S AGILE IN THE S CEO T TMT C AND ORGANIZATIONAL...
Transcript of S AGILE IN THE S CEO T TMT C AND ORGANIZATIONAL...
Staying Agile in the Saddle
1
STAYING AGILE IN THE SADDLE : CEO TENURE, TMT CHANGE , AND
ORGANIZATIONAL AMBIDEXTERITY
PAUL FERREIRA University of Geneva 40 Blvd. Pont-d’Arve
1205 Geneva, Switzerland Phone +41 22 379 85 86
[email protected] (CORRESPONDING AUTHOR)
SEBASTIAN RAISCH University of Geneva 40 Blvd. Pont-d’Arve
1205 Geneva, Switzerland [email protected]
PATRICIA KLARNER Munich School of Management
University of Munich Ludwigstr. 28
80539 Munich, Germany [email protected]
Staying Agile in the Saddle
2
STAYING AGILE IN THE SADDLE : CEO TENURE, TMT CHANGE ,
AND ORGANIZATIONAL AMBIDEXTERITY
ABSTRACT
This study examines how interrelated CEO and TMT changes affect an organization’s
ability to simultaneously explore and exploit. Drawing upon the upper echelons perspective, we
argue for a curvilinear CEO tenure-ambidexterity relationship. We then propose that the timing
of TMT change critically affects this relationship. While TMT change negatively impacts
ambidexterity in the early stages of CEO tenure, it has a positive relationship to ambidexterity in
the later stages of CEO tenure. We empirically test and find evidence for our arguments based on
a preliminary longitudinal sample of 30 European insurance companies between 2000 and 2011.
Our main contribution to the ambidexterity literature is the development of a dynamic model
revealing the interrelated effects of CEO and TMT changes on firm ambidexterity.
Keywords:
Ambidexterity; chief executive officers; top management team dynamics; panel data analysis
Staying Agile in the Saddle
3
INTRODUCTION
“Because once you decide to become a blackBerry user, you kind of stay there for life, and
let’s not be too penny-wise, pound-foolish when we do get very good absolute margin. ” (Co-
CEO James Balsillie, April 2008)
“We are on the right track to put these last two quarters behind us." (CEO Richard Fuld,
September 10, 2008)
Both Balsillie and Fuld experienced at Research in Motion and Lehman Brothers
respectively, a time when both firms were considered highly innovative or even pioneers and
outperforming the competition. Both CEOs, who by the time of their (forced) turnover, had been
characterized as long-tenured CEOs having grown “stale in the saddle”, were running firms
averse to innovation, being overwhelmed by the competition, and showing strong performance
decline. The pattern is so common (e.g., Kodak, the Edison Company, DEC) that it seems worth
investigating how can CEOs stay “agile in the saddle” – managing their daily operations, but also
continuing to explore new products, markets, and capabilities?
Organizational scholars increasingly apply the concept of ambidexterity – the human
ability to use both hands with equal skill – as a metaphor for an organization’s ability to
simultaneously explore and exploit (e.g., Gibson & Birkinshaw, 2004; O’Reilly & Tushman,
2008). To become ambidextrous, organizations have to reconcile internal tensions and
conflicting demands in their task environments. Research has shown that the top management
team (TMT) plays an important role in addressing these tensions and conflicts (e.g., Lubatkin,
Simsek, Ling, & Veiga, 2006; Smith & Tushman, 2005). For example, scholars have
documented the role of TMT composition (Beckman, 2006) and TMT processes (Jansen,
Tempelaar, van den Bosch, & Volberda, 2009) as antecedents to ambidexterity.
Staying Agile in the Saddle
4
While these studies provide rich insights, there are two important dimensions of the
TMT-ambidexterity relationship that have received scarce research attention. First, previous
studies describe the current state of the TMT, but did not capture how changes to the TMT affect
ambidexterity (Raisch & Birkinshaw, 2008). Since the upper echelons perspective shows that
TMT change – defined as changes to the composition of the TMT (Hambrick, 2007) – alters the
senior team’s demography (Williams & O’Reilly, 1998) and decision-making processes
(Hambrick, Cho, & Chen, 1996), analyzing its impact on ambidexterity may be worthwhile.
Second, previous ambidexterity studies generally tended to treat the TMT as a unit of
analysis, which did not allow for a more fine-grained analysis of the interrelated impact of the
chief executive officer (CEO) and other TMT members on ambidexterity (Cao, Simsek, &
Zhang, 2010). This distinction is particularly relevant for the TMT change-ambidexterity
relationship, since it is frequently the CEO who drives TMT change (Helmich & Brown, 1972),
and his or her propensity to undertake such change was found to vary over the tenure cycle
(Kesner & Dalton, 1994). Further research may thus be required into the dynamic interrelations
between CEO change, TMT change, and organizational ambidexterity.
In this study, we draw upon prior work from the upper echelons theory (Hambrick &
Mason, 1984), which explores the dynamics of CEO tenure (Hambrick & Fukutomi, 1991;
Miller, 2001). These studies are particularly insightful for our research objective, as they not
only describe temporal changes in CEO behavior, but also their interrelations with TMT
changes. Drawing upon these theoretical concepts, we first argue for a curvilinear CEO tenure-
ambidexterity relationship. We then introduce TMT change as a moderator of the CEO tenure-
ambidexterity relationship. Our arguments imply that TMT change is not only critical for
sustaining ambidexterity, but that it is the right timing of TMT change in relation to the CEO’s
Staying Agile in the Saddle
5
tenure that gives rise to or hinders ambidexterity. While TMT change negatively affects
ambidexterity in the early stages of CEO tenure, it has a positive relationship to ambidexterity in
the later stages of CEO tenure. We find evidence for our arguments from preliminary panel data
analyses based on a sample of 30 European insurance companies between 2000 and 2011.1
Drawing upon our findings, we make two contributions to the ambidexterity literature.
First, while prior research has stressed the importance of TMT characteristics for ambidexterity
(e.g., Jansen et al., 2009), we provide a dynamic perspective of how the timing of changes to the
TMT composition relate to ambidexterity. Second, while ambidexterity theory has generally
considered the TMT as a unit (e.g., Lubatkin et al., 2006), we show that CEO and TMT have
distinctive, but interrelated effects that need to be orchestrated to enable ambidexterity (see also
Cao et al., 2010). Finally, we discuss our study’s practical implications, especially how CEOs
can “stay agile in the saddle” by strategically timing TMT changes.
