
(Boal, 2004; Cannella & Monroe, 1997). One of the most popular theories of executive inuence is the upper echelon theory
(Hambrick & Mason, 1984), which argues for the importance of demographic factors (age, functional background, career
experiences, education, socioeconomic roots, etc.) as proxies for psychological aspects (cognition, values, and personality) in
inuencing rm outcomes. A number of studies have investigated the link between demographic variables of top management
teams and rm performance (e.g., Carpenter & Fredrickson, 2001; Hambrick, Cho, & Chen, 1996; Smith et al., 1994).
Other studies have tested the link between perceptions of CEO charisma and organizational performance. House, Spangler, and
Woycke (1991) found the charisma of the U.S. president to be associated with presidential performance, particularly during
periods of crisis. Using an industrial sample, Waldman et al. (2001) also showed the CEO's charismatic leadership to be
signicantly related to rm performance. Another study by Waldman and his colleagues Waldman et al. (2004) using 69 public
U.S. and Canadian rms showed that CEO charisma was related to subsequent organizational performance as measured by net
prot margins and return on equity. However, Tosi et al. (2004) did not nd a direct relationship between CEO charisma and
organizational performance. In addition, Agle et al. (2006) conducted a study to show that organization performance was
associated with subsequent perceptions of CEO charisma and that prior perceptions of CEO charisma did not lead to subsequent
organizational performance. It is possible that charisma may be an attribution rather than a cause of rm performance (Meindl et al.,
1985). In a very recent study, Ling et al. (2008b) found that a CEO's transformational leadership inuenced rm performance in
small, privately held rms.
We aim to contribute to a further understanding of this puzzle by examining the relationship between the CEO's leadership
behaviors and rm performance, mediated by employees' attitudes, an assumed mediation that has not been directly tested.
Previous studies have shown that aggregated employees' attitudes are positively related to rm performance (Harter, Schmidt, &
Hayes, 2002; Ostroff, 1992; Ryan, Schmit, & Johnson, 1996; Schleicher, Watt, & Greguras, 2004; Schneider, Hanges, Smith, &
Salvaggio, 2003) but these studies have not linked aggregated employees' attitudes to CEO's leadership behaviors.

2. Theory and hypotheses

2.1. CEO leadership behaviors

To respond to the call for a moratorium on the use of demographic variables as surrogates for psychosocial constructs (Boal &
Hooijberg, 2000: p.523), scholars have begun to focus on the inuence of the behaviors of CEOs on rm performance (e.g., Colbert,
Kristof-Brown, Bradley, & Barrick, 2008; Ling et al., 2008a; Ling et al., 2008b; Waldman et al., 2001). Most recent scholars have focused
on charismatic leadership and transformational leadership and their effects on organizational effectiveness. Charismatic leadership is
conceptually similar to transformational behavior. Shamir, House and Arthur (1993: p.1) stated, Such leadershipalternatively called
charismatic, visionary, or transformationalis claimed to affect followers in ways that are quantitatively greater and qualitatively
different than the effects specied in past theories. As described by Cannella and Monroe (1997), in contrast to the agency theory of
leadership, transformational leadership emphasizes the importance of leaders' relationships with followers. Key behaviors of
charismatic/transformational leaders include articulating a vision, making sense of missions, showing determination, and
communicating high performance expectations. The favorable effects of charismatic/transformational leadership behavior on
followers include generating followers' condence in the leader, making followers feel good in the leader's presence, and obtaining
strong admiration or respect from employees (Conger & Kanungo, 1987; House et al., 1991). In contrast, transactional leadership
behaviors focus on the motivation of followers through rewards or punishment. According to Burns (1978), transactional leadership is
related to an exchange relationship that meets the exchange partners' own self-interests. Transactional leadership behaviors primarily
aim at the maintenance and monitoring of organizational operations. These leadership behaviors shape the strategies to be pursued by
the organization, develop the structure to implement them, direct subordinates' efforts and attention, and correct any mistakes or
deviations from expectations. These actions are directed at enhancing organizational performance either through dealing with the
tasks directly or through inuencing the behaviors of followers.
Clearly, charismatic/transformational and transactional leadership involve two types of behavior. One focuses on the tasks or
performance of the rm, such as planning, articulating the vision or goals for the organization, monitoring subordinate activities,
and providing necessary support, equipment and technical assistance. The other focuses on relationships with employees,
including being supportive of and helpful to subordinates, showing trust and condence in employees, being friendly and
considerate, trying to understand subordinates' problems, showing appreciation for a subordinate's ideas, and providing
recognition for subordinates' contributions and accomplishments (Yukl, 2002). In fact, the task-oriented and relationship-oriented
behavior dichotomy of leadership is rooted in early research work at Ohio State University and the University of Michigan
(Stogdill, 1974).
94 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105

In an empirical study of Chinese organizations, Chen and Farh (1999) demonstrated that transformational leadership behaviors
could be grouped into two higher-order factors, one relationship oriented and the other task-oriented. They further showed that
relationship-oriented leadership behaviors have stronger impacts than do task-oriented leadership behaviors on employees' in-role
performance, organizational citizenship behaviors, and job satisfaction. In the current study, we focus on these two higher-order
factors of task- and relationship-oriented leadership behaviors of CEOs.

2.2. Task- and relationship-oriented behaviors of CEOs

CEOs inuence organizational effectiveness mainly through the roles they play and the behaviors they display during the
process of playing their roles (Quinn, 1988). Hooijberg and Quinn (1992) suggested that effective leaders are not only cognitively
complex but also able to perform a diverse set of roles and skills in the explicit behavioral realm. Behavioral complexity is the
ability to act out a cognitively complex strategy by playing multiple, even competing roles, in a highly integrated and
complementary way (Hooijberg & Quinn, 1992, p.164). Boal and Hooijberg (2000) further proposed that behavioral complexity
allows a CEO to communicate with different stakeholders effectively and to increase his/her capacity to change the organization.
Leadership behaviors are embedded in the executive leadership roles identied in the Competing Values Framework (Quinn,
1988; Quinn & Cameron, 1988). Hart and Quinn (1993) proposed a broad portfolio of CEO leadership roles including vision setter,
motivator, analyzer, and task master. The vision setter role is related to dening and articulating the rm's basic purpose and future
directions. To fulll this role, the executive leader must spend considerable time monitoring and studying emerging social,
economic, and technological trends. During this process, the leader selects relevant information from the environment and sets up
an appropriate vision for the organization. The motivator role refers to translating the vision of the rm into a cause worth ghting
for. To play this role, the leader must create a sense of excitement and vitality within the organization to motivate employees to
accomplish the organization's goals.
In playing the analyzer role, the leader focuses on making the internal operating system efcient. An executive leader sets the
context, shapes the decisions made by the operating system, and maintains control over the process of management. Finally, in the
task master role, the top manager focuses on the rm's performance and responsibilities. In a narrow sense, this role is associated
with economic performance and the demands of the market. In a broader sense, it results in social performance and responsibility.
To fulll this role, an executive leader needs not only to inuence decisions made at lower levels, but also to make trade-offs
explicit and to allocate resources to the highest priority activities.
A close examination suggests that these four roles are consistent with task-oriented and relationship-oriented leadership
behaviors. The roles of analyzer and task master are concerned with how CEOs evaluate proposed projects and make trade-off
decisions and allocate resources (Hart & Quinn, 1993: p.558). Through these roles, top managers focus on the efcient
management of the internal operating systems of their rms to serve existing product markets. The direct results of these two
roles are effective management systems, sustainable organizational structures, and solid nancial performance. The role of vision
setter requires a CEO to be sensitive to changes in the environment, scan the environment for information relevant to rm
development and growth, and set up a relevant vision for the organization. The company's mission and goals will align behaviors
of employees with their efforts. The nancial performance of the rms will consequently improve. This vision setter role therefore
also relates to a task focus and may have implications on rm performance.
On the other hand, the role of motivator is directly linked to employees' attitudes towards their organization. Specically, the
motivator role helps the CEO to maintain contact with followers and to ensure that basic and expected values are shared among
employees. Through behaviors that challenge followers with new goals and aspirations and create a sense of excitement (Hart
& Quinn, 1993: p.558), a CEO can communicate the vision of the organization among followers through personal examples and
inspirational articulation. Therefore, employees led by CEOs who are strong motivators are more likely to understand the
assumptions of their leaders, share their values, and identify with their mission and values. They will express high levels of
organizational commitment, perceive support from the organization as a whole, and have a positive perception of fairness.
The results of empirical research showed that leaders who perform multiple leadership roles enjoy high levels of organizational
effectiveness (Denison, Hooijerg & Quinn, 1995; Hart & Quinn, 1993). This is consistent with research that showed that high-
performing managers possess high levels of cognitive complexity (e.g., Streufert & Swezey, 1978) and are able to utilize multiple
frames of reference in dealing with problems (Dreyfus, Dreyfus, & Athanasion, 1986). Denison et al. (1995) also showed that more
effective executives played a greater variety of leadership roles than did their less effective counterparts.
In summary, the literature on executive leadership roles converges on identifying two meta roles played by CEOs, one directly
focusing on inuencing the accomplishment of the organization's tasks and the other focusing on motivating employees to have
organizational commitment and to pay attention to accomplishing organizational goals.

