Description
During this such a detailed information in regard to in pursuit of greatness ceo narcissism, entrepreneurial orientation, and firm performance.
1
This article is protected by copyright. All rights reserved.
In Pursuit of Greatness: CEO Narcissism, Entrepreneurial Orientation, and Firm Performance Variance
1
WILLIAM J. WALES
2
James Madison University
Department of Management
800 S. Main Street
Harrisonburg, VA 22801
Tel: (540) 568-4375
Fax: (540) 568.3275
e-mail: [email protected]
PANKAJ C. PATEL
Miller College of Business
Ball State University
Muncie, IN 47306
Tel: 765-285-3194
e-mail: [email protected]
and
This article has been accepted for publication and undergone full peer review but has not been through the
copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version
and the Version of Record. Please cite this article as doi: 10.1111/joms.12034
2
Corresponding author
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
2
This article is protected by copyright. All rights reserved.
G. THOMAS LUMPKIN
Syracuse University
Whitman School of Management
721 University Avenue
Syracuse, NY 13244-2450
Tel: (315) 443-3164
e-mail: [email protected]
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
3
This article is protected by copyright. All rights reserved.
IN PURSUIT OF GREATNESS: CEO NARCISSISM, ENTREPRENEURIAL
ORIENTATION, AND FIRM PERFORMANCE VARIANCE
ABSTRACT
Building upon the perspective that narcissism is a leadership trait with both ‘bright’ and ‘dark’
sides, the present study examines the question of whether companies led by narcissistic CEOs
exhibit higher levels of entrepreneurial orientation (EO). Moreover, this research examines
whether EO partially explains why narcissistic CEO led firms experience greater variability in
firm performance. Using survey data collected from 173 CEOs, and an archival measure of firm
performance variance, we find support for our model. These findings offer an improved
understanding of how CEO narcissism influences performance variance, and why the firms they
lead may even, at times, be viewed as on a path to success. Study implications are discussed.
Keywords:
CEO Leadership Traits; Narcissism; Strategic Posture; Entrepreneurial Orientation; Firm
Performance Variance
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
4
This article is protected by copyright. All rights reserved.
INTRODUCTION
Narcissism is broadly defined as an exaggerated, yet fragile self-concept of one’s
importance and influence (Resick et al., 2009). Most of the work on narcissism to date has
explored the negative implications of the phenomenon for individual-level functioning and
leadership (Judge et al., 2006; Kets De Vries and Miller, 1985). However, prior research also
indicates that many chief executive officers (CEOs) who are perceived as narcissistic lead highly
successful companies (Maccoby, 2000). How might this be so?
To begin, Chatterjee and Hambrick (2007) observe that narcissistic CEOs are associated
with more extreme firm performance—both bigger gains and bigger losses. While increased
variability in firm performance has neither an inherently positive or negative organizational
implication, higher variability can be interpreted as beneficial if the achievement of a superior
competitive position—i.e., above-average performance—is preferred over maintenance of the
firm’s extant competitive position which lower variance would help protect (March, 1991). Yet,
CEO narcissism itself is insufficient to fully account for such firm performance extremes in the
pursuit of superior performance. This raises the research question: What firm-level mechanisms
might narcissistic CEOs employ to obtain such wide variations in organizational outcomes?
Given that entrepreneurial processes and practices are often associated with stimulating
company renewal and growth (Guth and Ginsberg, 1990; Hoskisson et al., 2011), CEOs often
promote a firm-level entrepreneurial orientation (EO) within the organizations they lead (Rauch
et al., 2009). As will be discussed, firm-level entrepreneurial activities have also been identified
as an explanatory factor of firm performance variance, that is, relatively stronger and relatively
weaker firm performance (Wiklund and Shepherd, 2011). The present study proposes a linkage
between CEO narcissism and firm-level EO, and specifically asks whether EO might be a key
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
5
This article is protected by copyright. All rights reserved.
firm-level mechanism that enables narcissistic CEOs to lead organizations that experience a wide
range of performance outcomes. Chatterjee and Hambrick (2007) demonstrate that CEO
narcissism can affect firm strategy-making and performance variance independently. The present
study seeks to extend their findings by asking whether EO partially mediates the relationship
between CEO narcissism and firm performance variance.
By investigating these questions, this research makes three principal contributions to the
management literature. First, by further clarifying the relationship between CEO narcissism and
firm performance variance, we address the call by Rosenthal and Pittinsky (2006) to progress
beyond simple discussions of narcissistic leadership as inherently ‘good’ or ‘bad’, and focus
instead on its implications for company outcomes. In this vein we also build on the emerging
perspective that there are both “bright” and “dark” sides to narcissistic CEO leadership (Resick
et al., 2009), and thus it may have both positive and negative effects on their companies fate.
Second, our study extends the work of Chatterjee and Hambrick (2007) by identifying
EO as a strategy making construct that is not only affected by CEO narcissism, but also partially
mediates the CEO narcissism—performance variance relationship. Given the important and
sometimes dramatic role that entrepreneurial processes and practices may play in changing the
fate or direction of a company (Hoskisson et al., 2011), we posit that the tendency of narcissistic
CEOs to endorse firm-level EO helps explain why their organizations experience an increase in
performance variance, and may even, at times, be viewed as on a path to success. Third, while
narcissism has become a broad managerial phenomenon of interest (Bergman et al., 2010), little
consideration has been given to empirically examining this phenomenon using psychometrically
valid scales within broad samples of executives (Bollaert and Petit, 2010). Our study provides a
robust test of our theorized model using both survey data and archival performance measures.
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
6
This article is protected by copyright. All rights reserved.
The remainder of the paper is organized as follows. First we review narcissism as a
leadership trait. We then develop a research model and posit hypotheses linking CEO narcissism,
EO, and firm performance variance. This is followed by a discussion of methods and a report of
findings. Finally, we conclude with a discussion of implications and future research directions.
THEORETICAL BACKGROUND AND HYPOTHESIS DEVELOPMENT
Narcissism—an inflated but fragile view of one’s self-importance—is a construct that has
generated substantial interest across the social sciences for many decades (e.g., Morf and
Rhodewalt, 2001). The negative consequences of narcissism have been well documented in the
literature (e.g., Resick et al., 2009). Narcissistic individuals are principally concerned with
actions that reflect favorably upon themselves, which often leads to perceptions of entitlement,
shameless self-admiration, excessive arrogance, and hostility towards external criticism (Judge et
al., 2006; Lubit, 2002; Resick et al., 2009). Thus, in interpersonal relationships, they are often
viewed as abrasive and insensitive to their impact on others (Judge et al., 2009). Despite the
interpersonal issues typically associated with narcissism, maintaining unrealistic, inflated levels
of self-esteem creates a continuous need for narcissists to obtain external self-affirmation
through social interaction. One place narcissists receive such affirmation is in leadership roles.
A key reason narcissists are attracted to organizational leadership roles is their desire to
leave behind a grand, admirable legacy of achievement (Maccoby, 2000). Organizational settings
can provide narcissistic leaders with positions of authority and influence that can facilitate
achievement of their personal ambitions. In general, the role of CEO provides narcissists with the
greatest authority and hence studies in the management literature have previously focused on
examining narcissism among CEOs (e.g., Chatterjee and Hambrick, 2007; Resick et al., 2009).
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
7
This article is protected by copyright. All rights reserved.
An analysis of narcissism and leadership by Kets De Vries and Miller (1985) was among
the first to appear in the management literature. They noted that narcissism, while typically
viewed as detrimental, can also be conceived as beneficial for firm leaders and linked to higher
productivity. Subsequent analyses tend to support this view. Describing the tension between the
positive and negative elements of narcissistic leadership, Judge and colleagues (2009) classified
narcissism as a ‘dark’ leadership trait with some potentially ‘bright’ sides. In this regard,
Rosenthal and Pittinsky (2006) note that while it may at first seem ‘preposterous’ to suggest that
narcissists could possess strong leadership characteristics with both negative and positive
implications for organizational performance, it is certainly possible even if it does occur as a
result of self-aggrandizing motivations. For instance, it has been argued that narcissists, who are
generally viewed as arrogant, self-important and insensitive, can also be considered tough-
minded visionaries and evangelistic leaders (Maccoby, 2004). In this regard, some of the
positive attributes of narcissism operate similarly to those of charisma
1
. Both may inspire
devotion and commitment in others to achieve higher performance. This semblance may help
explain why individuals are sometimes attracted to narcissistic leaders and may even view their
leadership as transformational (Rosenthal and Pittinsky, 2006).