THEORETICAL BACKGROUND AND HYPOTHESES
TMT’s Role for Organizational Ambidexterity
There is growing evidence that ambidextrous organizations – those that have the ability to
simultaneously manage exploration and exploitation – outperform others in the long run (Gibson
& Birkinshaw, 2004; He & Wong, 2004; Lubatkin et al., 2006). Exploratory behaviors generate
internal variety, create new markets and products, and involve experimentation (March, 1991;
Rosenkopf & Nerkar, 2001). Conversely, exploitative behaviors decrease variance, ensure
operational efficiency, and involve implementation (Beckman, Haunschild, & Phillips, 2004,
1 The final sample is comprised of 115 insurers and banks and data collection effort is completed. We report findings for a subsample of 30 firms and 181 firm-year observations in this version of the paper and plan to finish our analysis by the end of January 2015.
Staying Agile in the Saddle
6
2004; March, 1991). Scholars have warned that a one-sided focus on exploitation results in
inertia and escalating conservatism (Leonard-Barton, 1992), while too much exploration may
deter firms from gaining returns on their knowledge (Levinthal & March, 1993). To become
ambidextrous, organizations have to manage the contradictory forces of exploitation and
exploration, since both compete for scarce resources and require distinct processes, strategies,
and structures (Benner & Tushman, 2003; Gupta, Smith, & Shalley, 2006).
A prominent research stream within this literature conceptualizes leadership characteristics
as an important antecedent to ambidexterity (for reviews, see Lavie, Stettner, & Tushman, 2010;
Raisch & Birkinshaw, 2008). Cognitive and behavioral inclinations of the senior management
team may affect an organization’s proclivity toward exploratory or exploitative actions (Lavie et
al., 2010). Smith and Tushman (2005) thus noted that the senior team’s abilities to develop
accurate cognitive representations of the opposing activities and engage in paradoxical thinking
are vital for effectively managing exploitation and exploration. Extending this view, Lubatkin et
al., (2006) describes “behavioral integration” – the synchronization of the senior team’s social
and task processes – as an important precursor of ambidexterity. More specifically, the authors
argue that the “level of behavioral integration directly influences how its members deal with the
contradictory knowledge processes that underpin the attainment of an exploitative and
exploratory orientation, such that greater integration enhances the likelihood of jointly pursuing
both” (Lubatkin et al., 2006: 647). Moreover, Jansen et al., (2009) empirically show that senior
team social integration and contingency rewards are enablers for ambidexterity.
While these studies focused on TMT processes, Beckman (2006) shows that the founding
team composition – in particular, members’ prior company affiliations – is an important
antecedent of ambidexterity. Firms whose founding teams had both diverse and common prior
Staying Agile in the Saddle
7
company affiliations demonstrated a higher degree of ambidexterity. Several authors argue that
heterogeneous senior management teams may be better equipped to attend to the contradictions
associated with exploration and exploitation (Boeker, 1997; Virany, Tushman, & Romanelli,
1992). Carmeli and Halevi (2009) found empirical evidence that TMT behavioral complexity is
positively related to ambidexterity. Finally, Cao et al., (2010) show that the CEO’s network
extensiveness (i.e., a system of relationships CEOs have inside and outside their organizations)
positively impacts ambidexterity, particularly when the CEO-TMT interactional interface
enables the TMT to process disparate information.
Despite these rich foundations, ambidexterity scholars have repeatedly criticized the
limited explanatory power of studies assessing TMT characteristics at a given point in time
(Raisch & Birkinshaw, 2008; Simsek, 2009). Ambidexterity has been described as a ‘dynamic
capability’ (O’Reilly & Tushman, 2008) that depends on the senior team’s ability to repeatedly
and intentionally orchestrate firm assets. It is thus important to investigate how changes to the
senior team’s composition and processes affect ambidexterity. Further, scholars have encouraged
studies to investigate how interrelated CEO-TMT effects relate to ambidexterity (Cao et al.,
2010). In this study, we investigate the dynamic interrelations between CEO tenure, TMT
change, and ambidexterity. We draw upon theoretical arguments from the upper echelons
literature on CEO tenure dynamics, which provide rich insights on the temporal dynamics of
CEO behaviors over different tenure phases and their interrelations to TMT changes.
CEO Tenure and Organizational Ambidexterity
A central argument of studies on the dynamics of CEO tenure is the notion of an ‘executive
life cycle’ (Hambrick & Fukutomi, 1991), which implies that CEOs vary in their patterns of
Staying Agile in the Saddle
8
behavior and strategic choices throughout different tenure phases. In the early stages of tenure,
new CEOs operate with limited repertoires, since their understanding of how the environment
functions and how the company should be run are still incomplete (Miller, 2001). As a result,
new CEOs strive to learn about their teams, firms, markets, and environments, a process that can
be expected to go hand in hand with exploration (Hambrick & Fukotomi, 1991). For example,
Miller and Shamsie (2001) showed, in a longitudinal study of the movie industry, that studio
chiefs were experimenting with their product lines in their early tenures to learn about their jobs.
Consistent with this logic, researchers also found evidence that related proxy variables such as
R&D spending (Barker & Mueller, 2002) and firm invention (Wu, Levitas, & Priem, 2005) are
greater at firms with short-tenured CEOs if compared to those with long-tenured CEOs. It is
understood that CEO behavior varies across different contingencies such as CEO origin, CEO
predecessor retained as chairman (Quigley & Hambrick, 2012), but we can expect that new
CEOs generally tend to engage in increased exploration activities.
If short-tenured CEOs tend to prioritize explorative activities, one can expect that they
engage less in exploitative activities during the same period. The reason is that exploitation and
exploration rely upon learning activities that are located at two ends of a continuum (Levinthal &
March, 1993), which makes it difficult for individuals to engage in extensive exploitation and
exploration at the same time (Gupta et al., 2006: 695). This view also derives support from
Ocasio’s (1997) work, which shows that executives’ attentional orientation serves to filter and
distort the abundant information in strategic decision-making processes. For example, a CEO
with an attentional orientation towards exploration would scan the firm’s environment for
exploratory opportunities, focus on comprehending and making sense of explorative demands,
and foster the generation of exploratory initiatives. Following Ocasio’s (1997) theory, such a
Staying Agile in the Saddle
9
CEO would be heavily limited in his or her ability to simultaneously dedicate attention to
exploitation activities. We thus argue that short-tenured CEOs tend to prioritize exploratory
activities, while paying less attention to exploitative activities.
Conversely, with increasing tenure, CEOs tend to select and retain a set of strategies that
define how their firm should be configured (Miller, 2001). Tenure can be equated with the
validation of the CEO’s past strategic approach, which in turn prevents him or her from
experimenting with other ways of doing things (Prahalad & Bettis, 1986). Consequently, in the
later stages of their tenure, CEOs are likely to draw strongly upon past experiences (Hambrick &
Fukutomi, 1991), which can be related to exploitative activities. As exploitation often leads to
early success (Levinthal & March, 1993), the exploitation activities are further reinforced, which
leads to an increasing path dependency on exploitation and restricts the search for new
competencies (Lavie & Rosenkopf, 2006; Lavie et al., 2010).