2.3. Leadership behaviors, employees' attitudes, and rm performance

There is robust evidence that employees' attitudes aggregated at the rm level have a positive relationship with the
performance of the rm (e.g., Harter et al., 2002; Koys, 2001; Ostroff, 1992; Ryan et al., 1996; Schleicher et al., 2004; Schneider et al.,
2003) and a negative relationship with employee turnover (Ryan et al., 1996). Employees' attitudes could be affected by a wide
variety of factors, including supervisory support, coworker inuences, and the employee's own personal attributes. We argue that
the CEO is another causal agent by inducing positive attitudes among employees (such as perceived organizational support and
 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105 95

organizational commitment). In this way, employees' attitudes are a mediating link between CEO leadership behaviors and rm
performance.
One possible way in which aggregated employees' attitudes mediate the link between CEO behaviors and rm performance is
through the idea of shared values. According to Ryan et al. (1996), if employees of an organization have shared values, they are
more likely to cooperate and collaborate with others, which in turn improve organizational performance. The CEO can motivate
employees to work hard to accomplish goals by using rewards and punishments. More importantly, the charismatic behaviors of
the CEO can motivate employees to do more than they are expected to do (Yukl, 1989: p.272). Such behaviors by the CEO help
employees to transcend their own self-interests for the sake of achieving the missions of their groups and/or organizations (Avolio
& Bass, 1987; Shamir et al., 1993).
Studies have largely conrmed a positive relationship between charismatic (Bryman (1992)) or transformational (Podsakoff,
MacKenzie, Moorman, & Fetter, 1990; Podsakoff, MacKenzie, & Bommer, 1996) leaders' behaviors and employees' satisfaction,
organizational commitment, trust, self-reported effort, and supervisor-rated performance at the individual level. Evidence on the
relationship between CEO leadership behaviors and employees' attitudes is more limited. In a recent meta-analysis study, Judge,
Piccolo, and Illies (2004) reported that the consideration dimension (relationship-oriented behavior) and the initiating structure
dimension (task-oriented behavior) of leadership behavior at the supervisory level have a moderately strong relationship with an
employee's job satisfaction, job performance, and group performance. Consequently, we expect that task-focused behaviors of the
CEO will be directly related to rm performance and that their relationship-focused behaviors will be associated with positive
employees' attitudes and, through these attitudes, rm performance. Our hypotheses are formally stated as follows:

Hypothesis 1. The CEO's task-focused behaviors will relate positively to organizational performance and his/her relationship-focused
behaviors will relate positively to employees' positive attitudes toward the organization.

Hypothesis 2. Employees' attitudes toward the organization will mediate the relationship between the CEO's relationship-focused
behaviors and rm performance.

When discussing the conditions under which and when CEOs have inuence on rm performance, Boal and Hooijberg
(2000) argued that in situations with wide latitudes of choice and action, individual differences in top managers may play key roles
in strategic choices and organizational outcomes. Cannella and Monroe (1997) proposed that Strategic leadership theory also
implies that executives need discretion, or latitude of action, in order to serve effectively as top managers (Cannella & Monroe,
1997: p.220). Such discretion may increase the capacity of CEOs to make a difference. We tested the above hypotheses using CEOs
in China because they operate in an environment characterized by rapid environmental shifts (Tsui, Schoonhoven, Meyer, Lau, &
Milkovich, 2004). In highly uncertain and unstable environments, the chief executive can provide the organization with direction
about the future. The transition from a planned economy to free market competition in China has created a highly uncertain task
environment for all rms. The context of China provides a unique and useful blend of historical, political, cultural, social, and
judicial changes that are rare in contemporary societies (Tsui et al., 2004). This context mirrors the condition of environmental
uncertainty under which CEOs would have more discretion (Hambrick & Finkelstein, 1987). Further, due to the lack of regulations
and standard norms of modern enterprise systems, CEOs in China have great latitude for discretion in a society characterized by
high power distance and traditional values of deference to authority and submission (Yang, 1993).
In linking CEO leadership behaviors to rm performance, we focus on the attitudes of employees who are proximal to the CEO
in the organizational hierarchy, i.e., the middle managers of the company. Middle managers have more opportunities to observe
the behaviors and actions of the CEO than do employees at lower levels in the hierarchy. They are responsible for carrying out the
goals set by top managers. They are involved in the day-to-day workings of the company and they may provide valuable
information to top managers to improve the organization's bottom line (Schilit, 1987). Through positive attitudes, these middle
managers will work toward achieving a high level of rm performance and will be committed to the organization. Turnover among
experienced managers is costly to the rm. Their positive attitudes will further develop a positive climate and they will serve as
positive role models to lower-level employees. While the middle managers may not represent the entire organization, their
attitudes toward the rm are important in their own right due to the above reasons. Findings from this study will contribute to
understanding leadership in China and also to leadership research in general.

3. Method

The present study proceeded in two stages. The rst stage was to develop a scale to measure CEO leadership behaviors that was
meaningful in the study context of China. We started with existing measures of charismatic/transformational and transactional
leadership, but supplemented these measures through an inductive effort to localize some of the indicators (items) and with a
scale-development factor analysis using responses from 542 managers and professional employees. In the second stage, structural
equation modeling analysis with a new sample of 739 middle managers and their supervisors in 125 companies was used to test
our hypotheses.

3.1. Scale-development study

We employed an inductive approach to identify CEO behaviors that are important to rm performance in the Chinese context
for two reasons. First, China is experiencing a very fast change from a planned economy to a market-oriented economy. The
96 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105

transitional economy and its deep culture suggest that CEOs may play roles that may be different from the roles of their
counterparts in the West. We believed that the inductive approach would help us to discover behaviors that may be different from
the existing measures of task and relationship orientations. Second, House et al. (1999) argued that the attributes rooted in a given
culture are predictive of the practices in that culture. Contextualization of existing leadership scales, using an inductive approach,
is necessary to ensure validity in the given culture (Farh, Cannella, & Lee, 2006). This approach requires gathering descriptions of
concrete incidents of leader behaviors and then classifying them into a number of categories by content analysis with an
agreement index constructed using multiple judges (e.g., Kerlinger, 1986; Farh, Early & Lin, 1997). This inductive approach is
particularly important in cross-cultural research where construct and measurement equivalence cannot be taken for granted (Van
de Vijver & Leung, 1997).