While extreme narcissism is a diagnosable personality disorder
2
, social science research
generally views it as a matter of degree, ranging from mild to severe (Foster and Campbell,
2007; Kets De Vries and Miller, 1985). Consistent with current research in social psychology
and strategic management, we view narcissism among executives as not necessarily reflective of
a personality disorder which presents itself as a binary state (i.e., sufferers v. non- sufferers).
Rather, we view narcissism as a personality trait which manifests in varying degrees along a
continuum. By adopting this view we recognize that different levels of narcissism among CEOs
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
8
This article is protected by copyright. All rights reserved.
can influence a range of organizational strategic behaviors. But the force of CEO personality
alone is insufficient to fully explain changes in firm performance. For CEO attributes and
ambitions to influence organizational outcomes, they must be translated into specific firm-level
strategic behaviors. In the following sub-section, we investigate the implications of CEO
narcissism for the manifestation of a particular firm-level strategic orientation, EO.
CEO Narcissism and Entrepreneurial Orientation
When CEOs possess higher levels of narcissism they are more likely to strive for bold,
daring actions which increase their stature by drawing attention to their vision and leadership
(Judge et al., 2009). Further, narcissists are driven by an overwhelming desire to compete
(Maccoby, 2004; Rosenthal and Pittinsky, 2006). To fulfill these needs, narcissistic CEOs will
likely direct their organizations to take bolder and more aggressive strategic actions, which
would be characteristic of an EO (Covin and Slevin, 1991; Lumpkin and Dess, 1996). We
propose that by inspiring their firms to assume a more entrepreneurial, competitively focused
strategic posture (Rauch et al., 2009), narcissistic CEOs can enhance their personal prestige and
fulfill their desire for positive accolades through the attributions they receive for their role in the
entrepreneurial initiatives their firm undertakes.
Entrepreneurial orientation (EO) refers to the strategy-making policies and practices that
firms use to identify and launch new ventures (Miller, 2011; Mintzberg, 1973; Dess and
Lumpkin, 2005). According to Miller (1983, p. 771), a firm is entrepreneurial when it “engages
in product-market innovation, undertakes somewhat risky ventures, and is first to come up with
‘proactive’ innovations, beating competitors to the punch.” As such, EO manifests through
higher levels of organizational innovativeness (engaging in creativity and experimentation by
introducing new products/services, as well as technological leadership via R&D); risk-taking
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
9
This article is protected by copyright. All rights reserved.
(venturing into the unknown, borrowing heavily, and/or committing significant resources in
uncertain environments); and proactiveness (forward-looking, opportunity-seeking initiatives
launched in anticipation of future demand) (Miller and Friesen, 1978; Venkatraman, 1989a;
Lumpkin and Dess, 1996). Hence, EO captures the extent to which organizations may be
considered entrepreneurial in their firm-level activities and strategic behaviors (Covin and
Wales, 2012; George, 2011; Hoskisson et al., 2011).
Miller and Friesen (1982) contend that executive goals and temperaments are central to
driving EO in a firm. It has been theorized that the top managers of EO firms adopt and exhibit
entrepreneurial management styles, which are evidenced within the firms’ strategic decisions and
operating management philosophy (Covin and Slevin, 1989). Thus, consistent with Hambrick
and Mason’s (1984) upper echelons perspective which posits that CEOs are responsible for
helping to set and direct the organization’s strategic orientation, CEOs have a major role in
influencing the manifestation of firm-level EO (Simsek et al., 2010; Wales et al., 2011).
Several features of a narcissistic personality increase the likelihood that a narcissistic
CEO’s actions will be closely aligned with EO organizational strategies. These personality
characteristics of narcissistic leaders include biased expectations that their decisions will be
successful, a desire to inspire awe and adoration among followers, and a lack of concern for
resource constraints when pursuing their agendas. We now explore each of these characteristics.
First, narcissists make decisions based on biased expectations that they will be successful
(Campbell et al., 2004), and exhibit confidence levels that notably exceed accurate perceptions of
a situation (Lakey et al., 2008). As a strategic orientation, EO is characterized by the frequency
with which organizations commit sizable portions of their resources towards the pursuit of
innovative product-market entries (Covin and Slevin, 1989; Lumpkin and Dess, 1996). To
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
10
This article is protected by copyright. All rights reserved.
manifest a high level of EO firm leadership must judge numerous risky, innovative projects to
have sufficient “upside” to support the commitment of organizational resources. When
considering the potential gains from engaging in higher-risk, higher-return entrepreneurial
initiatives, there is an increased likelihood that narcissistic CEOs will lead their firms to pursue a
greater number of new entry opportunities and thus manifest a higher level of EO.
Second, narcissistic CEOs’ attempts to inspire awe and foster adoration may further
increase their propensity to support organizational strategies aimed at pioneering bold
innovations. Adoration fosters an ‘invincibility complex’, which further increases the probability
that a narcissistic CEO will lead their organization to engage in high-risk, high-return projects
(Lubit, 2002). Developing a firm EO would be attractive to narcissistic CEOs because strategies
involving innovation and pioneering are more likely to enhance their power and influence, and
thus their ability to inspire awe and adoration among followers. Hence, entrepreneurial actions
provide the greatest opportunity for the firm to experience growth and expansion as opposed to
more conservative actions (Covin et al., 2006; Lumpkin and Dess, 1996).
Third, because narcissistic CEO’s have inflated perceptions of how others view their
leadership and abilities, they exhibit overconfidence and believe that the consequences of high
risk activities will be free of costs (Lobuts and Pennewill, 1986). Overconfident managers are
more likely to pursue opportunities, such as new product-market entry, with less concern for the
size of their available resource bases or capabilities (Hayward et al., 2006). Given their extreme
confidence narcissistic CEOs are also more likely to ignore counterfactual information when
considering opportunity costs (Campbell et al., 2004; Judge et al., 2006). Hence, narcissistic
CEOs have a tendency to exhibit judgment and decision-making biases which increase their
likelihood of supporting bolder firm strategies irrespective of resource limitations.
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
11
This article is protected by copyright. All rights reserved.
In sum, narcissistic CEOs are likely to pioneer their grand innovative visions without the
expected concern for risk, resource limitations, or feedback. Narcissists are not fond of either
dissent or differing opinions, possess an exaggerated sense of control over the world around
them, and seek to expand their power and influence (e.g., Resick et al., 2009; Watson et al.,
1991). This suggests that they are likely to be among the first to champion bold innovations, and
push their firms toward new entry. Given this propensity, we suggest that a higher level of EO
will be found among firms lead by narcissistic CEOs. Therefore, we hypothesize:
Hypothesis 1: CEO narcissism is positively associated with entrepreneurial orientation.
Entrepreneurial Orientation and Firm Performance Variance
In the sub-section that follows, we discuss the potential influence of CEO narcissism on
firm performance variability as partially mediated by EO. Here, our aim is to argue that EO is a
likely source of variability in firm performance. Relative to other strategic postures, EO is a
more risky strategic orientation (Miller, 1983; Lumpkin and Dess, 1996; Covin and Slevin,
1989). Evidence suggests that higher risk is associated with both higher positive and higher
negative returns to firm performance (Fama and MacBeth, 1973; Miller and Bromiley, 1990).
Given that few theoretical and empirical studies have explicitly considered the link between EO
and firm performance variance, we now examine three areas in which EO is a more risky
strategic orientation—organizational uncertainty, opportunity selection, and experimentation.
First, uncertainty, or inestimable risk, is a fundamental aspect of the entrepreneurial
process (Knight, 1921). Uncertainty makes the final performance returns from developing an
entrepreneurial orientation unknowable. The degree to which entrepreneurial efforts succeed or
fail to create performance gains is impossible to predict a priori. Theoretically, if the returns
from entrepreneurial initiatives were not shrouded in uncertainty, all new product-market entries
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
12
This article is protected by copyright. All rights reserved.
would return positive gains since a firm would choose only those entries destined to succeed and
thereby improve firm financial positions. Given this, EO is a more risky strategic orientation.
Second, previous research suggests that a firm may be considered more (or less)
entrepreneurial based upon the types of opportunities which it pursues (e.g., Baird and Thomas,
1985; Covin and Slevin, 1989; Lumpkin and Dess, 1996; Miller, 1983; Miller, 2011).