Empirical studies provide supportive evidence for these theoretical arguments. For
example, Miller and Shamsie (2001) found that movie industry executives with increasing tenure
progressively shift their focus towards a set of proven managerial routines. Their growing
knowledge and security renders them less pressed to learn by experimenting, but rather to resort
to increasing exploitation (Levitt & March, 1988). Long-tenured CEOs thus tend to become
complacent and no longer follow market trends; they stick to an obsolete formula and apply it
well beyond its utility, which adversely affects firm performance. They get more internally
focused, which facilitates exploitation while driving out exploration (Hambrick, Finkelstein, &
Mooney, 2005; O’Reilly & Tushman, 2008). Consistent with these arguments, Barker and
Mueller (2002) found that R&D spending decreases with CEO tenure and Wu et al., (2005)
Staying Agile in the Saddle
10
found a curvilinear relationship between CEO tenure and firm invention. We thus assume that
long-tenured CEOs generally prioritize exploitative over exploratory activities.
Since ambidexterity requires simultaneous exploration and exploitation (Gupta et al., 2006;
He & Wong, 2004), neither retaining long-tenured CEOs – which can lead to downward spirals
caused by one-sided exploitation (Hambrick & Fukutomi, 1991) – nor frequently replacing them
– which implies the risk of getting caught in unrewarding exploration cycles (Volberda & Lewin,
2003) – can be expected to lead to ambidexterity. However, these developments are likely to be
continuous (Miller, 2001): exploration gradually decreases, while exploitation gradually
increases with CEO tenure. We thus expect that intermediate CEO tenure is related to the
strongest orientation towards organizational ambidexterity.
Hypothesis 1: CEO tenure has an inverted U-shaped relationship to organizational ambidexterity.
CEO Tenure, TMT Change, and Organizational Ambidexterity
While the curvilinear CEO tenure-ambidexterity relationship highlights the central role of
the CEO, it may be an incomplete representation of how he or she affects ambidexterity. Upper
echelons scholars have stressed the added value of extending the focus to the TMT when
exploring the dynamics of CEO tenure (e.g., Shen & Cannella, 2002). For example, Tushman
and Rosenkopf (1996) argue that TMT changes have “more profound effects on organization
outcomes than simple CEO succession” (1996: 950). Williams and O’Reilly (1998: 99) noted
that “once turnover does occur, the heterogeneity of the group will, by definition, change.” Since
TMT characteristics were found to affect ambidexterity (e.g., Beckman, 2006), we can expect
that changes to TMT composition may as well play an important role for ambidexterity. In this
section, we thus introduce TMT change as a moderator of the CEO tenure-ambidexterity
Staying Agile in the Saddle
11
relationship. We argue that the timing of TMT change in relationship to the CEO’s tenure is
critical for ambidexterity.
Prior research on CEO tenure dynamics and the related literature on CEO succession claim
that TMT changes often occur in close alignment with CEO changes. If new CEOs take office,
they are likely to make immediate changes to the TMT composition to clear “executive
deadwood” and facilitate strategic reorientation (Kesner & Dalton, 1994). The need for TMT
changes may be explained by the fact that CEO changes by themselves fail to introduce
sufficient new experience and knowledge to radically alter established understandings and
entrenched activity patterns, and power positions (Shen & Cannella, 2002; Virany et al., 1992,
Wiersema & Bantel, 1993). For example, Ancona and Chong (1996: 276) argue for a combined
CEO and TMT change, suggesting that only if the “clocks at both levels are reset to zero at
once,” this provides a powerful mechanism to drive radical strategic change.
As argued above, CEOs may exhibit high levels of exploration and low levels of
exploitation in their early stages of tenure. TMT changes in these early stages can be expected to
further reinforce the CEO’s inclination towards exploration. The newly appointed members
reduce the team’s average tenure, which has been related to increased experimentation and
exploration activities (Miller & Shamsie, 2001). Furthermore, TMT change increases the
heterogeneity of the senior team’s experiences (Williams & O’Reilly, 1998). Such heterogeneity
can create task conflicts that force senior team members to share more varied information (Jehn,
Chadwick, & Thatcher, 1997) and engage in greater experimentation (Grinyer & McKiernan,
1990). Finally, the disruptive effects of TMT change are likely to encourage the executive team
to challenge established routines and experiment with novel approaches (Lavie et al., 2010).
From an ambidexterity perspective, TMT change in the early stages of CEO tenure may thus lead
Staying Agile in the Saddle
12
to a one-sided orientation of both the CEO and the TMT towards exploration, which negatively
affects ambidexterity.
Conversely, refraining from discretionary TMT changes – those that the CEO can
influence - in the early stages of CEO tenure may have positive effects on ambidexterity.
Maintaining stability in TMT composition increases the TMT’s average tenure. Long-tenured
TMTs have accumulated firm-specific human capital and develop common sets of experiences,
similar attitudes, and shared problem-solving behaviors (Allen & Cohen, 1969). In other words,
TMTs that continue with relatively few changes – some may be unavoidable due to causes of
force majeure such as death or criminal charges – tend to develop organizational routines that
they use to identify, frame and make convergent types of strategic decisions (Prahalad & Bettis,
1986). Routines are self-reinforcing (March & Simon, 1958), they give rise to a “dominant
logic” (Prahalad & Bettis, 1986: 485) the longer the TMT members have worked together.
Ambidexterity scholars have related the emergence of such organizational routines to
exploitative activities (Baum, Li & Usher, 2000). Since exploitation tends to drive out
exploration (Levinthal & March, 1993), TMT members who opt for proven responses to
environmental changes refrain from exploring new capabilities.
From an ambidexterity lens, the combination of a new CEO with a long-tenured TMT
should thus allow for balancing exploration and exploitation “across domains” (Gupta et al.,
2006). Short-tenured CEOs with a natural tendency towards exploration are paired with long-
tenured TMTs with an inclination towards exploitation. However, balancing across these
competing activities entails the inherent risk that the CEO and the TMT may not know how to
coordinate the activities and manage the task-oriented conflict induced by such an arrangement
of roles and responsibilities. In our study, we follow Smith and Tushman’s (2005) previous work
Staying Agile in the Saddle
13
suggesting that “team design” reflected in features such as “distinct roles, goals and rewards” is a
decisive antecedent in shaping the CEO and TMT’s ability to engage in paradoxical cognitive
processes and manage strategic contradictions. Accordingly, it is precisely the mismatch of their
tenure that gives rise to cognitive frames and processes that allow both the CEO and the TMT to
effectively handle potential tensions (Smith & Lewis, 2011). We therefore argue that low levels
of TMT change in the early stages of CEO tenure positively affect ambidexterity.