3.1.1. Item generating and category sorting

Two samples were used in the scale-development study. The rst sample was for generating items, and it consisted of 65 managers
participating in a management course at a university in a major city in north China. The managers came from different companies in a
wide range of industries including high-tech, service, manufacturing, etc. Fifty-one percent of them were middle-level managers and
49% were senior managers. Their average age was 37 years and their average company tenure was ten years. On a blank sheet of paper,
each respondent answered this open-ended question anonymously: what kind of leadership behavior does the CEO in your company
exhibit? The respondents could provide up to ve responses to the question. Answers were subsequently summarized and given to the
participants for class discussion.
The 65 respondents generated a total of 312 statements (approximately 4.8 items per respondent). Following Eisenhardt
(1989), Farh et al. (1997), and Tsui, Wang, and Xin (2006), a sorting procedure was employed to analyze the 312 items. Three Ph.D.
students majoring in management coded the 312 items. They had taken organizational behavior courses and were therefore
familiar with the literature on leadership. However, they did not know the purpose and hypotheses of the study.
The coding process was divided into two steps. First, based on the existing literature, broad denitions of charismatic/
transformational leadership and transactional leadership and several exemplar behaviors were provided. The three Ph.D. students
sorted the statements independently using these denitions. They were asked to create categories to group the statements that
seemed similar to each other. After sorting the rst 100 statements, they met to discuss the categories they had created. The purpose of
this sharing was to give the coders a chance to learn from one another and to make the categorization more understandable. After
sharing the coding results, each student re-sorted all 312 statements. They could create new categories if any items did not t the
existing categories. Eleven categories were obtained after the rst round of coding.
The three raters and the authors discussed these eleven categories. Categories with less than three items were discarded. The
three raters re-sorted the items based on the new category system. By the end of this second round of coding, nine categories were
retained. These categories were: 1) Being sensitive to the environment, 2) Motivating, 3) Monitoring operations, 4) Interpersonal
relationships, 5) Articulating a vision, 6) Being creative and risk-taking, 7) Being benevolent, 8) Authoritative, and 9) Being moral.
The agreement among the three raters on the two rounds of coding was computed. In the rst round, the agreement among the
three raters was 57.7% and between two raters it was 84.3%. The agreement in the second round of coding was 87.2% among the
three raters and between two raters it was 92.6%. We further examined the inter-rater agreement on each of the categories. The
agreement on the nine categories among the three coders was from 72.3% to 91.5%. The agreement between two raters was from
83.5% to 95.8%. The results suggest that the coding was reliable.

3.1.2. Scale development

A different sample of 542 managers was used for the scale development. These managers were part-time MBA students at three
universities in a major city in north China. The average age of this sample was 31 (SD = 4.43 years) and their average company tenure
was ve years (SD = 4.87 years). They had an average of six years of post-high school education and 63.8% of them were male. Ninety-
three percent of the respondents held management titles (from front line to executive level). The rest were non-management
professional-level employees (6.6%).
We selected ve most frequently mentioned statements in each of the nine categories to represent the dimensions. Forty-ve
items were used to measure CEO leadership behaviors. Each item was measured on a ve-point Likert scale, ranging from 1
(strongly disagree) to 5 (strongly agree) in response to the question of whether their CEOs exhibited the behavior described by
the item. The managers completed the survey during a class on leadership, prior to a class discussion on leadership issues. The
survey results were given to the class for a subsequent discussion of executive leadership styles in China.
Table 1 presents the results of the exploratory factor analysis with the extraction method of principal axis factoring and Oblimin
rotation with Kaiser normalization. It identies six dimensions with 24 items after a series of factor analyses. Factor loadings for all the
items were above 0.50 with no cross loadings of greater than 0.40. The six factors explained 62.35% of the total variance. The six
dimensions were labeled as Being Creative and Risk-taking (5 items), Relating and Communicating (5 items), Showing Benevolence
(4 items), Articulating a Vision (4 items), Being Authoritative (3 items), and Monitoring Operations (3 items). The alpha coefcients
ranged from 0.70 to 0.92.1

1
These six dimensions were briey reported in a different published paper (the authors names remain anonymous to preserve the principle of blind review).
The focus of that paper was leadership proles involving a cluster analysis of the six dimensions. It had a different purpose and involved other samples not used
in this paper. The only commonality was the scale development sample. Since that paper did not present details on the scale development process, in this paper,
we provide more detailed information on how these six dimensions were developed.
 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105 97

Table 1
Factor analysis results of CEO leadership behaviors.

Variables Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6

EFA CFA EFA CFA EFA CFA EFA CFA EFA CFA EFA CFA

Being creative and risk-taking

Willing to take risks 0.84 0.63 0.02 0.05 0.08 0.05 0.06
Being bold with innovation 0.83 0.71 0.02 0.04 0.01 0.06 0.05
Willing to try new projects and ideas 0.82 0.68 0.01 0.06 0.04 0.02 0.05
Having rich entrepreneurial spirit 0.78 0.57 0.03 0.07 0.06 0.02 0.05
Being creative 0.67 0.67 0.02 0.02 0.13 0.03 0.13
Relating and communicating
Having good skills in dealing with interpersonal 0.01 0.90 0.71 0.13 0.02 0.03 0.05
relationships effectively
Being able to communicate well with employees 0.10 0.79 0.66 0.00 0.04 0.05 0.03
Being good at balancing interpersonal relationships 0.01 0.77 0.70 0.03 0.04 0.01 0.04
Getting along with employees very well 0.05 0.67 0.70 0.11 0.06 0.00 0.08
Being able to facilitate interpersonal relationships 0.04 0.66 0.60 0.16 0.08 0.05 0.01
Showing benevolence
Showing concern for employee's family members 0.09 0.02 0.84 0.60 0.05 0.05 0.13
Showing concern for employee's personal life 0.01 0.00 0.81 0.64 0.01 0.02 0.13
Treating employees like family members 0.00 0.12 0.75 0.62 0.07 0.01 0.05
Showing love and care for subordinates 0.04 0.17 0.68 0.65 0.06 0.17 0.05
Articulating a vision
Clearly communicating his/her vision about the future 0.05 0.01 0.00 0.95 0.71 0.04 0.06
of the company
Articulating a bright future for employees 0.05 0.01 0.04 0.72 0.71 0.10 0.10
Clearly handling on the development of the company 0.07 0.02 0.02 0.70 0.61 0.14 0.13
in the next ve years
Emphasizing the long-term planning of the company 0.06 0.06 0.05 0.60 0.65 0.10 0.05
Being authoritative
Asking employees to obey him/her fully and 0.06 0.02 0.10 0.00 0.82 0.88 0.14
completely
Centralizing decisions on him/herself 0.01 0.10 0.02 0.03 0.69 0.62 0.10
Making unilateral decisions and taking individual 0.06 0.00 0.07 0.01 0.65 0.43 0.10
actions
Monitoring operations
Having a good control over different projects and plans 0.11 0.10 0.03 0.01 0.03 0.68 0.57
Monitoring the operation of the organization 0.00 0.04 0.09 0.03 0.04 0.56 0.56
Having strict requirements on quality of products and 0.10 0.10 0.05 0.11 0.13 0.53 0.42
services
Eigen value 8.71 2.86 1.93 1.44 1.17 1.12
Percent variance explained 34.87 10.34 6.26 4.61 3.58 2.69
Cumulative percent variance explained 34.87 45.21 51.47 56.08 59.66 62.35
Alpha coefcients 0.92 0.89 0.87 0.87 0.75 0.70

Note: EFA results came from the pilot study with 542 part-time MBA students; CFA results came from the sample in the main study with 739 middle-level
managers.