Opportunities for new entry can range from more incrementally focused product-market
advances with relatively low financial risk, to more growth focused and therefore financially
risky, product-market entries or technology-based transformations (March, 1991). Along this
continuum, EO represents a strategic orientation characterized by the pursuit of more risky,
growth oriented opportunities (Bierly et al., 2009; Wiklund and Shepherd, 2011).
Finally, EO has recently been discussed as a process of experimentation conducted by
organizations as they search for profitable new entry opportunities (Wiklund and Shepherd,
2011). An experimentation perspective suggests that new entries are essentially untested
hypotheses which the firm must examine to determine if their insights were correct (Garvin,
2004). Since some of the experiments will result in failures, this emerging perspective serves to
further illustrate that EO is a risky strategic orientation, and that greater variation in firm
performance from higher levels of EO is inevitable. Therefore, based upon these fundamental
characteristics of the strategic orientation, we offer the following hypothesis:
Hypothesis 2: EO is positively associated with firm performance variance.
CEO Narcissism, Firm Performance Variance and Entrepreneurial Orientation
Given the work of Chatterjee and Hambrick (2007), which observes a positive relationship
between narcissistic CEO leadership and more extreme performance levels, we expect that CEO
narcissism will have a direct impact on firm performance variance. Nonetheless, there is limited
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
13
This article is protected by copyright. All rights reserved.
understanding of how narcissism may manifest itself through firm’s strategic behaviors to
increase performance variability. While the direct effects of narcissism on performance could
represent an aggregation of various outcomes resulting from narcissism, a partial mediation
model controls for such direct effects while enabling theorizing concerning the unique effects of
narcissism on performance variability through EO. In this sub-section, therefore, we argue that
EO partially mediates the CEO narcissism—performance variance relationship.
Our first rationale for CEO narcissism leading to greater EO and hence higher firm
performance variance builds upon the observation that CEOs are central to driving the scope of
organizational project investment (Hayward et al., 2006). CEOs are in a position to initiate and
expand the scope of investments into an organization’s entrepreneurially oriented firm processes,
practices, and projects. EO is a resource intensive strategic orientation (Covin and Slevin, 1991)
and the influence of EO on performance outcomes should increase when more resources are
available (Wiklund and Shepherd, 2011). Given that, narcissistic CEOs may be emboldened to
garner control over more resources when pursuing entrepreneurial initiatives.
In general, narcissistic CEOs are likely to be organizational resource ‘hogs’ who take
possession of whatever resources are accessible, even when such acquisitions might result in
resource depletions that damage the effectiveness of other firm objectives (Campbell et al.,
2005). Given their exaggerated perceptions that their decisions will be successful (Lakey et al.,
2008), narcissistic CEOs will often attempt to gain control over the largest possible resource
pools from which to support aggressive levels of investments into firm initiatives which promise
to help achieve their grandiose ambitions. When those grandiose ambitions are entrepreneurially
oriented as discussed in hypothesis one, investments in EO activities will naturally grow in
magnitude. But since not all EO activities will have a positive pay-off (Wiklund and Shepherd,
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
14
This article is protected by copyright. All rights reserved.
2011), in firms where narcissistic CEOs influence greater investment into higher levels of EO,
increased variability in firm performance is likely.
Secondly, the creation of firm performance variance through EO may also be influenced by
the erratic behavior of narcissistic CEOs. Because entrepreneurial initiatives often take longer to
reach fruition and payoff than more incremental or conservative firm actions (Covin and Slevin,
1991; Hoskisson et al., 2011), investments into EO firm initiatives must be given time to
develop. Yet, narcissistic CEOs lack self-control and are prone to change the strategic behavior
of the firm on an impulse (Chatterjee and Hambrick, 2007; Lubit, 2002: Vazire and Funder,
2006; Wink and Donahue, 1997). When narcissistic CEOs become impatient or bored, or
perceive that a new entrepreneurial initiative could more immediately enhance the positive
accolades they require to sustain their self-concept, they may abandon their current
entrepreneurial investments without concern for the depth of their resource allocations. Such
volatile strategic changes to their entrepreneurial initiatives could, at times, help them develop
more profitable emerging opportunities, and thus lead to increased performance. However, it is
also likely that such behavior will waste resources and prematurely scuttle promising
opportunities, leading to decreased performance. Hence, the tendency of narcissistic CEOs to
produce grand swings in resource deployments also makes it more likely that they will contribute
to greater variability in firm performance when encouraging higher levels of firm EO.
While it is probable that CEO narcissism will contribute to firm performance variance
through EO, it is plausible that narcissistic leadership will also manifest other, additional,
indirect effects on firm performance variance. It would be challenging, if not impossible, to
theoretically rule out other strategic firm-level factors that narcissistic CEOs might influence,
such as the organizations level of resource-orchestration capability (Sirmon et al., 2007) or
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
15
This article is protected by copyright. All rights reserved.
strategic flexibility (Anderson et al., 2009). As such firm-level strategic factors may allow
narcissistic CEOs to more proficiently manipulate their organizations resource bases to conform
to their ambitions, they may also partially explain CEO narcissism’s impact on firm performance
variance. Therefore, we theorize that EO will play a partial mediating role in the relationship.
Taken together, we argue that some, however not all, of the influence which CEO
narcissism has on firm performance variance occurs through EO. Both CEO narcissism and EO
are likely to drive firm performance variance. However, a meaningful portion of the increased
variance may be explained by the partial mediating effect of entrepreneurial strategic orientations
in narcissistic CEO led companies. Therefore, we hypothesize:
Hypothesis 3: EO partially mediates the relationship between CEO narcissism and firm
performance variance (i.e., in addition to directly increasing performance variance, CEO
narcissism indirectly increases variability through promoting higher firm-level EO).
Building upon and extending past theory, Figure 1 presents our research model.
--------Insert Figure 1 about here ---------
METHODS
Data Collection Procedure and Data Quality
Sampling frame. We draw on archival performance data and survey data of high-
technology manufacturing firms in the mid-western United States with 10 to 250 employees and
10 years of age or younger. Smaller, younger firms represent an ideal context for studying as
executive’s influence on EO and performance outcomes is likely to be stronger and more direct
in this context (Rauch et al., 2009; Wiklund, 1999). By limiting our focus to a single industry
sector we limit potential heterogeneity in factors affecting performance variance, reduce the
effects of unobserved factors driving variously narcissistic individuals to lead more or less EO
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
16
This article is protected by copyright. All rights reserved.
companies, and control for industry specific factors which may either reinforce or mitigate the
effects of narcissistic leadership. Additionally, an empirical rationale driving our sampling frame
was the ability to identify performance data from third-party sources. The potential common
method bias resulting from self-report performance could be severe when investigating CEO
narcissism as prior research would suggest that respondents high in narcissism are more likely to
underreport lower performance and overreport higher performance (Campbell et al., 2004). By
focusing on the high-tech manufacturing sector in the present study we were able to triangulate
performance information from two third-party sources containing secondary sales information –
Corptech, a directory of high-tech firms, and Dun and Bradstreet, a directory of private firms.
We started with the Corptech directory as our sampling frame. Corptech is considered a
reliable source of technology firm listings, and has been used widely for research on high
technology firms (e.g., Lee and Lieberman, 2010). In addition, Corptech provides yearly sales
and firm size (number of employees) information collected through surveys and annual phone
interviews. To balance survey cost and scope we focused on high-tech manufacturing firms in
the U.S. states of Illinois, Indiana, Kentucky, Ohio, and Missouri. We were able to identify 1,526
listed in the 2009 Corptech directory, representing thirty different NAICS codes
3
.
Contacting CEOs. CEOs play a central role in the strategic direction and day-to-day
management of small firms. They are well-informed of strategic issues that explicitly entail an
organizationally-wide or external focus (Sharfman, 1998). Thus, traditionally top management
team members in general, and the CEO in particular is surveyed regarding his or her perceptions
of the firm’s EO (Covin and Slevin, 1989; Rauch et al., 2009; Wales et al., 2011). A packet
containing our survey, along with a cover letter and pre-paid business reply envelope was sent to
the CEO of each firm. To enhance typically low response rates among CEOs, in the initial mail
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
17
This article is protected by copyright. All rights reserved.
survey we informed the respondents that we would donate USD $20 for every complete survey
to a charity of their choice. After the first mailed survey, and three follow-up reminder emails
conducted between October 2009 and January 2010, we received responses from 219 CEOs, for
a response rate of 14.35%. Low response rates (10-15%) are typical for mailed surveys to top
executives, particularly when asking sensitive questions, and are comparable to other similar
studies (Cycyota and Harrison, 2002; Ling et al., 2008). In the final data set, we excluded 46
firms with incomplete data on narcissism scale items. This yielded a final sample of 173 firms.