If a CEO refrains from discretionary TMT change, his or her organization eventually
reaches a stage characterized by the simultaneous presence of a long-tenured CEO and a long-
tenured TMT. As argued above, long-tenured CEOs face the increasing risk of getting overly
committed to increasingly outdated paradigms (Miller & Shamsie, 2001), and long-tenured
TMTs tend to exhibit a greater commitment to the status quo and the culture and norms of the
organization (Bantel & Jackson, 1989). Consequently, the combined presence of a long-tenured
CEO and TMT has the potential to decrease task-oriented conflict and the related exploratory
activities. As the orientation at both levels is directed towards exploitation, which can be
expected to mutually reinforce one another, the attentional orientation to exploration vanishes.
Refraining from discretionary TMT changes in the advanced stages of CEO tenure may thus
negatively affect ambidexterity.
Conversely, TMT change in the advanced stages of CEO tenure can prompt TMT members
to consider fresh insights and discuss alternative solutions (Amason & Sapienza, 1997).
Furthermore, new TMT members increase tenure heterogeneity, which has been related to
frictions and relational conflicts that push the TMT towards greater risk-taking and
experimentation (Hambrick et al., 1996). TMT change contributes new skills and shakes up
engrained group processes, which may shift previous TMT attentional orientations toward new
Staying Agile in the Saddle
14
ones. For example, Cho and Hambrick (2006) have shown in the context of airline deregulation
how firms that changed their TMT composition – notably their TMT industry tenure – have
experienced a shift in managerial attention toward an entrepreneurial orientation (relative to an
engineering orientation), which generally goes hand in hand with exploration activities.
From an ambidexterity perspective, we thus expect that new TMT members, who drive the
TMT towards greater exploration, challenge or revitalize long-tenured CEOs whose natural
inclination is towards exploitation. The tenure mismatch between the CEO and the TMT leads to
contradictory orientations and task-related conflicts, which are essential for stimulating
ambidextrous behavior (Smith & Lewis, 2011). We know from Arrow and McGrath’s (1995)
work on group dynamics that the more past continuity teams have experienced, the stronger the
immediate effects of changes to team composition. We thus argue that opposed to what we had
argued above for the early stages, TMT changes in the later stages of CEO tenure positively
affect ambidexterity.
Hypothesis 2: TMT change moderates the CEO tenure-ambidexterity relationship in such a way that TMT change combined with short-tenured CEOs decreases ambidexterity, while TMT change combined with long-tenured CEOs increases ambidexterity.
RESEARCH METHODOLOGY
Research Setting and Data Collection
To test our hypotheses, we composed a sample of insurance firms from the Stoxx Europe
TMI Finance Index for the period from 2000 to 2011. Due to the considerable industry dynamics
during our observation period (e.g., two global financial and economic crises and subsequent
rebounds; important regulatory changes), firms in this industry had to both exploit their
incumbent business and explore future growth platforms. The European financial services
Staying Agile in the Saddle
15
industry is thus a suitable context for our study. In addition, a single industry study allows us to
ensure comparability among firms (Klarner & Raisch, 2013).
In total, there were 68 financial services companies in the Stoxx Europe TMI Finance
Index. While 47 of these companies made up the index in 2000, the remaining 21 companies
were added to the index between 2000 and 2011. During our analysis period, the financial
services industry has experienced multiple mergers and acquisitions, as well as several delistings.
To account for these changes, we constructed a detailed list indicating the “family tree” of each
of the firms in our sample for the time period between 2000 and 2011. Due to insufficient data
availability (e.g., unavailability of annual reports or missing values in databases), we had to
discard 12 firms, leading to a final sample of 56 firms and 392 firm-year observations.2 The
number of firm-year observations was reduced due to missing values resulting from mergers,
acquisitions, and delistings. Since we included several lagged variables, we traced all firms back
to 2000.
We used different sources to collect our data. First, we used content analysis of press
releases on corporate websites to collect data on firm ambidexterity. Second, data on the CEO
and TMT members was obtained by studying the TMT composition of each firm and year using
annual reports. Third, firm-level controls were obtained through the Thomson Analytics and
Capital IQ databases.
Dependent Variable
We operationalized exploration and exploitation in accordance with Uotila, Maula, Keil,
and Zahra (2009) by quantifying the yearly relative amount of a firm’s exploitative orientation
2 The final sample is comprised of 115 insurers and banks and data collection effort is completed. We report findings for a subsample of 30 firms and 181 firm-year observations in this version of the paper and plan to finish our analysis by the end of January 2015.
Staying Agile in the Saddle
16
and explorative orientation. We collected textual data in form of press releases available on the
corporate websites (Duriau, Reger, & Ndofor, 2000). Each document was analyzed according to
the wordlist provided by March (1991). Organizational ambidexterity is operationalized as the
interaction between exploration and exploitation (Gibson & Birkinshaw, 2004; He & Wong,
2004). We measure a firm’s ambidexterity in year t as the interaction of exploration i,t and
exploitation i,t. This measure is calculated by the arithmetic average of all the exploration-
exploitation interactions for the focal (t) period, taking a minimum value of 0 (for firms that have
a one-sided focus on either exploration or exploitation) and a maximum of .25 (for firms that
show maximum values of both exploration and exploitation).
Independent Variables
CEO tenure refers to the total number of years a specific individual had held the CEO
position within a company (Shen & Canella, 2002). TMT change was measured for all
individuals that reported directly to the CEO (Tushman & Rosenkopf, 1996). For each year and
firm, we used annual reports to identify all the executives who were not on the team in the
previous year (the entering executives), and the executives who were listed in the previous year,
but not in the current year (the exiting executives). TMT change was measured as the number of
entering and exiting executives in a given year. We mean centered CEO tenure and TMT change
prior to testing interaction effects to avoid potential multicollinearity (Wu et al., 2005).
Control Variables
We included five general classes of control variables to mitigate potential misinterpretations of
our findings: Firm performance (Return on Equity), top management team size (measured by
counting all top management team members in a given year), firm age (years since founding),
Staying Agile in the Saddle
17
and prior firm ambidexterity. We also controlled for firm size (logarithm of the number of
employees) but had to drop this variable due to multicollinearity concerns. We lagged all
independent and control variables by one year to avoid simultaneity bias (Weng & Lin,
forthcoming).
Data Analysis
Since our dataset consists of multiple observations for each firm, we tested our hypotheses
using panel data analysis. A Wooldridge test (Wooldridge, 2010) revealed autocorrelation in our
data (p<0.001). In addition, following Greene (2000), we had to reject the null hypothesis of
homoscedasticity (p<0.001). We therefore used Prais-Winston regression with a panel-specific
autoregressive disturbance structure (the “xtpcse” command in STATA 11), specifying that there
is first-order autocorrelation and that the coefficient of the AR(1) process is specific to each
panel (“psar1” option) (Marquis & Huang, 2009). We also chose the “pairwise” option in
STATA, which includes all available observations with non-missing pairs (StataCorp, 2009).