We asked 40 of these respondents to complete the survey again after two months and 30 responded. The average test-retest
reliability coefcient among the six scales over a two-month interval was 0.76 from the lowest of 0.68 to the highest of 0.85. These
results in total suggest that the measure had good reliability.
The six dimensions seem to capture both the task and relationship orientations of the CEO's behaviors. Specically, the
dimensions of Articulating a Vision, Monitoring Operations, and Being Creative and Risk-taking are related to making strategic
decisions, developing organizational structures, and monitoring and controlling the rm's production and operation systems.
These behaviors are directed at managing the task activities of the organization. The dimensions of Relating and Communicating,
Showing Benevolence, and Being Authoritative focus on the relationship aspect of a CEO's job. Through these behaviors, the CEO
has direct effects on employees' attitudes and indirect effects on rm performance. Interestingly, Showing Benevolence and
Being Authoritative are considered paternalistic leadership behaviors in the Chinese context (Cheng, 1995; Farh & Cheng, 2000;
Redding, 1990; Silin, 1976; Westwood, 1997). In general, the six dimensions that emerged from the factor analysis are consistent
with the argument that CEO leadership behaviors manifest both a relationship focus (benevolence, communicating, and
authoritative) and a task focus (vision, risk taking, and monitoring). A second-order factor analysis using these six dimensions
produced two unambiguous factors with their respective a priori dimensions.

3.2. The hypothesis testing study

3.2.1. Sample and procedures

The sample used to test our hypotheses consisted of two groups of respondents. The rst included top managers from 125
rms, including CEOs (44.8%), vice presidents or senior managers (52.4%). The second were 739 middle-level managers in these
98 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105

companies. The top managers were students in Executive MBA classes of a university in a large city in central China. These students
came from many provinces in China. They worked in different industries including telecommunications, information technology,
chemical and pharmacy, banking and investing, manufacturing, real estate, trade and commerce, and services. If the student was
not a CEO, he or she was asked to give a CEO survey to the CEO in his/her company. A self-addressed envelope was used to return
the completed survey by the CEOs.
Each student in the class was asked to invite six to eight middle managers in their companies to complete a questionnaire. The
questionnaire included items measuring CEO behaviors and the respondent's attitude variables (including organizational
commitment, perceived organizational support, and perceptions of justice). A self-addressed return envelope was provided to
each respondent to mail his/her questionnaires directly to a researcher at the university. Participation was voluntary. An average
of six middle managers from each rm returned the questionnaires, for an 83.9% response rate from the original sample.
The average age of the CEOs was 44 years (SD = 7.56 years) and their average company tenure was eight years
(SD = 6.34 years). The educational level of the CEOs was six years beyond high school (SD = 1.66 years) and 89.5% of them
were males. The average age of the middle managers was 33.7 years (SD = 7.49 years) with an average of four years of education
after high school (SD = 2.64 years). They had ve years of company tenure on average (SD = 5.81 years) and 483 of the middle
managers in the sample (65.4%) were male.

3.2.2. Measures
CEO leadership behaviors were measured using the 24-item scale resulting from the scale-development study. The middle
managers were asked to indicate on a ve-point scale the extent to which the items described the behaviors of the CEO in their
organizations. The scale ranged from 1 (strongly disagree) to 5 (strongly agree). A Conrmatory Factor Analysis (CFA) was
conducted. In the CFA, the 24 items loaded on the six a priori dimensions. Furthermore, the dimensions of Articulating a Vision,
Being Creative and Risk-taking, and Monitoring Operations loaded on one higher-order factor, namely task-focused behaviors. The
dimensions of Relating and Communicating, Showing Benevolence, and Being Authoritative loaded on another higher-order
factor, namely relationship-focused behaviors. The goodness of t indices showed that the overall Chi-square of the model was
839.45 with 245 degrees of freedom; RMSEA was 0.06; GFI was 0.91; CFI was 0.94; NNFI was 0.93; and IFI was 0.94. The coefcient
alphas for the six dimensions ranged from 0.69 to 0.89. The factor loading structure from the CFA is provided in Table 1, along with
the EFA loadings from the scale-development study.
There is a lack of systematic reliable nancial data on Chinese rms (e.g., Peng & Luo, 2000). Therefore, similar to other
researchers (Peng & Luo, 2000; Tan & Litschert, 1994; Wang, Tsui, Zhang, & Ma, 2003), we utilized perceptual measures of rm
performance. Five items from Tan and Litschert (1994) and Wang et al. (2003) were used to assess the rm's performance relative
to other rms in its industry. The items were protability, sales growth, asset growth, market share, and competitive status in the
rm's industry. The CEOs provided data on this measure based on a ve-point Likert scale from 1, very low, to 5, very high. We
also asked the CEOs to provide information on the rm's actual assets, sales, and prots. As we suspected, many executives
declined to provide this information (suggesting that the data were not available, not seen to be reliable, or not for public
consumption). We obtained this information from 69 executives. The correlation between the perceptual measures of
performance and ROA (return on assets) was 0.26 (p b 0.05) and the correlation between the same measure and ROS (return on
sales) was 0.37 (p b 0.01). In the Wang, Tsui, Zhang, and Ma (2003) study, the correlation between the perceptual performance
measure and ROA was 0.33 (p b 0.05). These correlations gave us some condence about the validity of the subjective measures.
The coefcient for the subjective rm performance measure was 0.70.
We used four measures of Employees' attitudes that focused on the organization rather than on the job or on supervisors. They
were perceived organizational support, organizational commitment, perception of the fairness of outcomes (distributive justice)
and fairness of procedures (procedural justice) in the organization. These are the most frequently studied employees' attitudes
toward their rms. These attitudes have been found directly and indirectly to inuence employees' psychological and behavioral
commitments to the rm, their task performance and their extra role behaviors. Thus, the behavioral outcomes of these attitudes
should contribute to rm performance as a whole. Petty and Bruning (1978) demonstrated that employees' attitudes accounted
for 45% of the variance in direct administrative costs across 51 public welfare organizations.
Perceived Organizational Support (POS) was measured by eight items (Eisenberger, Cummings, Armeli & Lynch, 1997). The Chinese
version of this scale is available from Chen, Aryee, and Lee (2005). Sample items include My organization really cares about my well-
being and Help is available from my organization when I have a problem. The scale's reliability coefcient was 0.87. Organizational
Commitment was measured by the scale developed by Mowday, Porter, and Steers (1982) and was used in a study in China (Chen, Tsui,
& Farh, 2002). Nine items from this scale were selected by excluding items with reverse wording. The coefcient for this measure
was 0.88. A scale developed by Farh et al. (1997), who conducted their study in Taiwan, was used to measure perceptions of Procedural
and Distributive Justice. Each construct was measured using eight items reecting different types of fairness such as pay, promotion,
and evaluation decisions. The coefcient alphas of these two measures were 0.93 and 0.88, respectively. All the items relevant to
employees' attitudes used in the present study were measured by a ve-point Likert scale with 1 for strongly disagree and 5 for
strongly agree. They were translated into Chinese and validated in the Chinese context in previous studies. CFA was used to conrm
their factor structures. The model with four factors plus one overall second-order factor structure (i.e., overall attitude) showed
reasonable t to our data, with a Chi-square value of 1380.04 with 460 degrees of freedom, RMSEA of 0.07, CFI of 0.89, NNFI of 0.88, and
IFI of 0.89.
We included several control variables that may inuence rm performance and employees' attitudes. Firm size was measured
by the total number of employees in the company. Firm age was the years since founding. Industry was measured by categorical
 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105 99

variables and eight dummy codes were used. The CEOs' age, gender, educational level, and company tenure (measured in years)
were additional control variables in this study.