Assessing key informant quality and nonresponse bias. To assess the extent of CEO
involvement in the manifestation of firm–level EO, we used three self-report items (1 – ‘not at
all,’ to 5 – ‘to a great extent’): (a) ‘How involved are you in developing and implementing your
venture’s innovation strategies?’ [mean = 4.83; s.d.=.24], (b) ‘How actively are you engaged in
managing venture’s risk?’ [mean = 4.65; s.d.=.19], (c) ‘How actively are your involved in
identifying, pursuing, and exploiting opportunities in the environment’ [mean = 4.25; s.d.=.31].
This provides the present investigation with some assurance that the respondents had detailed
knowledge concerning their organization’s EO.
We tested non-response bias for early and late respondents and also the mean responses
of respondents and non-respondents on age, sales revenues, firm size (number of employees),
industry, CEO age, CEO gender, and CEO industry work experience. We found no significant
differences, leading us to conclude that non-response bias was not a significant threat to study
estimates. Furthermore, we calculated sampling error based on all firms in the thirty NAICS
codes sampled from the Corptech directory in 2009. The sampling error of 5.87% (at a 95%
confidence level) is below recommended limits and indicates that the sample is generally
representative of the population of high-technology firms (Särndal et al., 2003).
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
18
This article is protected by copyright. All rights reserved.
Dependent variable. Firm performance variance. The importance of sales as a
meaningful dependent variable in the case of small firms has been previously theorized
(Chandler et al., 2009; Hmieleski and Corbett, 2008). Moreover, we chose to examine sales data
given its secondary reporting within the Corptech directory and Dun and Bradstreet Million
Dollar Directory. We triangulate the secondary sales data from the Corptech and Dun and
Bradstreet Directories. Apart from our survey, triangulating sales data from two secondary
sources lessens concerns regarding common method bias and provides a more consistent
measure of sales. The efficacy and use of such directories for the compilation of secondary
performance measures has been supported by numerous studies (Lee and Lieberman, 2010; Lee
and Paruchuri, 2008; Puranam et al., 2009; Puranam and Srikanth, 2007; Sine et al., 2006).
We began by compiling sales for 2006, 2007, 2008, 2009, and 2010 from Corptech and
the Dun and Bradstreet Million Dollar Directory. We also compiled information on industry-
level sales by collecting data for each firm continuously listed for all five years (2006-2010) at
the six-digit NAICS code in Corptech and Dun and Bradstreet Million Dollar Directory. We used
the mean reported sales values from Corptech and the Dun and Bradstreet Million Dollar
Directory for each six-digit NAICS sector. Examining this data, we were able to calculate the
standard deviation of the sales for each firm, and also calculate the standard deviation of sales for
all industry firms over the five-year period. Although firms may experience significant
fluctuations in sales, by adjusting for the standard deviation of industry sales we further control
for systematic industry specific effects. We then subtracted the median standard deviation of
sales at the industry-level from each firm’s standard deviation of sales to arrive at our final
measure of industry-adjusted firm performance variance.
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
19
This article is protected by copyright. All rights reserved.
Independent variable. CEO narcissism. Previous studies have used proxies of
narcissism based on information available in the public domain (e.g., 10-K annual reports,
ExecuComp, Chatterjee and Hambrick, 2007). It is important to note that all measurement
approaches are imperfect. Information within the public domain is influenced by sources beyond
the CEO such as inputs from the top management team, the board of directors, and other
stakeholders. Several individuals may be involved in preparing shareholder letters or grooming
CEOs for interviews, and further, the content analysis used to develop proxies for CEO
prominence may introduce researcher bias (Krippendorff, 2004). Therefore, we offer a different,
but a complementary test of the CEO-narcissism—Firm-performance-variance relationship using
a previously validated and well established psychometric scale.
Narcissism was assessed using a continuous measure based on the 16-item measure (NPI-
16) proposed and validated by Ames and colleagues (2006). The scale consists of 16-paired
statements that require a forced choice between a narcissistic response (=1) and a non-
narcissistic response (=0). This scale has been psychometrically validated (Ames and Kammrath,
2004) and is a shorter version of Raskin and Terry’s (1988) NPI-40. Although the 16-item scale
used in the current study has been validated only recently, the longer version of this scale (NPI-
40) has found increased support over the years. Samuel and Widiger (2008) provide a review of
the face, content, convergent, divergent and nomological validity of the narcissism scale, and
several other meta-analyses offer evidence of the validity of the NPI-40 scale (Campbell et al.,
2000; Campbell et al., 2002; Campbell et al., 2007; Foster et al., 2003; Twenge and Foster, 2008;
Twenge et al., 2008). As a derivative of NPI-40, the shorter 16-item version proposed by Ames
and colleagues (2006) is rooted within a robust tradition of psychometric-research exploring
narcissism.
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
20
This article is protected by copyright. All rights reserved.
The forced-choice scale includes items such as ‘I really like to be the center of attention
vs. It makes me uncomfortable to be the center of attention,’ and ‘I think I am a special person
vs. I am no better or no worse than most people.’ The respondent is asked to choose one of the
two statements in each scale item. For example, is the respondent chooses ‘I really like to be the
center of attention’ it is coded as 1, if the respondent chooses ‘It makes me uncomfortable to be
the center of attention’ then it is coded as 0. We list scores of either 1 or 0 for each item in Table
A.I. The final narcissism score, based on Ames et al. (2006), is the sum of the scores divided by
16. The measure of narcissism is the average of all of the forced choice responses. The final 16-
item scale measure is continuous, ranging from 0 to 1, where ‘1’ represents the highest level of
narcissism and ‘0,’ represents the lowest level of narcissism.
The recommended cut-off values for item loadings, reliability, average variance
exteracted (AVE), constrained-unconstrained models, and confirmatory factor analysis (CFA)
are listed in Table I. Significant item loadings indicate stronger evidence for a relationship
among the observed indicators and their respective latent factors. In addition to Cronbach's
alpha, we also use two other measures of reliability—Raykov's ? and ordinal reliability alpha.
For discriminant validity we use AVE, and constrained-unconstrained models as suggested by
Venkatraman (1989b).
--------Insert Table I about here ---------
The lowest item loading for the narcissism measure was 0.742 (t-value = 8.337) and the
highest item loading was 0.869 (t-value=8.355). The scale reliability for narcissism was above
the recommended cut-off of 0.7 (?=0.782; ?=0.783; ordinal reliability = 0.785). The AVE
(=0.653) was above the recommended cut-off, and the change in chi-square was significant for
constrained-unconstrained models between narcissism and EO, and EO’s sub-scales. Finally,
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
21
This article is protected by copyright. All rights reserved.
confirmatory factor analysis (CFA) in Table I, shows acceptable fit (?
2
(df)=125.348 (103);
CFI=0.942, TLI=0.937; RMSEA=0.065; SRMR=0.031).
Mediating variable. Entrepreneurial orientation. To measure EO we use the widely
adopted scale developed by Covin and Slevin (1989). The three components of EO are:
innovativeness (3 items), proactiveness (3 items), and risk taking (3 items). Sample items include
“My business places a strong emphasis on: Tried and tested practices, equipment, and products
or services vs. Innovation, technological leadership and R&D” [innovativeness], “My business
typically Responds to initiatives my competitors initiate vs. Initiates action to which my
competitors then respond” [proactiveness], “My business is inclined toward low risk projects
with normal rates of return vs. High risk projects with a chance of very high returns” [risk-
taking]. As shown in Table A.I. (Appendix), all item loadings were significant. The lowest item
loading for EO was for EO-risk-taking (0.778; t-value=5.328) and highest loading was for EO-
Innovativeness (0.882; t-value=7.875). The scale reliability for EO was above the recommended
cut-off of 0.7 (?=0.836; ?=0.839; ordinal reliability = 0.839). Table A.II (Appendix) additionally
confirms a robust second-order fit for the EO measure. Furthermore, as listed in Table A.I, AVE
for EO (=0.745) was above the recommended cut-off of 0.5. Constrained-unconstrained analysis
listed in Table 1 also shows that the lowest change in chi-square was 8.505 (1), p
During this such a detailed information in regard to in pursuit of greatness ceo narcissism, entrepreneurial orientation, and firm performance.