A possible concern with our analysis and findings is endogeneity, since CEO tenure – the
independent variable – may correlate with unobservable factors in the error term (Wooldridge,
2010). In order to address this concern, we tested a two-stage instrumental variable model
(2SLS) using the “xtivreg2” command in STATA. We used CEO origin, CEO duality, and mean
CEO tenure in the industry as instrumental variables and designated CEO tenure as endogenous
variable. Prior research has found that CEO origin relates to CEO tenure (Karaevli, 2007; Zhang,
2008). In addition, CEO duality can influence the tenure of a CEO (Goyal & Park, 2002; Quigley
& Hambrick, 2012). Finally, a firm’s decision to maintain or replace the CEO may be influenced
by competitors’ tendency to keep their CEOs in office or replace them (Karaevli & Zajac, 2013;
Staying Agile in the Saddle
18
Zhang & Rajagopalan, 2003). Results from the 2SLS model indicated the consistency of OLS (p
= 0.4546).
We also tested each model for potential multicollinearity by calculating the variance
inflation factors (VIF). VIFs ranged between 1.23 and 7.04, which was below the suggested
cutoff value and thus not problematic (Neter, Wasserman, & Kutner, 1985).
RESULTS
Table 1 presents the descriptive statistics and the correlation matrix. First, we examined
Hypothesis 1 regarding the proposed inverted U-shaped relationship between CEO tenure and
organizational ambidexterity. Two models were estimated, as shown in Table 2. Model 1
included only the control variables and the main effects, and model 2 also included the squared
CEO tenure term. Both models were significant. The results in model 2 support Hypothesis 1
since the coefficient of CEO tenure is positive and significant (b = 0.0018, p < 0.05) and the
coefficient of CEO tenure squared is negative and significant (b = -0.0001, p < 0.01).
In Hypothesis 2, we argued that top management team change would moderate the
curvilinear relationship between CEO tenure and organizational ambidexterity. Model 3 in Table
2 included the control variables, main effects, CEO tenure squared, and the interaction of CEO
tenure and TMT change. Model 4 included all these terms, as well as the interaction term of
CEO tenure squared and TMT change. Both models were significant. The coefficient of the
interaction of CEO tenure squared and TMT change is positive and significant (b = 0.0001, p <
0.10) in model 4, thus supporting Hypothesis 2.
------------------------------------------------- Please insert Tables 1 and 2 about here
-------------------------------------------------
Staying Agile in the Saddle
19
DISCUSSION
In this study, we revisited the emerging research stream on the role of senior management
in attending to contradictory demands of exploration and exploitation (e.g., Cao et al., 2010;
Lubatkin et al., 2006; Smith & Tushman, 2005). Contrary to prior studies, our emphasis was on
the dynamics of the senior management-ambidexterity relationship. In essence, we investigated
how TMT changes at different stages of CEO tenure contribute to or hinder ambidexterity. We
find that the CEO tenure-organizational ambidexterity relationship is a dynamic one with TMT
change having varying effects on ambidexterity at different stages of CEO tenure. Drawing upon
these findings, we continue to discuss our paper’s contributions to the ambidexterity literature, as
well as its managerial implications, limitations, and avenues for future research.
Managing Ambidexterity: A Dynamic Perspective
Prior research has stressed the importance of TMT characteristics and processes for
organizational ambidexterity (e.g., Beckman, 2006; Lubatkin et al., 2006). In a recent review of
the ambidexterity literature, Lavie et al., (2010) criticize the static character of previous studies
and argue that shifting our research focus from the rather static notion of balance towards the
process of balancing over time could generate important new insights. In this paper, we respond
to this call for further research by providing a dynamic perspective of how senior team dynamics
relate to ambidexterity. This novel perspective allows us to derive several insights on the
dynamics of managing ambidexterity within the corporate team.
Staying Agile in the Saddle
20
First, our results contribute a new dimension to ambidexterity research, revealing the thus
far neglected role that the timing of TMT change can play in successfully managing exploration-
exploitation tensions. Following executive life cycle theory arguments (Hambrick & Fukotomi,
1991; Miller, 2001), we show that ambidexterity varies with CEO tenure and that TMT change
critically affects this relationship. While we provide further evidence that long-tenured CEOs
generally face the increasing risk of growing “stale in the saddle” (Miller, 1991), which may
cause organizations to get caught in a “success trap” that exacerbates exploitation at the cost of
exploration (Levinthal & March, 1993), we also reveal that the appropriate timing of deliberate
TMT changes may be an important means to break these downward spirals and allow
organizations to foster and sustain ambidexterity. Future research can build upon these
foundations to explore how different paces, frequencies, and rhythms of change (Klarner &
Raisch, 2013) across different organizational levels and dimensions affect ambidexterity.
Second, prior ambidexterity studies increasingly argue that successful organizations do not
just strive for balancing exploration and exploitation, but also dynamically shift their relative
attention between these poles (Boumgarden, Nickerson, & Zenger, 2012; Lavie et al., 2010). The
gradual adaptations allow organizations to ensure continued alignment with changing
environmental conditions (Siggelkow, 2002; Simsek, Heavy, Veiga, & Souder, 2009). While
prior studies describe such adaptations, they did not explore the drivers and capabilities that
enable organizations to enact these shifts in their exploration-exploitation balance (Raisch,
Birkinshaw, Probst, & Tushman, 2009). In our study, we show that TMT change is one potential
driver for organizations to foster “adaptation capabilities” (Smith & Tushman, 2005). Future
research could build upon our findings to explore whether and how organizations can
intentionally and strategically time their TMT changes to shift their exploration-exploitation
Staying Agile in the Saddle
21
balance. Such modification routines (Adler, Goldoftas, & Levine, 1999) support firms’
adaptation to environmental change and could enable them to sustain ambidexterity over time.
Finally, we present an alternative, more positive ambidexterity perspective of CEO tenure.
While upper echelons scholars traditionally accentuate the negative effects of CEO tenure such
as increasing path dependencies and reduced behavioral repertoires (Hambrick & Fukutomi,
1991; Miller, 1990), we argue that CEOs accumulate valuable organizational knowledge over
time. If combined with deliberate TMT changes, which introduce new knowledge for exploration
and challenge the status quo, long-tenured CEO’s accumulated organizational knowledge
represents the complement that ensures simultaneous exploitation.
Our theoretical arguments are contingent upon the quality and richness of exchanges
between the CEO and his or her senior management team. Future research should explore how
the CEO-TMT interactional interface (Cao et al., 2010), which includes communication richness
(Smith et al., 1994), functional complementarity (Taylor & Greve, 2006) and power
decentralization (Finkelstein, 1992), enables the executive team to consider the different
knowledge bases when addressing exploration-exploitation tensions.