3.2.3. Aggregation and preliminary analyses

Following the common approach used in unit-level research (e.g., Ostroff, 1992; Ryan et al., 1996), we aggregated individual-
level variables to the rm level. Since the data on CEO leadership behaviors and employees' attitudes were from the same source,
we separated the responses from each rm randomly into two groups in order to avoid common method bias (Ostroff, Kinicki, &
Clark, 2002). The data from the rst random half of middle managers were used to compute the CEOs' behaviors scores. The data
from the second random half were used to compute the employees' attitudes scores. Before this procedure, we computed Rwg
(James, Demaree, & Wolf, 1984) on all the variables from each rm to make sure that there was agreement among the middle
managers and that it was acceptable to aggregate individual-level variables to the rm level. We rst computed Rwg to assess
within-group agreement. We then computed ICC (intraclass correlation) (1) and ICC (2) to evaluate the variance between groups
and within groups. ICC (1) compares the between-group sum of squares to the total sum of squares. It represents the proportion of
variance in individuals' perceptions accounted for by differences in groups (James et al., 1984; Ostroff, 1992). ICC (2) is a measure
of proportional consistency of variance and can be computed by comparing the mean-square-between minus the mean-square-
within to the mean-square-between based on the results of a one-way analysis of variance (James et al., 1984; Ostroff, 1992).
Table 2 shows the agreement indices for the six leadership scales and the four attitude scales. The Rwg values ranged from 0.58 to
1.00 with a mean of 0.90 for both leadership behaviors and attitude variables. The ICC (1) for CEO leadership behavior and
attitudinal measures ranged from 0.12 to 0.40 with an average of 0.31. The ICC (2) ranged from 0.62 to 0.98 with an average of 0.79.
These results conrm a high level of agreement and reliability in the leadership and attitude ratings.
Due to the small sample size and many industry dummy codes, we performed an ANOVA to test the differences in rm performance
in different industries. The results showed no signicant differences among the eight industries (F(7114) = 0.61, p b 0.77). We also
examined the effects of other control variables on rm performance and employees' attitudes. Only the CEO's age was signicantly
correlated with Being Creative and Risk-taking and with Monitoring Operations, r = 0.21 and r = 0.21, respectively. Firm size was
negatively related to Showing Benevolence and Monitoring Operations behaviors, r = 0.18 and r = 0.19, respectively. Given the
results from the control variables, we included rm size and CEO age in the hypothesis testing but not the other control variables to
conserve statistical power. The means, standard deviations, and inter-correlation values of all the variables at the rm level are
presented in Table 3.

3.2.4. Hypothesis testing

We used structural equation modeling (Bollen, 1989) to test our two hypotheses. In the mediation model, task-focused and
relationship-focused CEO behaviors were the two latent variables. The former was manifested by the Articulating a Vision, Being
Creative and Risk-taking, and Monitoring Operations dimensions, while the latter was manifested by the Relating and
Communicating, Showing Benevolence, and Being Authoritative dimensions. Employees' attitudes were manifested by perceived
organizational support, organizational commitment, and the two fairness perceptions. Relationship-focused CEO behaviors were
found to relate positively to employees' attitudes and rm performance, and employees' attitudes were found to be associated
positively with rm performance. Task-focused CEO behaviors were found to relate directly to rm performance. We reduced the
number of items measuring rm performance by averaging the items with the highest and lowest factor loadings to form a new
indicator of rm performance. Then, the items with the next highest and the next lowest loadings were averaged to form the
second new indicator of rm performance. By repeating this process, we had three new indicators for rm performance. This
procedure of reducing the number of indicators is common in the literature on structural equation analysis in order to increase the
statistical power of the analysis (e.g., Mathieu & Farr, 1991). Lastly, we included CEO age and rm size as control variables for rm
performance. We tested the model using the LISREL 8.30 program.

Table 2
Agreement indices on six CEO leadership behaviors and four employees' attitudes.

Variables Rwg ICC (1) ICC (2)

CEO leadership behaviors

Being creative and risk-taking 0.96 0.38 0.72
Relating and communicating 0.92 0.39 0.62
Showing benevolence 0.89 0.40 0.71
Articulating a vision 0.89 0.12 0.75
Being authoritative 0.58 0.38 0.65
Monitoring operations 0.89 0.40 0.73
Employees' attitudes
Perceived organizational support 0.86 0.36 0.89
Organizational commitment 0.99 0.23 0.93
Procedural justice 1.00 0.18 0.98
Distributive justice 0.98 0.26 0.89
Median 0.90 0.38 0.74
Mean 0.90 0.31 0.79

Note: Based on 739 middle managers in 125 rms, an average of 5.91 managers per rm.
 100
Table 3
Mean, SD, and inter-correlations of CEO leadership behaviors, employees' attitudes, and rm performance (N = 125).

Variables Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Outcome variables

H. Wang et al. / The Leadership Quarterly 22 (2011) 92105

1. Firm performance 3.63 0.54 (0.70)
Employees' attitudes
2. Perceived organizational support 3.53 0.44 0.18 (0.87)
3. Organizational commitment 3.80 0.45 0.31 0.78 (0.88)
4. Procedural justice 3.41 0.48 0.20 0.80 0.79 (0.93)
5. Distributive justice 3.56 0.42 0.23 0.70 0.68 0.87 (0.88)
CEO leadership behaviors
6. Being creative and risk-taking 3.93 0.49 0.21 0.29 0.28 0.26 0.21 (0.89)
7. Relating and communicating 3.87 0.49 0.13 0.44 0.43 0.46 0.44 0.55 (0.89)
8. Showing benevolence 3.56 0.46 0.11 0.50 0.43 0.46 0.45 0.57 0.72 (0.87)
9. Articulating a vision 3.94 0.46 0.34 0.35 0.39 0.30 0.32 0.64 0.60 0.63 (0.87)
10. Being authoritative 3.13 0.47 0.13 0.31 0.25 0.28 0.27 0.01 0.32 0.26 0.21 (0.69)
11. Monitoring operations 3.97 0.44 0.26 0.44 0.38 0.40 0.40 0.62 0.69 0.60 0.68 0.18 (0.70)
Control variables
12. CEO age 43.52 7.56 0.06 0.05 0.13 0.09 0.10 0.21 0.10 0.02 0.10 0.12 0.21
13. CEO gender 1.94 0.24 0.03 0.09 0.03 0.08 0.15 08 0.10 0.11 0.07 0.08 0.02 0.03
14. CEO education 2.24 1.67 0.09 0.00 0.06 0.01 0.03 0.12 0.10 0.04 0.12 0.07 0.10 0.04 0.14
15. CEO tenure 8.23 6.42 0.13 0.12 0.18 0.13 0.01 0.17 0.11 0.13 0.13 0.09 0.13 0.16 0.02 0.19
16. Firm age 15.79 20.96 0.07 0.05 0.08 0.15 0.08 0.01 0.19 0.13 0.07 0.03 0.14 0.15 0.01 0.10 0.50
17. Firm size 2.78 0.83 0.12 0.18 0.12 0.13 0.06 0.11 0.14 0.18 0.10 0.13 0.19 0.26 0.09 0.11 0.36 0.29

Note: Firm size is measured with logarithm values of employee number in the rm.
p b 0.05.
p b 0.01.
 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105 101

4. Results

Fig. 1 shows the estimated path coefcients related to the direct effects of CEOs' behaviors and the mediating effects of
employees' attitudes on rm performance. The overall Chi-square of the model was 145.76 with 87 degrees of freedom (p b 0.01);
RMSEA was 0.08; IFI was 0.95; CFI was 0.95; and NNFI was 0.93. These results and the signicance of the path coefcients indicate
that there was a good t between the data and the hypothesized mediating model.
We also compared the hypothesized model with two nested alternative models. In the rst model, we specied a path from
task-oriented behavior to employees' attitudes. The overall Chi-square of the alternative model was 143.12 with 86 degrees of
freedom (p b 0.01); RMSEA was 0.08; IFI was 0.95; CFI was 0.95; and NNFI was 0.93. The chi-square difference was 2.40 with
1 degree of freedom, which was not signicant. The nding suggests that adding the path from task-oriented behavior to
employees' attitudes does not signicantly improve the hypothesized model. In the second alternative model, we removed the
path from relationship-oriented leadership behaviors to rm performance. The overall Chi-square of the alternative model was
150.53 with 88 degrees of freedom (p b 0.01); RMSEA was 0.08; IFI was 0.94; CFI was 0.94; and NNFI was 0.92. The difference in
chi-square, 4.77 with 1 degree of freedom, was signicant. This result suggests that the hypothesized model was superior to the
alternative model.
The overall model showed that rm performance was explained well by aggregated middle managers' attitudes and task-focused
CEOs' behaviors (R2 = 0.28). Specically, organizations whose middle managers expressed more positive attitudes towards their
organizations had higher levels of rm performance ( = 0.36, p b 0.05). Task-focused CEO leadership behaviors were also directly
related to rm performance ( = 0.53, p b 0.05). Employees' attitudes were strongly inuenced by relationship-focused CEO behaviors
( = 0.45, p b 0.01). The relationship-focused CEO behaviors had no direct effects on rm performance ( = 0.32, n.s.). The control
variables of rm size and CEO age were not signicant.