1
This article is protected by copyright. All rights reserved.
In Pursuit of Greatness: CEO Narcissism, Entrepreneurial Orientation, and Firm Performance Variance
1
WILLIAM J. WALES
2
James Madison University
Department of Management
800 S. Main Street
Harrisonburg, VA 22801
Tel: (540) 568-4375
Fax: (540) 568.3275
e-mail: [email protected]
PANKAJ C. PATEL
Miller College of Business
Ball State University
Muncie, IN 47306
Tel: 765-285-3194
e-mail: [email protected]
and
This article has been accepted for publication and undergone full peer review but has not been through the
copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version
and the Version of Record. Please cite this article as doi: 10.1111/joms.12034
2
Corresponding author
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
2
This article is protected by copyright. All rights reserved.
G. THOMAS LUMPKIN
Syracuse University
Whitman School of Management
721 University Avenue
Syracuse, NY 13244-2450
Tel: (315) 443-3164
e-mail: [email protected]
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
3
This article is protected by copyright. All rights reserved.
IN PURSUIT OF GREATNESS: CEO NARCISSISM, ENTREPRENEURIAL
ORIENTATION, AND FIRM PERFORMANCE VARIANCE
ABSTRACT
Building upon the perspective that narcissism is a leadership trait with both ‘bright’ and ‘dark’
sides, the present study examines the question of whether companies led by narcissistic CEOs
exhibit higher levels of entrepreneurial orientation (EO). Moreover, this research examines
whether EO partially explains why narcissistic CEO led firms experience greater variability in
firm performance. Using survey data collected from 173 CEOs, and an archival measure of firm
performance variance, we find support for our model. These findings offer an improved
understanding of how CEO narcissism influences performance variance, and why the firms they
lead may even, at times, be viewed as on a path to success. Study implications are discussed.
Keywords:
CEO Leadership Traits; Narcissism; Strategic Posture; Entrepreneurial Orientation; Firm
Performance Variance
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
4
This article is protected by copyright. All rights reserved.
INTRODUCTION
Narcissism is broadly defined as an exaggerated, yet fragile self-concept of one’s
importance and influence (Resick et al., 2009). Most of the work on narcissism to date has
explored the negative implications of the phenomenon for individual-level functioning and
leadership (Judge et al., 2006; Kets De Vries and Miller, 1985). However, prior research also
indicates that many chief executive officers (CEOs) who are perceived as narcissistic lead highly
successful companies (Maccoby, 2000). How might this be so?
To begin, Chatterjee and Hambrick (2007) observe that narcissistic CEOs are associated
with more extreme firm performance—both bigger gains and bigger losses. While increased
variability in firm performance has neither an inherently positive or negative organizational
implication, higher variability can be interpreted as beneficial if the achievement of a superior
competitive position—i.e., above-average performance—is preferred over maintenance of the
firm’s extant competitive position which lower variance would help protect (March, 1991). Yet,
CEO narcissism itself is insufficient to fully account for such firm performance extremes in the
pursuit of superior performance. This raises the research question: What firm-level mechanisms
might narcissistic CEOs employ to obtain such wide variations in organizational outcomes?
Given that entrepreneurial processes and practices are often associated with stimulating
company renewal and growth (Guth and Ginsberg, 1990; Hoskisson et al., 2011), CEOs often
promote a firm-level entrepreneurial orientation (EO) within the organizations they lead (Rauch
et al., 2009). As will be discussed, firm-level entrepreneurial activities have also been identified
as an explanatory factor of firm performance variance, that is, relatively stronger and relatively
weaker firm performance (Wiklund and Shepherd, 2011). The present study proposes a linkage
between CEO narcissism and firm-level EO, and specifically asks whether EO might be a key
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
5
This article is protected by copyright. All rights reserved.
firm-level mechanism that enables narcissistic CEOs to lead organizations that experience a wide
range of performance outcomes. Chatterjee and Hambrick (2007) demonstrate that CEO
narcissism can affect firm strategy-making and performance variance independently. The present
study seeks to extend their findings by asking whether EO partially mediates the relationship
between CEO narcissism and firm performance variance.
By investigating these questions, this research makes three principal contributions to the
management literature. First, by further clarifying the relationship between CEO narcissism and
firm performance variance, we address the call by Rosenthal and Pittinsky (2006) to progress
beyond simple discussions of narcissistic leadership as inherently ‘good’ or ‘bad’, and focus
instead on its implications for company outcomes. In this vein we also build on the emerging
perspective that there are both “bright” and “dark” sides to narcissistic CEO leadership (Resick
et al., 2009), and thus it may have both positive and negative effects on their companies fate.
Second, our study extends the work of Chatterjee and Hambrick (2007) by identifying
EO as a strategy making construct that is not only affected by CEO narcissism, but also partially
mediates the CEO narcissism—performance variance relationship. Given the important and
sometimes dramatic role that entrepreneurial processes and practices may play in changing the
fate or direction of a company (Hoskisson et al., 2011), we posit that the tendency of narcissistic
CEOs to endorse firm-level EO helps explain why their organizations experience an increase in
performance variance, and may even, at times, be viewed as on a path to success. Third, while
narcissism has become a broad managerial phenomenon of interest (Bergman et al., 2010), little
consideration has been given to empirically examining this phenomenon using psychometrically
valid scales within broad samples of executives (Bollaert and Petit, 2010). Our study provides a
robust test of our theorized model using both survey data and archival performance measures.
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
6
This article is protected by copyright. All rights reserved.
The remainder of the paper is organized as follows. First we review narcissism as a
leadership trait. We then develop a research model and posit hypotheses linking CEO narcissism,
EO, and firm performance variance. This is followed by a discussion of methods and a report of
findings. Finally, we conclude with a discussion of implications and future research directions.
THEORETICAL BACKGROUND AND HYPOTHESIS DEVELOPMENT
Narcissism—an inflated but fragile view of one’s self-importance—is a construct that has
generated substantial interest across the social sciences for many decades (e.g., Morf and
Rhodewalt, 2001). The negative consequences of narcissism have been well documented in the
literature (e.g., Resick et al., 2009). Narcissistic individuals are principally concerned with
actions that reflect favorably upon themselves, which often leads to perceptions of entitlement,
shameless self-admiration, excessive arrogance, and hostility towards external criticism (Judge et
al., 2006; Lubit, 2002; Resick et al., 2009). Thus, in interpersonal relationships, they are often
viewed as abrasive and insensitive to their impact on others (Judge et al., 2009). Despite the
interpersonal issues typically associated with narcissism, maintaining unrealistic, inflated levels
of self-esteem creates a continuous need for narcissists to obtain external self-affirmation
through social interaction. One place narcissists receive such affirmation is in leadership roles.
A key reason narcissists are attracted to organizational leadership roles is their desire to
leave behind a grand, admirable legacy of achievement (Maccoby, 2000). Organizational settings
can provide narcissistic leaders with positions of authority and influence that can facilitate
achievement of their personal ambitions. In general, the role of CEO provides narcissists with the
greatest authority and hence studies in the management literature have previously focused on
examining narcissism among CEOs (e.g., Chatterjee and Hambrick, 2007; Resick et al., 2009).
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
7
This article is protected by copyright. All rights reserved.
An analysis of narcissism and leadership by Kets De Vries and Miller (1985) was among
the first to appear in the management literature. They noted that narcissism, while typically
viewed as detrimental, can also be conceived as beneficial for firm leaders and linked to higher
productivity. Subsequent analyses tend to support this view. Describing the tension between the
positive and negative elements of narcissistic leadership, Judge and colleagues (2009) classified
narcissism as a ‘dark’ leadership trait with some potentially ‘bright’ sides. In this regard,
Rosenthal and Pittinsky (2006) note that while it may at first seem ‘preposterous’ to suggest that
narcissists could possess strong leadership characteristics with both negative and positive
implications for organizational performance, it is certainly possible even if it does occur as a
result of self-aggrandizing motivations. For instance, it has been argued that narcissists, who are
generally viewed as arrogant, self-important and insensitive, can also be considered tough-
minded visionaries and evangelistic leaders (Maccoby, 2004). In this regard, some of the
positive attributes of narcissism operate similarly to those of charisma
1
. Both may inspire
devotion and commitment in others to achieve higher performance. This semblance may help
explain why individuals are sometimes attracted to narcissistic leaders and may even view their
leadership as transformational (Rosenthal and Pittinsky, 2006).