Managing Ambidexterity: A Multi-Level Perspective
With the notable exception of Cao et al. (2011), prior ambidexterity research has
considered the TMT as a unit of analysis (Lubatkin et al., 2006). Scholars criticized this
approach since it does not allow for more fine-grained analysis of the interrelated effects of the
CEO and other TMT members on ambidexterity (Cao et al., 2001; Raisch et al., 2009). In this
paper, we contribute to the emerging multi-level perspective to managing ambidexterity at the
Staying Agile in the Saddle
22
corporate level by providing several new insights into the dynamics of the CEO-TMT
interrelations.
First, we show that CEO and TMT changes are interrelated and reveal how changes to
these two distinct types of actors have to be orchestrated to enable ambidexterity. While our
findings indicate that CEOs by themselves may be able to temporarily hold the exploitation-
exploration tension in mid-tenure, they also show that CEOs must rely upon the TMT’s support
to sustain ambidexterity over time. These insights confirm prior arguments that exploitation and
exploration are difficult to reconcile “within a single domain” (Gupta et al., 2006: 697) and
illustrate how these tensions may be resolved across domains (Raisch & Birkinshaw, 2008). This
requires future work to adopt a contingency perspective when exploring TMT characteristics and
processes, since their outcomes may vary dependent on the CEO’s background, behaviors, and
cognitions.
Second, an emergent perspective in recent ambidexterity research suggests domain
separation as an alternative means to address exploration-exploitation tensions (Lavie &
Rosenkopf, 2006; Lavie et al., 2010). While this line of work has so far been limited to inter-
organizational domains, we provide theoretical arguments and empirical evidence for its
usefulness in a top management team context. Domain separation in this context means that the
CEO and his or her senior team take different roles with regard to exploration and exploitation at
different stages of CEO tenure. Integration across domains is ensured through collaborative
behavior and joint decision-making within the senior team. By separating and investigating the
distinctive roles of the CEO and the senior management team, we are able to propose an
alternative approach to balancing exploration and exploitation at the corporate level. Future
Staying Agile in the Saddle
23
research should take a micro-level approach to explore the differentiation and integration
processes at the CEO-TMT interactional interface.
Finally, our work has more general implications for the upper echelons theory. We
contribute a novel ambidexterity perspective to the executive succession literature (e.g.,
Finkelstein & Hambrick, 1990; Karaevli, 2007; Tushman & Rosenkopf, 1996). A core argument
in this literature is that CEO change drives first-order learning, but concurrent and extensive
TMT change is required to realize more fundamental, second-order learning (Dalton & Keener,
1985; Virany et al., 1992). Consequently, upper echelons scholars suggest that firms tend to
align CEO change with discretionary TMT change (Shen & Cannella, 2002). While this solution
may be appropriate for extensive exploration, the ambidexterity literature warns firms against
overly focusing on one-sided exploration (Gupta et al., 2006; Levinthal & March, 1993). More
importantly, the ambidexterity literature found strong empirical evidence that firms that balance
exploration and exploitation generally outperform those with a one-sided alignment in the long
run (for an overview, see Junni, Sarala, Taras, & Tarba, 2013). Firms that strive for
ambidexterity may benefit less from aligning CEO change and TMT change. Such alignment
may even be harmful since it implies the risk of reinforcing a one-sided orientation. In this paper,
we lay the foundations for an alternative perspective of executive succession, which accentuates
the benefits of decoupling CEO and TMT change to foster the senior team’s ability to deal with
the contradictory demands of exploration and exploitation.
Practical Implications
Staying Agile in the Saddle
24
The results of our study have particularly important implications for top managers that
attempt to achieve ambidexterity within their organizations. “staying agile in the saddle” is
particularly challenging for CEOs since it forces them to address multiple challenges
simultaneously: First, they need to host and harmonize the conflicting requirements of
exploration and exploitation within their senior management teams. Second, they need to
withstand the temptations of falling into the trap of increasing alignment with one or the other
objective in different stages of their tenure. Finally, they need to identify “early warning signs”
that signal that “the time for a course correction has come” (Probst & Raisch, 2005: 100), which
allows them to strategically time their discretionary TMT changes to maintain ambidexterity.
Since this is a quite demanding set of expectations towards CEOs, we suggest that boards
of directors play a more active and supportive role in managing CEO and TMT changes. This
includes boards’ development of a long-term succession plan for TMT members in addition to
the CEO. A promising alternative to systematically replacing CEOs at intermediate tenures
(Fama, 1980) may be to only renew experienced CEOs’ contracts under the condition of
refreshing their senior management teams. The board of directors could even play an active role
in suggesting candidates that ensure increasing diversity in experience (Zhang, 2008). Similarly,
boards of directors could advice newly appointed CEOs – particularly if they have been recruited
from outside the firm – to retain some of the previous top management team’s members. This
would ensure adequate tenure diversity right from the start and should allow the TMT to
immediately engage in exploitative activities alongside the exploratory activities.
Limitations and Future Research
Staying Agile in the Saddle
25
As most empirical studies, our research has its limitations. First, our findings are
preliminary and rely upon a first data analysis effort based on a subsample of 30 firms. We plan
to complete our data collection in the spring of 2014 and to present findings for the full sample
during the AoM Meetings in August 2014. At that point, we also plan to complement our
preliminary analyses with additional robustness checks and post-how analyses to ensure the
validity and reliability of our findings.
Second, reliance on a single industry implies that the findings cannot be simply generalized
to other industry settings. For example, industry characteristics may effect whether and to what
extent different TMT characteristics and processes matter for organizational ambidexterity
(Lavie et al., 2010). Furthermore, ambidexterity may not be equally beneficial to firms across
industries (Junni et al., 2013; Simsek et al., 2009). Future research should thus test the concept
of interrelated CEO and TMT changes in a broader set of industries.
Third, future research might also examine how CEOs, TMTs, and boards of directors can
implement and sustain organizational ambidexterity through the strategic timing of CEO and
TMT changes. While we introduced the concept of “staying agile in the saddle” and provided
evidence of its importance, our discussion of potential solutions to implement it within
organizations is limited. It remains to be explored how senior managers and directors may
accomplish it. We provided first ideas and pointed towards potentially promising avenues for
future research.