5. Discussion

Building on strategic leadership theories in the Western literature, we inductively identied two sets of CEO leadership
behaviors in the Chinese context. The six-dimension measure of CEO leadership behavior has some Chinese characteristics at the
behavioral level but also reects the dual focus on the relationship and task domains of an executive's job. Results showed that
task-focused CEO leadership behaviors are directly linked to rm performance. Relationship-focused behaviors relate directly to

Perceived Org.
Support
Relating and
Communicating .39**
.47** Organizational
.37** Commitment

Showing
Relationship-Focused .45** Employees
.44** CEO Leadership
Benevolence Attitudes .37**
Behaviors
-.27** Distributive
Justice
.46**
Being
Authoritative
Procedural
-.32 n.s. .36* Justice

Articulating a
Vision Indicator 1
.45** .42**

Being Creative .34** Task-Focused .53* Firm .33**

and Risk-tasking CEO Leadership Indicator 2
Behaviors
Performance
.28**
.50**
.10 n.s.
Monitoring .03 n. s. Indicator 3
Operations
Firm size CEO age
* p<.05; **p<.01

Fig. 1. The results of the mediating effect of employees' attitudes on CEO leadership behaviors and rm performance.
102 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105

employees' attitudes and indirectly to rm performance. These results conrm our argument that CEOs can achieve higher levels
of rm performance by both focusing on task performance directly and inducing positive attitudes of employees through their
relationship-oriented behaviors.
The existing research on CEO characteristics and organizational effectiveness is tting of the metaphor of the blind men and
the elephant with each of the studies only touching on one CEO characteristic. For example, Hambrick et al. (1996) focused on
demographic variables; Waldman et al. (2001) on charismatic leadership behaviors; and Peterson et al. (2003) on the
personalities of top managers. Our study adds another piece to this larger puzzle regarding the role of the CEO by contributing to
the behavioral theory of CEO leadership. The dimension of Articulating a Vision is critical for the CEO to instill aspiration, energy,
and intensity in pursuing organizational goals. The Being Creativity and Risk-taking and Monitoring Operations dimensions are
relevant during the process of operating and managing the organization's tasks. The behaviors of Relating and Communicating
and Showing Benevolence, while echoing the importance of relationships in China (Yang, Yu, & Yeh, 1989), may not be unique to
the Chinese context. They may be relevant across contexts. Future research should explore or conrm the relevance of these
dimensions in other social-cultural settings. Even the dimension of Being Authoritative may not be a culture-specic Chinese
behavior. For example, abusive supervision, a manifestation of dysfunctional workplace behavior, has attracted the attention of
academic researchers in China (Aryee, Sun, Chen, & Debrah, 2008) and the U.S. (Tepper, 2000). Studies could compare the
prevalence and inuence of such leadership behavior in different cultures. Future research could explore the specic function of
this dimension on rm performance and employees' attitudes since it has negative and signicant correlations with other
relationship-oriented behaviors as well as with the four employees' attitudes variables.

5.1. Theoretical implications

The present ndings extend current theories by extending to the behaviors of top managers and demonstrating that the
behaviors of the CEO not only inuence rm performance, but also the middle managers' attitudinal responses. These results offer
a glimpse into the black box of leadership behaviors (e.g., Carpenter et al., 2004) and provide conrmatory evidence in a different
national context that organizations that have more satised employees are more productive and protable than organizations
whose employees are less satised (Ostroff, 1992: p.963).
The current study focused on the leadership behaviors of the chief operating ofcer. Future research could also study the
behavior of the entire top management team. The top management team may have behavioral tendencies that mirror the
behaviors of the CEO if the team is characterized by demographic homogeneity. However, the top management team members
also may not share all the behavioral characteristics of the CEO, especially in teams with demographic and personality
heterogeneity. Future research should include both demographic and behavioral characteristics of both the CEO and the top
management team for a more comprehensive assessment of executive leadership inuences on the organization. The concept of
team behavior at the executive level may be worthy of theoretical and empirical attention.
The employees surveyed in this study were middle managers who would more likely have access to and interaction with the
CEO than would lower-level employees. This access provides them with opportunities to observe the behaviors of the CEO. This
also means that they would be more directly inuenced by the behaviors of the executive leader. Would the mediating role of
attitudes be generalizable to employees at lower organizational levels? It seems plausible that favorable attitudes, such as
perceptions of organizational support, justice, and commitment, of employees at all levels contribute to positive rm outcomes.
However, would such attitudes at lower organizational levels be inuenced by the CEO, by middle managers, or by their
immediate superiors? Further research should examine the effect of the CEO relative to middle managers and the immediate
superior on lower-level employees. Would the CEO's leadership behavior cascade down the hierarchy such that their inuence on
lower-level employees is through the leadership behaviors of managers below them?
This study was conducted in a context where managers are likely to have great latitude for discretion due to both institutional
and cultural reasons. Could societal culture play a moderating role between CEO behaviors and employees' attitudes? Future
research could explore the joint inuence that national culture and the environment have on executive leadership behaviors and
rm outcomes.

5.2. Limitations

The study has several limitations. The rst is the use of a cross-sectional design. Although it is argued that CEO leadership behaviors
have effects on employees' attitudes and on rm performance, one can still make the argument that high levels of rm performance
lead to positive employees' attitudes, or that high levels of rm performance lead to a positive evaluation of CEO leadership behaviors
by the top managers. Such attribution is clearly possible. A longitudinal research design is needed to conrm the causal relationship
between CEO leadership behaviors, employees' attitudes, and rm performance. However, by using multiple sources of data (top
managers and middle managers) and splitting the sample of middle managers into two random sub-samples, we have at least ruled
out common method variance as a possible source of bias.
The second limitation of the study is the measure of rm performance. Though the use of a subjective measure of performance
is not ideal, this is a pragmatic constraint in conducting research in China at this stage due to the lack of reliable performance data
and the variety of rms and industries in our sample, making it difcult (if not impossible) to have objective performance
indicators that are meaningful and valid for all rms. Further, a self-serving bias by the CEOs might have reduced the variance in
performance across the rms. Future research could include other indicators of organizational effectiveness to assess the impact of
 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105 103

the CEO. For example, the CEO deals with multiple stakeholders, such as employees, customers, suppliers, and the community in
which the organization is located. Assessment by these stakeholders might provide alternative measures of CEO effectiveness,
recognizing that a rm's nancial performance is but one aspect of an organization's overall effectiveness.

5.3. Practical implications

In addition to the theoretical implications, the current study makes practical contributions. The results suggest that both task-
oriented and relationship-oriented leadership behaviors are important for the CEO. Ideally, a CEO should be able to demonstrate
both types of behaviors. Such behavioral exibility is important to strategic leadership according to Boal and Hooijberg (2000). If a
CEO is inclined toward only one set of behaviors or is singular in focus, the missing behaviors should be provided by another
executive in the top management team to ensure the rm's optimal performance. Given the young age of these Chinese CEOs
(average age of 43), executive education that aims at developing the full range of leadership skills of CEOs might be a highly
advisable course of action for Chinese companies.