While extreme narcissism is a diagnosable personality disorder
2
, social science research
generally views it as a matter of degree, ranging from mild to severe (Foster and Campbell,
2007; Kets De Vries and Miller, 1985). Consistent with current research in social psychology
and strategic management, we view narcissism among executives as not necessarily reflective of
a personality disorder which presents itself as a binary state (i.e., sufferers v. non- sufferers).
Rather, we view narcissism as a personality trait which manifests in varying degrees along a
continuum. By adopting this view we recognize that different levels of narcissism among CEOs
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
8
This article is protected by copyright. All rights reserved.
can influence a range of organizational strategic behaviors. But the force of CEO personality
alone is insufficient to fully explain changes in firm performance. For CEO attributes and
ambitions to influence organizational outcomes, they must be translated into specific firm-level
strategic behaviors. In the following sub-section, we investigate the implications of CEO
narcissism for the manifestation of a particular firm-level strategic orientation, EO.
CEO Narcissism and Entrepreneurial Orientation
When CEOs possess higher levels of narcissism they are more likely to strive for bold,
daring actions which increase their stature by drawing attention to their vision and leadership
(Judge et al., 2009). Further, narcissists are driven by an overwhelming desire to compete
(Maccoby, 2004; Rosenthal and Pittinsky, 2006). To fulfill these needs, narcissistic CEOs will
likely direct their organizations to take bolder and more aggressive strategic actions, which
would be characteristic of an EO (Covin and Slevin, 1991; Lumpkin and Dess, 1996). We
propose that by inspiring their firms to assume a more entrepreneurial, competitively focused
strategic posture (Rauch et al., 2009), narcissistic CEOs can enhance their personal prestige and
fulfill their desire for positive accolades through the attributions they receive for their role in the
entrepreneurial initiatives their firm undertakes.
Entrepreneurial orientation (EO) refers to the strategy-making policies and practices that
firms use to identify and launch new ventures (Miller, 2011; Mintzberg, 1973; Dess and
Lumpkin, 2005). According to Miller (1983, p. 771), a firm is entrepreneurial when it “engages
in product-market innovation, undertakes somewhat risky ventures, and is first to come up with
‘proactive’ innovations, beating competitors to the punch.” As such, EO manifests through
higher levels of organizational innovativeness (engaging in creativity and experimentation by
introducing new products/services, as well as technological leadership via R&D); risk-taking
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
9
This article is protected by copyright. All rights reserved.
(venturing into the unknown, borrowing heavily, and/or committing significant resources in
uncertain environments); and proactiveness (forward-looking, opportunity-seeking initiatives
launched in anticipation of future demand) (Miller and Friesen, 1978; Venkatraman, 1989a;
Lumpkin and Dess, 1996). Hence, EO captures the extent to which organizations may be
considered entrepreneurial in their firm-level activities and strategic behaviors (Covin and
Wales, 2012; George, 2011; Hoskisson et al., 2011).
Miller and Friesen (1982) contend that executive goals and temperaments are central to
driving EO in a firm. It has been theorized that the top managers of EO firms adopt and exhibit
entrepreneurial management styles, which are evidenced within the firms’ strategic decisions and
operating management philosophy (Covin and Slevin, 1989). Thus, consistent with Hambrick
and Mason’s (1984) upper echelons perspective which posits that CEOs are responsible for
helping to set and direct the organization’s strategic orientation, CEOs have a major role in
influencing the manifestation of firm-level EO (Simsek et al., 2010; Wales et al., 2011).
Several features of a narcissistic personality increase the likelihood that a narcissistic
CEO’s actions will be closely aligned with EO organizational strategies. These personality
characteristics of narcissistic leaders include biased expectations that their decisions will be
successful, a desire to inspire awe and adoration among followers, and a lack of concern for
resource constraints when pursuing their agendas. We now explore each of these characteristics.
First, narcissists make decisions based on biased expectations that they will be successful
(Campbell et al., 2004), and exhibit confidence levels that notably exceed accurate perceptions of
a situation (Lakey et al., 2008). As a strategic orientation, EO is characterized by the frequency
with which organizations commit sizable portions of their resources towards the pursuit of
innovative product-market entries (Covin and Slevin, 1989; Lumpkin and Dess, 1996). To
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
10
This article is protected by copyright. All rights reserved.
manifest a high level of EO firm leadership must judge numerous risky, innovative projects to
have sufficient “upside” to support the commitment of organizational resources. When
considering the potential gains from engaging in higher-risk, higher-return entrepreneurial
initiatives, there is an increased likelihood that narcissistic CEOs will lead their firms to pursue a
greater number of new entry opportunities and thus manifest a higher level of EO.
Second, narcissistic CEOs’ attempts to inspire awe and foster adoration may further
increase their propensity to support organizational strategies aimed at pioneering bold
innovations. Adoration fosters an ‘invincibility complex’, which further increases the probability
that a narcissistic CEO will lead their organization to engage in high-risk, high-return projects
(Lubit, 2002). Developing a firm EO would be attractive to narcissistic CEOs because strategies
involving innovation and pioneering are more likely to enhance their power and influence, and
thus their ability to inspire awe and adoration among followers. Hence, entrepreneurial actions
provide the greatest opportunity for the firm to experience growth and expansion as opposed to
more conservative actions (Covin et al., 2006; Lumpkin and Dess, 1996).
Third, because narcissistic CEO’s have inflated perceptions of how others view their
leadership and abilities, they exhibit overconfidence and believe that the consequences of high
risk activities will be free of costs (Lobuts and Pennewill, 1986). Overconfident managers are
more likely to pursue opportunities, such as new product-market entry, with less concern for the
size of their available resource bases or capabilities (Hayward et al., 2006). Given their extreme
confidence narcissistic CEOs are also more likely to ignore counterfactual information when
considering opportunity costs (Campbell et al., 2004; Judge et al., 2006). Hence, narcissistic
CEOs have a tendency to exhibit judgment and decision-making biases which increase their
likelihood of supporting bolder firm strategies irrespective of resource limitations.
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
11
This article is protected by copyright. All rights reserved.
In sum, narcissistic CEOs are likely to pioneer their grand innovative visions without the
expected concern for risk, resource limitations, or feedback. Narcissists are not fond of either
dissent or differing opinions, possess an exaggerated sense of control over the world around
them, and seek to expand their power and influence (e.g., Resick et al., 2009; Watson et al.,
1991). This suggests that they are likely to be among the first to champion bold innovations, and
push their firms toward new entry. Given this propensity, we suggest that a higher level of EO
will be found among firms lead by narcissistic CEOs. Therefore, we hypothesize:
Hypothesis 1: CEO narcissism is positively associated with entrepreneurial orientation.
Entrepreneurial Orientation and Firm Performance Variance
In the sub-section that follows, we discuss the potential influence of CEO narcissism on
firm performance variability as partially mediated by EO. Here, our aim is to argue that EO is a
likely source of variability in firm performance. Relative to other strategic postures, EO is a
more risky strategic orientation (Miller, 1983; Lumpkin and Dess, 1996; Covin and Slevin,
1989). Evidence suggests that higher risk is associated with both higher positive and higher
negative returns to firm performance (Fama and MacBeth, 1973; Miller and Bromiley, 1990).
Given that few theoretical and empirical studies have explicitly considered the link between EO
and firm performance variance, we now examine three areas in which EO is a more risky
strategic orientation—organizational uncertainty, opportunity selection, and experimentation.
First, uncertainty, or inestimable risk, is a fundamental aspect of the entrepreneurial
process (Knight, 1921). Uncertainty makes the final performance returns from developing an
entrepreneurial orientation unknowable. The degree to which entrepreneurial efforts succeed or
fail to create performance gains is impossible to predict a priori. Theoretically, if the returns
from entrepreneurial initiatives were not shrouded in uncertainty, all new product-market entries
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
12
This article is protected by copyright. All rights reserved.
would return positive gains since a firm would choose only those entries destined to succeed and
thereby improve firm financial positions. Given this, EO is a more risky strategic orientation.
Second, previous research suggests that a firm may be considered more (or less)
entrepreneurial based upon the types of opportunities which it pursues (e.g., Baird and Thomas,
1985; Covin and Slevin, 1989; Lumpkin and Dess, 1996; Miller, 1983; Miller, 2011).