CONCLUSION
Staying Agile in the Saddle
26
Prior research on the TMT-ambidexterity relationship has primarily described how the
current state of the TMT affects ambidexterity, but did not capture how changes to the TMT alter
ambidexterity over time. This study explores how TMT changes at different stages of CEO
tenure contribute to or hinder ambidexterity. Our findings show that the CEO tenure-
organizational ambidexterity relationship is a dynamic one with TMT changes distinctively
affecting ambidexterity along the different stages of CEO tenure. Owing to the evidence that
ambidextrous organizations outperform others in the long run, as well as the crucial role played
by senior managers in this strategic endeavor, the analysis of how interrelated CEO and TMT
changes influence ambidexterity is particularly relevant. We hope that our findings pave the way
for future studies in this field towards a more dynamic conception of organizational
ambidexterity and its antecedents.
Staying Agile in the Saddle
27
REFERENCES
Adler, P. S., Goldoftas, B., & Levine, D. I. 1999. Flexibility Versus Efficiency? A Case Study of Model
Changeovers in the Toyota Production System. Organization Science, 10(1): 43-68.
Allen, T. J., & Cohen, S. I. 1969. Information flow in research and development laboratories. Administrative
Science Quarterly, 12-19.
Ancona, D. G., & Chong, C. L. 1996. Entrainment: Pace, cycle, and rhythm in organizational behavior. In B. M.
Staw & L. L. Cummings (Eds.), Research in organizational behavior, vol. 18: 251–284. Greenwich, CT: JAI
Press.
Amason, A. C., & Sapienza, H. C. 1997. The effects of top management team size and interaction norms on
cognitive and affective conflict. Journal of Management, 23: 495–516.
Arrow, H., & McGrath, J. E. 1995. Membership dynamics in groups at work: A theoretical framework. In B. M.
Staw & L. L. Cummings (Eds.) Research in organizational behavior, vol 17: 373-411. Greenwich, CT: JAI
Press.
Bantel, K., & Jackson, S. 1989. Top management and innovation in banking: Does the composition of the top team
make a difference? Strategic Management Journal, 10: 107–124.
Barker, V. L., & Mueller, G. C. 2002. CEO characteristics and firm R&D spending. Management Science, 48(6):
782– 801.
Baum, J. A. C., Li, S. X., & Usher, J. M. 2000. Making the Next Move: How Experiential and Vicarious Learning
Shape the Locations of Chains' Acquisitions. Administrative Science Quarterly, 45(4): 766-801.
Beckman, C. M. 2006. The Influence of Founding Team Company Affiliations on Firm Behavior. Academy of
Management Journal, 49(4): 741-758.
Beckman, C. M., Haunschild, P. R., & Phillips, D. J. 2004. Friends or strangers? Firm-specific uncertainty, market
uncertainty, and network partner selection. Organization science, 15(3): 259-275.
Benner, M. J., & Tushman, M. L. 2003. Exploitation, Exploration, and Process Management: The Productivity
Dilemma Revisited. Academy of Management Review, 28(2): 238-256.
Boeker, W. 1997. Strategic change: The influence of managerial characteristics and organizational growth. Academy
of Management Journal, 40(1): 152-170.
Staying Agile in the Saddle
28
Boumgarden, P., Nickerson, J. A., & Zenger, T. R. 2012. Sailing into the Wind: Exploring the Relationship among
Ambidexterity, Vacilation, and Organizational Performance. Strategic Management Journal, 33: 587-610.
Cao, Q., Simsek Z., & Zhang H. 2010. Modelling the Joint Impact of the CEO and the TMT on Organizational
Ambidexterity. Journal of Management Studies, 47: 1272-1296.
Carmeli, A., & Halevi, M. Y. 2009. How top management team behavioral integration and behavioral complexity
enable organizational ambidexterity: The moderating role of contextual ambidexterity. The Leadership
Quarterly, 20(2): 207-218.
Cho, T. S., & Hambrick, D. 2006. Attention patterns as mediators between top management team characteristics and
strategic change: The case of airline deregulation. Organization Science, 17: 453-469.
Duriau, V. J., Reger, R. K., & Ndofor, H. 2000. Content Analysis of Firms' Web Sites: Methodological Foundations,
Software Solutions, and Implementation Issues, Academy of Management Conference. Toronto, Canada.
Fama, E. F. 1980. Agency problems and the theory of the firm. Journal of Political Economy, 88: 288-307.
Finkelstein, S. 1992. Power in top management teams: Dimensions, measurement, and validation. Academy of
Management Journal, 35(3): 505-538.
Gibson, C. B., & Birkinshaw, J. 2004. The Antecedents, Consequences, and Mediating Role of Organizational
Ambidexterity. Academy of Management Journal, 47(2): 209-226.
Goyal, V. K., & Park, C. W. 2002. Board leadership structure and CEO turnover. Journal of Corporate Finance,
8(1): 49-66.
Greene, W. 2000. Econometric analysis. New York: Prentice-Hall. .
Grinyer, P. H., Mayes, D., & McKiernan, P. 1990. The sharpbenders: Achieving a sustained improvement in
performance. Long Range Planning, 23: 116-125.
Gupta, A. K., Smith, K. G., & Shalley, C. E. 2006. The Interplay between Exploration and Exploitation. Academy of
Management Journal, 49(4): 693-706.
Hambrick, D.C. 2007. Upper echelons theory—An update. Academy of Management Review, 32(2): 334–343.
Hambrick, D. C., Cho, T., & Chen, M. 1996. The influence of top management team heterogeneity on firms'
competitive moves. Administrative Science Quarterly, 41: 659-684.
Hambrick, D. C., Finkelstein, S., & Mooney, A. C. 2005. Executive job demands: New insights for explaining
strategic decisions and leader behaviors. Academy of Management Review, 30(3): 472-491.
Staying Agile in the Saddle
29
Hambrick, D. C., & Fukutomi G. 1991. The seasons of a CEO’s tenure. Academy of Management Review, 16: 719–
742.
Hambrick, D. C., & Mason, P. 1984. Upper echelons: The organization as a reflection of its top managers. Academy
of Management Review 9: 193–206.
He, Z. L., & Wong, P. K. 2004. Exploration Vs. Exploitation: An Empirical Test of the Ambidexterity Hypothesis.
Organization Science, 15(4): 481-494.
Helmich, D. L., & Brown, W. B. 1972. Successor type and organizational change in corporate enterprise.
Administrative Science Quarterly, 17: 371-381.
Jansen, J. J. P., Tempelaar, M. P., Van den Bosch, F. A .J., & Volberda, H. W. 2009. Structural differentiation and
ambidexterity: The mediating role of integration mechanisms. Organization Science, 20(4): 797–811.
Jehn, K. A., Chadwick, C., & Thatcher, S. M. B. 1997. To Agree or Not to Agree: The Effects of Value Congruence,
Individual Demographic Dissimilarity, and Conflict on Workgroup Outcomes. International Journal of Conflict
Management, 8(4), 287–306.
Junni, P., Sarala, R. M., Taras, V., & Tarba, S. Y. 2013. Organizational ambidexterity and performance: A meta-
analysis. The Academy of Management Perspectives, 27(4): 299-312.