6. Conclusion

This study makes several contributions to the literature. First, this study provided evidence that CEO leadership could inuence
organizational performance through inducing positive attitudinal responses of the company's middle managers. Second, the
mediating role of employees' attitudes enriches the content of upper echelon theory by revealing one of the mechanisms through
which executive leaders inuence rm performance. Third, this study offers some culture-specic insight about executive
leadership. Lastly, this study suggests some avenues for further exploration of the complex relationship between leadership
behavior, employees' responses, and rm outcomes. Despite almost 100 years of research and many thousands studies on
leadership (Bass, 1990) and a rejuvenation of interest in executive leadership in the past 20 years (Boal & Hooijberg, 2000), the
eld does not yet have denitive knowledge about the role of the CEO in the rm. Much remains to be explored on how, when,
where and why CEOs inuence rms. We hope that the current study adds another small piece to this large puzzle.

Acknowledgments

This research is supported by a grant from the Hong Kong Research Grants Council (RGC number is HKUST6218/00H), a China
Business and Management Area of Excellence grant from the Hong Kong University of Science and Technology, and grants from
the Natural Science Foundation of China (NSFC number is 70672010 and 71032001).

References

Agle, B. R., Nagarajan, N. J., Sonnenfeld, J. A., & Srinivasan, D. (2006). Does CEO charisma matter? An empirical analysis of the relationship among organizational
performance, environmental uncertainty, and top management team perceptions of CEO charisma. Academy of Management Journal, 49, 161174.
Aryee, S., Sun, L., Chen, Z. X., & Debrah, Y. A. (2008). Abusive supervision and contextual performance: The mediating role of emotional exhaustion and the
moderating role of work unit structure. Management and Organization Review, 4, 393411.
Avolio, B. J., & Bass, B. M. (1987). Charisma and beyond. In J. G. Hunt (Ed.), Emerging leadership vistas. Elmasford, New York: Pergamon Press.
Bass, B. (1990). Bass and Stogdill's handbook of leadership: Theory, research, and managerial applications. New York: Free Press.
Boal, K. (2004). Strategic leadership. In G. R. Goethals, G. J. Sorenson, & J. M. Burns (Eds.), Encyclopedia of leadership. CA: Sage, Thousand Oaks.
Boal, K., & Hooijberg, R. (2000). Strategic leadership research: Moving on. Leadership Quarterly, 11, 515549.
Bollen, K. A. (1989). Structural equation with latent variables. New York: Wiley.
Bryman, A. (1992). Charisma and leadership in organizations. Newbury Park, CA: Sage.
Burns, J. M. (1978). Leadership. New York: Harper & Row.
Cannella, A. A., Jr., & Monroe, M. J. (1997). Contrasting perspectives on strategic leaders: Toward a more realistic view of top managers. Journal of Management, 23,
213238.
Carpenter, M. A., & Fredrickson, J. W. (2001). Top management teams, global strategic posture, and the moderating role of uncertainty. Academy of Management
Journal, 44, 533545.
Carpenter, M. A., Geletkanycz, M. A., & Sanders, W. G. (2004). Upper echelons research revisited: Antecedents elements, and consequences of top management
team composition. Journal of Management, 30, 749778.
Chen, Z. X., Aryee, S., & Lee, C. K. (2005). Test of a mediation model of perceived organizational support. Journal of Vocational Behavior, 66, 457470.
Chen, X., & Farh, J. L. (1999). The effectiveness of transactional and transformational leader behaviors in Chinese organizations: Evidence from Taiwan. AOM
Conference paper, 1999, Chicago.
Chen, Z. X., Tsui, A. S., & Farh, J. L. (2002). Loyalty to supervisor versus organizational commitment: Relationship to employee performance in China. Journal of
Occupational and Organizational Psychology, 75, 339356.
Cheng, B. S. (1995). Authoritarian values and executive leadership: The case of Taiwanese family enterprises: Report prepared for Taiwan's National Science Council.
Taiwan: National Taiwan University (In Chinese).
Colbert, A. E., Kristof-Brown, A. L., Bradley, B. H., & Barrick, M. R. (2008). CEO transformational leadership: The role of goal importance congruence in top
management teams. Academy of Management Journal, 51, 8196.
Conger, J. A., & Kanungo, R. N. (1987). Toward a behavioral theory of charismatic leadership in organizational settings. Academy of Management Review, 12,
637647.
Denison, D. R., Hooijerg, R., & Quinn, R. E. (1995). Paradox and performance: A theory of behavioral complexity in managerial leadership. Organization Science, 6,
524540.
Dreyfus, H., Dreyfus, S., & Athanasion, T. (1986). Mind over machine: The power of human intuition and expertise in the era of the computer. New York: Free Press.
Eisenberger, R., Cummings, J., Armeli, S., & Lynch, P. (1997). Perceived organizational support, discretionary treatment, and job satisfaction. The Journal of Applied
Psychology, 82, 812820.
Eisenhardt, K. (1989). Building theories from case study research. Academy of Management Review, 14, 532550.
104 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105