Opportunities for new entry can range from more incrementally focused product-market
advances with relatively low financial risk, to more growth focused and therefore financially
risky, product-market entries or technology-based transformations (March, 1991). Along this
continuum, EO represents a strategic orientation characterized by the pursuit of more risky,
growth oriented opportunities (Bierly et al., 2009; Wiklund and Shepherd, 2011).
Finally, EO has recently been discussed as a process of experimentation conducted by
organizations as they search for profitable new entry opportunities (Wiklund and Shepherd,
2011). An experimentation perspective suggests that new entries are essentially untested
hypotheses which the firm must examine to determine if their insights were correct (Garvin,
2004). Since some of the experiments will result in failures, this emerging perspective serves to
further illustrate that EO is a risky strategic orientation, and that greater variation in firm
performance from higher levels of EO is inevitable. Therefore, based upon these fundamental
characteristics of the strategic orientation, we offer the following hypothesis:
Hypothesis 2: EO is positively associated with firm performance variance.
CEO Narcissism, Firm Performance Variance and Entrepreneurial Orientation
Given the work of Chatterjee and Hambrick (2007), which observes a positive relationship
between narcissistic CEO leadership and more extreme performance levels, we expect that CEO
narcissism will have a direct impact on firm performance variance. Nonetheless, there is limited
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
13
This article is protected by copyright. All rights reserved.
understanding of how narcissism may manifest itself through firm’s strategic behaviors to
increase performance variability. While the direct effects of narcissism on performance could
represent an aggregation of various outcomes resulting from narcissism, a partial mediation
model controls for such direct effects while enabling theorizing concerning the unique effects of
narcissism on performance variability through EO. In this sub-section, therefore, we argue that
EO partially mediates the CEO narcissism—performance variance relationship.
Our first rationale for CEO narcissism leading to greater EO and hence higher firm
performance variance builds upon the observation that CEOs are central to driving the scope of
organizational project investment (Hayward et al., 2006). CEOs are in a position to initiate and
expand the scope of investments into an organization’s entrepreneurially oriented firm processes,
practices, and projects. EO is a resource intensive strategic orientation (Covin and Slevin, 1991)
and the influence of EO on performance outcomes should increase when more resources are
available (Wiklund and Shepherd, 2011). Given that, narcissistic CEOs may be emboldened to
garner control over more resources when pursuing entrepreneurial initiatives.
In general, narcissistic CEOs are likely to be organizational resource ‘hogs’ who take
possession of whatever resources are accessible, even when such acquisitions might result in
resource depletions that damage the effectiveness of other firm objectives (Campbell et al.,
2005). Given their exaggerated perceptions that their decisions will be successful (Lakey et al.,
2008), narcissistic CEOs will often attempt to gain control over the largest possible resource
pools from which to support aggressive levels of investments into firm initiatives which promise
to help achieve their grandiose ambitions. When those grandiose ambitions are entrepreneurially
oriented as discussed in hypothesis one, investments in EO activities will naturally grow in
magnitude. But since not all EO activities will have a positive pay-off (Wiklund and Shepherd,
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
14
This article is protected by copyright. All rights reserved.
2011), in firms where narcissistic CEOs influence greater investment into higher levels of EO,
increased variability in firm performance is likely.
Secondly, the creation of firm performance variance through EO may also be influenced by
the erratic behavior of narcissistic CEOs. Because entrepreneurial initiatives often take longer to
reach fruition and payoff than more incremental or conservative firm actions (Covin and Slevin,
1991; Hoskisson et al., 2011), investments into EO firm initiatives must be given time to
develop. Yet, narcissistic CEOs lack self-control and are prone to change the strategic behavior
of the firm on an impulse (Chatterjee and Hambrick, 2007; Lubit, 2002: Vazire and Funder,
2006; Wink and Donahue, 1997). When narcissistic CEOs become impatient or bored, or
perceive that a new entrepreneurial initiative could more immediately enhance the positive
accolades they require to sustain their self-concept, they may abandon their current
entrepreneurial investments without concern for the depth of their resource allocations. Such
volatile strategic changes to their entrepreneurial initiatives could, at times, help them develop
more profitable emerging opportunities, and thus lead to increased performance. However, it is
also likely that such behavior will waste resources and prematurely scuttle promising
opportunities, leading to decreased performance. Hence, the tendency of narcissistic CEOs to
produce grand swings in resource deployments also makes it more likely that they will contribute
to greater variability in firm performance when encouraging higher levels of firm EO.
While it is probable that CEO narcissism will contribute to firm performance variance
through EO, it is plausible that narcissistic leadership will also manifest other, additional,
indirect effects on firm performance variance. It would be challenging, if not impossible, to
theoretically rule out other strategic firm-level factors that narcissistic CEOs might influence,
such as the organizations level of resource-orchestration capability (Sirmon et al., 2007) or
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
15
This article is protected by copyright. All rights reserved.
strategic flexibility (Anderson et al., 2009). As such firm-level strategic factors may allow
narcissistic CEOs to more proficiently manipulate their organizations resource bases to conform
to their ambitions, they may also partially explain CEO narcissism’s impact on firm performance
variance. Therefore, we theorize that EO will play a partial mediating role in the relationship.
Taken together, we argue that some, however not all, of the influence which CEO
narcissism has on firm performance variance occurs through EO. Both CEO narcissism and EO
are likely to drive firm performance variance. However, a meaningful portion of the increased
variance may be explained by the partial mediating effect of entrepreneurial strategic orientations
in narcissistic CEO led companies. Therefore, we hypothesize:
Hypothesis 3: EO partially mediates the relationship between CEO narcissism and firm
performance variance (i.e., in addition to directly increasing performance variance, CEO
narcissism indirectly increases variability through promoting higher firm-level EO).
Building upon and extending past theory, Figure 1 presents our research model.
--------Insert Figure 1 about here ---------
METHODS
Data Collection Procedure and Data Quality
Sampling frame. We draw on archival performance data and survey data of high-
technology manufacturing firms in the mid-western United States with 10 to 250 employees and
10 years of age or younger. Smaller, younger firms represent an ideal context for studying as
executive’s influence on EO and performance outcomes is likely to be stronger and more direct
in this context (Rauch et al., 2009; Wiklund, 1999). By limiting our focus to a single industry
sector we limit potential heterogeneity in factors affecting performance variance, reduce the
effects of unobserved factors driving variously narcissistic individuals to lead more or less EO
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
16
This article is protected by copyright. All rights reserved.
companies, and control for industry specific factors which may either reinforce or mitigate the
effects of narcissistic leadership. Additionally, an empirical rationale driving our sampling frame
was the ability to identify performance data from third-party sources. The potential common
method bias resulting from self-report performance could be severe when investigating CEO
narcissism as prior research would suggest that respondents high in narcissism are more likely to
underreport lower performance and overreport higher performance (Campbell et al., 2004). By
focusing on the high-tech manufacturing sector in the present study we were able to triangulate
performance information from two third-party sources containing secondary sales information –
Corptech, a directory of high-tech firms, and Dun and Bradstreet, a directory of private firms.
We started with the Corptech directory as our sampling frame. Corptech is considered a
reliable source of technology firm listings, and has been used widely for research on high
technology firms (e.g., Lee and Lieberman, 2010). In addition, Corptech provides yearly sales
and firm size (number of employees) information collected through surveys and annual phone
interviews. To balance survey cost and scope we focused on high-tech manufacturing firms in
the U.S. states of Illinois, Indiana, Kentucky, Ohio, and Missouri. We were able to identify 1,526
listed in the 2009 Corptech directory, representing thirty different NAICS codes
3
.
Contacting CEOs. CEOs play a central role in the strategic direction and day-to-day
management of small firms. They are well-informed of strategic issues that explicitly entail an
organizationally-wide or external focus (Sharfman, 1998). Thus, traditionally top management
team members in general, and the CEO in particular is surveyed regarding his or her perceptions
of the firm’s EO (Covin and Slevin, 1989; Rauch et al., 2009; Wales et al., 2011). A packet
containing our survey, along with a cover letter and pre-paid business reply envelope was sent to
the CEO of each firm. To enhance typically low response rates among CEOs, in the initial mail
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
17
This article is protected by copyright. All rights reserved.
survey we informed the respondents that we would donate USD $20 for every complete survey
to a charity of their choice. After the first mailed survey, and three follow-up reminder emails
conducted between October 2009 and January 2010, we received responses from 219 CEOs, for
a response rate of 14.35%. Low response rates (10-15%) are typical for mailed surveys to top
executives, particularly when asking sensitive questions, and are comparable to other similar
studies (Cycyota and Harrison, 2002; Ling et al., 2008). In the final data set, we excluded 46
firms with incomplete data on narcissism scale items. This yielded a final sample of 173 firms.