Farh, J. L., Cannella, A. A., Jr., & Lee, C. (2006). Approaches to scale development in Chinese management research. Management and Organization Review, 2,
301318.
Farh, J. L., & Cheng, B. S. (2000). A cultural analysis of paternalistic leadership in Chinese organizations. In J. T. Li, A. S. Tsui, & E. Weldon (Eds.), Management and
organizations in the Chinese context. London: Macmillan Press Ltd.
Farh, J. L., Early, P. C., & Lin, S. C. (1997). Impetus for action: A culture analysis justice and organizational citizenship behavior in Chinese society. Administrative
Science Quarterly, 42, 421444.
Finkelstein, S., & Hambrick, D. C. (1996). Strategic leadership: Top executives and their effects on organizations. St. Paul, MN: West Publishing Company.
Hambrick, D. C., Cho, T. S., & Chen, M. J. (1996). The inuence of top management team heterogeneity on rm's competitive moves. Administrative Science Quarterly,
41, 659684.
Hambrick, D. C., & Finkelstein, S. (1987). Managerial discretion: A bridge between polar views of organizational fates. In B. Staw & L.L. Cummings (Eds.), Research in
organizational research, 9. (pp. 369406)Greenwich, CT: JAI Press.
Hambrick, D. C., & Mason, P. A. (1984). Upper echelons: The organization as a reection of its top managers. Academy of Management Review, 9, 193206.
Hart, S. L., & Quinn, R. E. (1993). Roles executives play: CEOs, behavioral complexity, and rm performance. Human Relations, 46, 543574.
Harter, J. K., Schmidt, F. L., & Hayes, T. L. (2002). Business-unit level relationship between employee satisfaction, employee engagement, and business outcomes: A
meta-analysis. The Journal of Applied Psychology, 87, 268279.
Hooijberg, R., & Quinn, R. E. (1992). Behavioral complexity and the development of effective managers. In R. L. Phillips & J.G. Hunt (Eds.), Strategic leadership: A
multiorganizational-level perspective. London: Quorum Books.
House, R. J., Hanges, P. J., Ruiz-Quintanilla, S. A., Dorfman, P. W., Javidan, M., Dickson, M., et al. (1999). Cultural inuences on leadership and organizations. Advances
in global leadership, Vol. 1. (pp. 171233)Stanford, CT: JAI Press.
House, R. J., Spangler, W. D., & Woycke, J. (1991). Personality and charisma in the U.S. presidency: A psychological theory of leader effectiveness. Academy of
Management Journal, 36, 364396.
Hunt, J. G. (1991). Leadership: A new synthesis. Newbury Park, Calif: Sage Publication.
James, F. R., Demaree, R. G., & Wolf, G. (1984). Estimating within-group interrater reliability with and without response bias. The Journal of Applied Psychology, 68,
8598.
Judge, T. A., Piccolo, R. F., & Illies, R. (2004). The forgotten ones? The validity of consideration and initiating structure in leadership research. The Journal of Applied
Psychology, 89, 3651.
Katz, D., & Kahn, R. L. (1978). The social psychology of organizations. New York: John Wiley & Sons Ltd.
Kerlinger, F. N. (1986). Foundations of behavioral research. Fort Worth: Holt: Rinehart and Winston.
Koys, D. J. (2001). The effects of employee satisfaction, organizational citizenship behavior, and turnover on organizational effectiveness: A unit-level, longitudinal
study. Personnel Psychology, 54, 101114.
Lieberson, S., & O'Connor, J. F. (1972). Leadership and organizational performance: A study of large corporate. American Sociological Review, 37, 117130.
Ling, Y., Simsek, Z., Lubatkin, M. H., & Veiga, J. F. (2008a). Transformational leadership's role in promoting corporate entrepreneurship: Examining the CEO-TMT
interface. Academy of Management Journal, 51, 557576.
Ling, Y., Simsek, Z., Lubatkin, M. H., & Veiga, J. F. (2008b). The impact of transformational CEOs on the performance of small- to medium-sized rms: Does
organizational context matter? The Journal of Applied Psychology, 93, 923934.
Mathieu, J. E., & Farr, J. L. (1991). Further evidence for the discriminant validity of measures of organizational commitment, job involvement, and job satisfaction.
The Journal of Applied Psychology, 76, 127133.
Meindl, J. R., Ehrlich, S. B., & Dukerich, J. M. (1985). The romance of leadership. Administrative Science Quarterly, 30, 78102.
Mowday, R. T., Porter, L. W., & Steer, R. M. (1982). Employee-organization linkages: The psychology of commitment, absenteeism, and turnover. New York: Academic
Press.
Ostroff, C. (1992). The relationship between satisfaction, attitudes, and performance: An organizational level analysis. The Journal of Applied Psychology, 77,
963974.
Ostroff, C., Kinicki, A. J., & Clark, M. A. (2002). Substantive and operational issues of response bias across levels of analysis: An example of climate-satisfaction
relationship. The Journal of Applied Psychology, 87, 355368.
Peng, M. M., & Luo, Y. (2000). Managerial ties and rm performance in a transition economy: The nature of a micromacro link. Academy of Management Journal, 43,
486501.
Petty, M. M., & Bruning, N. S. (1978). Leader behavior and subordinate attitudes as predictors of leader effectiveness and agency inefciency in public welfare
agencies. Proceedings of the Academy of Management, 1978, 357360.
Peterson, R. S., Smith, D. B., Martorana, P. V., & Owens, P. D. (2003). The impact of chief executive ofcer personality on top management team dynamics: One
mechanism by which leadership affects organizational performance. The Journal of Applied Psychology, 88, 795808.
Pfeffer, J. (1977). The ambiguity of leadership. Administrative Science Review, 2, 104112.
Podsakoff, P. M., MacKenzie, S. N., Moorman, R. H., & Fetter, R. (1990). Transformational leader behaviors and their effects on followers' trust in leader, satisfaction,
and organizational citizenship behavior. Leadership Quarterly, 1, 107142.
Podsakoff, P. M., MacKenzie, S. N., & Bommer, W. H. (1996). Transformational leader behaviors and substitutes for leadership as determinants of employee
satisfaction, commitment, trust, and organizational citizenship behavior. Journal of Management, 22, 259298.
Quinn, J. B. (1988). Beyond rational management: Mastering the paradoxes and competing demands of high performance. San Francisco: Jossey-Bass.
Quinn, J. B., & Cameron, K. (1988). Paradox and transformation. Cambridge, MA: Ballinger.
Redding, S. G. (1990). The spirit of Chinese capitalism. New York: de Gruyter.
Ryan, A. M., Schmit, M. J., & Johnson, R. (1996). Attitudes and effectiveness: Examining relations at an organizational level. Personnel Psychology, 49, 853882.
Schilit, W. K. (1987). An examination of the inuence of middle-level managers in formulating and implementing strategic decisions. Journal of Management
Studies, 24, 271293.
Schleicher, D. J., Watt, J. D., & Greguras, G. J. (2004). Reexamining the job satisfaction-performance relationship: The complexity of attitudes. The Journal of Applied
Psychology, 89, 165177.
Schneider, B., Hanges, P. J., Smith, D. B., & Salvaggio, A. N. (2003). Which comes rst: Employee attitudes or organizational nancial and market performance. The
Journal of Applied Psychology, 88, 836851.
Shamir, B., House, R. J., & Arthur, M. B. (1993). The motivational effects of charismatic leadership: A self concept based theory. Organization Science, 4, 117.
Silin, R. H. (1976). Leadership and value: The organization of large-scale Taiwan enterprises. Cambridge, MA: Harvard University Press.
Smith, K. G., Smith, K. A., Olian, H. P., Sims, H. P., O'Bannon, D. P., & Scully, J. A. (1994). Top management team demography and process: The role of social
integration and communication. Administrative Science Quarterly, 39, 412438.
Stogdill, R. M. (1974). Handbook of leadership. New York: Free Press.
Streufert, S., & Swezey, R. W. (1978). Complexity, managers, and organizations. Orlando, FL: Academic Press.
Tan, J. J., & Litschert, R. J. (1994). Environment-strategy relationship and its performance implications: An empirical study of the Chinese electronic industry.
Strategic Management Journal, 15, 120.
Tepper, B. J. (2000). Consequences of abusive supervision. Academy of Management Journal, 43, 178191.
Thomas, A. B. (1988). Does leadership make a difference to organizational performance? Administrative Science Quarterly, 33, 388400.
Tosi, H. L., Misangyi, V. F., Fanelli, A., Waldman, D. A., & Yammarino, F. J. (2004). CEO charisma, compensation, and rm performance. Leadership Quarterly, 15,
405421.
Tsui, A. S., & Lau, C. M. (2002). Management of enterprises in the People's Republic of China. Boston: Kluwer Academic Press.
Tsui, A. S., Schoonhoven, C. B., Meyer, M. W., Lau, C., & Milkovich, G. T. (2004). Organization and management in the midst of societal transformation: The People's
Republic of China. Organization Science, 15, 133144.
 H. Wang et al. / The Leadership Quarterly 22 (2011) 92105 105

Tsui, A. S., Wang, H., & Xin, R. (2006). Organizational culture in the People's Republic of China: An analysis of culture dimensions and culture types. Management
and Organization Review, 2, 345376.
Van de Vijver, F., & Leung, K. (1997). Methods and data analysis for cross-cultural research. London: Sage.
Waldman, D. A., Javidan, M., & Varella, P. (2004). Charismatic leadership at the strategic level: A new application of upper echelons theory. Leadership Quarterly, 15,
355381.
Waldman, D. A., Ramirez, G. G., House, R. J., & Puranam, P. (2001). Does leadership matter? CEO leadership attributes and protability under condition of perceived
environmental uncertainty. Academy of Management Journal, 44, 134143.
Wang, D., Tsui, A. S., Zhang, Y., & Ma, L. (2003). Employment relationships and rm performance: Evidence from an emerging economy. Journal of Organizational
Behavior, 24, 511535.
Westwood, R. I. (1997). Harmony and patriarchy: The cultural basis for paternalistic headship among the overseas Chinese. Organization Studies, 18, 445480.
Yang, K. S. (1993). Chinese social orientation: An integrative analysis. In L. Y. Cheng, F. M. C. Cheung, & C. N. Cheng (Eds.), Psychotherapy for the Chinese
(pp. 1956). Hong Kong: The Chinese University of Hong Kong.
Yang, K. S., Yu, A. B., & Yeh, M. H. (1989). Chinese individual modernity and traditionality: construct denition and measurement. Proceedings of the interdisciplinary
conference on Chinese psychology and behavior (pp. 287354). (in Chinese).
Yukl, G. A. (1989). Managerial leadership: A review of theory and research. Yearly Review of Management, 15, 251289.
Yukl, G. A. (2002). Leadership in organizations (5th Ed.). Upper Saddle River, New Jersey: Prentice Hall.