Assessing key informant quality and nonresponse bias. To assess the extent of CEO
involvement in the manifestation of firm–level EO, we used three self-report items (1 – ‘not at
all,’ to 5 – ‘to a great extent’): (a) ‘How involved are you in developing and implementing your
venture’s innovation strategies?’ [mean = 4.83; s.d.=.24], (b) ‘How actively are you engaged in
managing venture’s risk?’ [mean = 4.65; s.d.=.19], (c) ‘How actively are your involved in
identifying, pursuing, and exploiting opportunities in the environment’ [mean = 4.25; s.d.=.31].
This provides the present investigation with some assurance that the respondents had detailed
knowledge concerning their organization’s EO.
We tested non-response bias for early and late respondents and also the mean responses
of respondents and non-respondents on age, sales revenues, firm size (number of employees),
industry, CEO age, CEO gender, and CEO industry work experience. We found no significant
differences, leading us to conclude that non-response bias was not a significant threat to study
estimates. Furthermore, we calculated sampling error based on all firms in the thirty NAICS
codes sampled from the Corptech directory in 2009. The sampling error of 5.87% (at a 95%
confidence level) is below recommended limits and indicates that the sample is generally
representative of the population of high-technology firms (Särndal et al., 2003).
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
18
This article is protected by copyright. All rights reserved.
Dependent variable. Firm performance variance. The importance of sales as a
meaningful dependent variable in the case of small firms has been previously theorized
(Chandler et al., 2009; Hmieleski and Corbett, 2008). Moreover, we chose to examine sales data
given its secondary reporting within the Corptech directory and Dun and Bradstreet Million
Dollar Directory. We triangulate the secondary sales data from the Corptech and Dun and
Bradstreet Directories. Apart from our survey, triangulating sales data from two secondary
sources lessens concerns regarding common method bias and provides a more consistent
measure of sales. The efficacy and use of such directories for the compilation of secondary
performance measures has been supported by numerous studies (Lee and Lieberman, 2010; Lee
and Paruchuri, 2008; Puranam et al., 2009; Puranam and Srikanth, 2007; Sine et al., 2006).
We began by compiling sales for 2006, 2007, 2008, 2009, and 2010 from Corptech and
the Dun and Bradstreet Million Dollar Directory. We also compiled information on industry-
level sales by collecting data for each firm continuously listed for all five years (2006-2010) at
the six-digit NAICS code in Corptech and Dun and Bradstreet Million Dollar Directory. We used
the mean reported sales values from Corptech and the Dun and Bradstreet Million Dollar
Directory for each six-digit NAICS sector. Examining this data, we were able to calculate the
standard deviation of the sales for each firm, and also calculate the standard deviation of sales for
all industry firms over the five-year period. Although firms may experience significant
fluctuations in sales, by adjusting for the standard deviation of industry sales we further control
for systematic industry specific effects. We then subtracted the median standard deviation of
sales at the industry-level from each firm’s standard deviation of sales to arrive at our final
measure of industry-adjusted firm performance variance.
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
19
This article is protected by copyright. All rights reserved.
Independent variable. CEO narcissism. Previous studies have used proxies of
narcissism based on information available in the public domain (e.g., 10-K annual reports,
ExecuComp, Chatterjee and Hambrick, 2007). It is important to note that all measurement
approaches are imperfect. Information within the public domain is influenced by sources beyond
the CEO such as inputs from the top management team, the board of directors, and other
stakeholders. Several individuals may be involved in preparing shareholder letters or grooming
CEOs for interviews, and further, the content analysis used to develop proxies for CEO
prominence may introduce researcher bias (Krippendorff, 2004). Therefore, we offer a different,
but a complementary test of the CEO-narcissism—Firm-performance-variance relationship using
a previously validated and well established psychometric scale.
Narcissism was assessed using a continuous measure based on the 16-item measure (NPI-
16) proposed and validated by Ames and colleagues (2006). The scale consists of 16-paired
statements that require a forced choice between a narcissistic response (=1) and a non-
narcissistic response (=0). This scale has been psychometrically validated (Ames and Kammrath,
2004) and is a shorter version of Raskin and Terry’s (1988) NPI-40. Although the 16-item scale
used in the current study has been validated only recently, the longer version of this scale (NPI-
40) has found increased support over the years. Samuel and Widiger (2008) provide a review of
the face, content, convergent, divergent and nomological validity of the narcissism scale, and
several other meta-analyses offer evidence of the validity of the NPI-40 scale (Campbell et al.,
2000; Campbell et al., 2002; Campbell et al., 2007; Foster et al., 2003; Twenge and Foster, 2008;
Twenge et al., 2008). As a derivative of NPI-40, the shorter 16-item version proposed by Ames
and colleagues (2006) is rooted within a robust tradition of psychometric-research exploring
narcissism.
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
20
This article is protected by copyright. All rights reserved.
The forced-choice scale includes items such as ‘I really like to be the center of attention
vs. It makes me uncomfortable to be the center of attention,’ and ‘I think I am a special person
vs. I am no better or no worse than most people.’ The respondent is asked to choose one of the
two statements in each scale item. For example, is the respondent chooses ‘I really like to be the
center of attention’ it is coded as 1, if the respondent chooses ‘It makes me uncomfortable to be
the center of attention’ then it is coded as 0. We list scores of either 1 or 0 for each item in Table
A.I. The final narcissism score, based on Ames et al. (2006), is the sum of the scores divided by
16. The measure of narcissism is the average of all of the forced choice responses. The final 16-
item scale measure is continuous, ranging from 0 to 1, where ‘1’ represents the highest level of
narcissism and ‘0,’ represents the lowest level of narcissism.
The recommended cut-off values for item loadings, reliability, average variance
exteracted (AVE), constrained-unconstrained models, and confirmatory factor analysis (CFA)
are listed in Table I. Significant item loadings indicate stronger evidence for a relationship
among the observed indicators and their respective latent factors. In addition to Cronbach's
alpha, we also use two other measures of reliability—Raykov's ? and ordinal reliability alpha.
For discriminant validity we use AVE, and constrained-unconstrained models as suggested by
Venkatraman (1989b).
--------Insert Table I about here ---------
The lowest item loading for the narcissism measure was 0.742 (t-value = 8.337) and the
highest item loading was 0.869 (t-value=8.355). The scale reliability for narcissism was above
the recommended cut-off of 0.7 (?=0.782; ?=0.783; ordinal reliability = 0.785). The AVE
(=0.653) was above the recommended cut-off, and the change in chi-square was significant for
constrained-unconstrained models between narcissism and EO, and EO’s sub-scales. Finally,
A
c
c
e
p
t
e
d
A
r
t
i
c
l
e
21
This article is protected by copyright. All rights reserved.
confirmatory factor analysis (CFA) in Table I, shows acceptable fit (?
2
(df)=125.348 (103);
CFI=0.942, TLI=0.937; RMSEA=0.065; SRMR=0.031).
Mediating variable. Entrepreneurial orientation. To measure EO we use the widely
adopted scale developed by Covin and Slevin (1989). The three components of EO are:
innovativeness (3 items), proactiveness (3 items), and risk taking (3 items). Sample items include
“My business places a strong emphasis on: Tried and tested practices, equipment, and products
or services vs. Innovation, technological leadership and R&D” [innovativeness], “My business
typically Responds to initiatives my competitors initiate vs. Initiates action to which my
competitors then respond” [proactiveness], “My business is inclined toward low risk projects
with normal rates of return vs. High risk projects with a chance of very high returns” [risk-
taking]. As shown in Table A.I. (Appendix), all item loadings were significant. The lowest item
loading for EO was for EO-risk-taking (0.778; t-value=5.328) and highest loading was for EO-
Innovativeness (0.882; t-value=7.875). The scale reliability for EO was above the recommended
cut-off of 0.7 (?=0.836; ?=0.839; ordinal reliability = 0.839). Table A.II (Appendix) additionally
confirms a robust second-order fit for the EO measure. Furthermore, as listed in Table A.I, AVE
for EO (=0.745) was above the recommended cut-off of 0.5. Constrained-unconstrained analysis
listed in Table 1 also shows that the lowest change in chi-square was 8.505 (1), p