Karaevli, A. 2007. Performance consequences of new CEO ‘Outsiderness: Moderating effects of pre‐and post‐
succession contexts. Strategic Management Journal, 28(7): 681-706.
Karaevli, A., & Zajac, E. J. 2013. When do outsider CEOs generate strategic change? The enabling role of corporate
stability. Journal of Management Studies, 50(7): 1267-1294.
Kesner, I., & Dalton, D. 1994. Top management turnover and CEO succession: an investigation of the effects of
turnover on performance. Journal of Management Studies, 31: 701–713.
Klarner, P., & Raisch S. 2013. Move to the beat - Rhythms of change and firm performance. Academy of
Management Journal, 56(1): 160-184.
Lavie, D., & Rosenkopf, L. 2006. Balancing Exploration and Exploitation in Alliance Formation. Academy of
Management Journal, 49(4): 797-818.
Lavie, D., Stettner, U., & Tushman, M. L. 2010. Exploration and exploitation within and across organizations. The
Academy of Management Annals, 4(1): 109-155.
Staying Agile in the Saddle
30
Leonard-Barton, D. 1992. Core capabilities and core rigidities: A paradox in managing new product development.
Strategic Management Journal, 13(5): 111–125.
Levinthal, D. A., & March, J. G. 1993. The Myopia of Learning. Strategic Management Journal, 14: 95-112.
Levitt, B., & March, J. G. 1988. Organizational Learning. Annual Review of Sociology, 14: 319-340
Lubatkin, M. H., Simsek, Z., Ling, Y., & Veiga, J. F. 2006. Ambidexterity and Performance in Small- to Medium-
Sized Firms: The Pivotal Role of Top Management Team Behavioral Integration. Journal of Management,
32(5): 646-672.
March, J. G. 1991. Exploration and Exploitation in Organizational Learning. Organization Science, 2(1): 71-87.
Marquis, C., & Huang Z. 2009. The contingent nature of public policy and the growth of U.S. commercial banking.
Academy of Management Journal, 52(6): 1222-1246.
Miller, D. 1991. Stale in the saddle: CEO tenure and the match between organization and environment.
Management Science, 37: 34–52.
Miller, D., & Shamsie J. 2001. Learning across the life cycle: Experimentation and performance among the
Hollywood studio heads. Strategic Management Journal, 22: 725–745.
Neter, J., Wasserman, W., & Kutner, M. 1985. Applied linear statistical model (2nd ed.). Homewood, IL: Irwin.
Ocasio, W. 1997. Towards an attention-based view of the firm. Strategic Management Journal, 18: 187–206.
O'Reilly, C. A., & Tushman, M. L. 2008. Ambidexterity as a Dynamic Capability: Resolving the Innovator's
Dilemma. In A. P. Brief & B. M. Staw (Eds.), Research in Organizational Behavior, Vol 28: An Annual Series
of Analytical Essays and Critical Reviews, Vol. 28: 185-206.
Prahalad, C. K., & Bettis, R. A. 1986. The Dominant Logic - a New Linkage between Diversity and Performance.
Strategic Management Journal, 7(6): 485-501.
Probst, G., & Raisch, S. 2005. Organizational Crisis: The Logic of Failure. Academy of Management Executive,
19(1): 90-105.
Quigley, T. J., & Hambrick, D. C. 2012. When the former CEO stays on as board chair: effects on successor
discretion, strategic change, and performance. Strategic Management Journal, 33(7): 834-859.
Raisch, S., & Birkinshaw, J. 2008. Organizational Ambidexterity: Antecedents, Outcomes, and Moderators. Journal
of Management, 34(3): 375-409.
Staying Agile in the Saddle
31
Raisch, S., Birkinshaw, J., Probst, G., & Tushman, M. L. 2009. Organizational Ambidexterity: Balancing
Exploitation and Exploration for Sustained Performance. Organization Science, 20(4): 685-695.
Rosenkopf, L., & Nerkar, A. 2001. Beyond local search: Boundary-spanning, exploration,and impact in the optical
disk industry. Strategic Management Journal, 22(4): 287–306.
Salancik, G. R., & Pfeffer, J. 1980. Effects of ownership and performance on executive tenure in U.S. corporations.
Academy of Management Journal, 23: 653-664.
Shen, W., Cannella, A. A. Jr. 2002. Revisiting the performance: consequences of CEO succession: The impacts of
successor type, postsuccession senior executive turnover, and departing CEO tenure. Academy of Management
Journal, 43: 717-734.
Siggelkow, N. 2002. Evolution toward Fit. Administrative Science Quarterly, 47(1): 125-159.
Simsek, Z. 2009. Organizational ambidexterity: towards a multilevel understanding. Journal of Management
Studies, 46: 597–624.
Simsek, Z., Heavey, C., Veiga, J. F., & Souder, D. 2009. A Typology for Aligning Organizational Ambidexterity's
Conceptualizations, Antecedents, and Outcomes. Journal of Management Studies, 46(5): 864-894.
Smith, K. G., Smith, K. A., Olian, J. D., Sims Jr, 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, 412-438.
Smith, W. K., & Lewis, M. W. 2011. Toward a Theory of Paradox: A Dynamic Equilibrium Model of Organizing.
Academy of Management Review, 36(2): 381-403.
Smith, W. K., & Tushman, M. L. 2005. Managing Strategic Contradictions: A Top Management Model for
Managing Innovation Streams. Organization Science, 16(5): 522-536.
StataCorp. 2009. STATA longitudinal-data/panel-data reference manual. College Station, Tx: Stata Press.
Taylor, A., & Greve, H. R. 2006. Superman or the fantastic four? Knowledge combination and experience in
innovative teams. Academy of Management Journal, 49(4): 723-740.
Tushman, M. L., & Rosenkopf, L. 1996. Executive succession, strategic reorientation and performance growth: a
longitudinal study in the US Cement Industry. Management Science, 42: 939-953.
Uotila, J., Maula, M., Keil, T., & Zahra, S. A. 2009. Exploration, Exploitation, and Financial Performance: Analysis
of S&P 500 Corporations. Strategic Management Journal, 30(2): 221-231.
Staying Agile in the Saddle
32
Virany, B., Tushman, M. L., & Romanelli E. 1992. Executive succession and organization outcomes in turbulent
environments: an organization learning approach. Organization Science, 3: 72-91.
Volberda, H., & Lewin, A. 2003. Co-evolutionary dynamics within and between firms: From evolution to
coevolution. Journal of Management Studies, 40: 2111-2136.
Weng, D. H., & Lin, Z. forthcoming. Beyond CEO tenure: The effect of CEO newness on strategic changes.
Journal of Management.
Wiersema, M. F., & Bantel, K. A. 1993. Top management team turnover as an adaptation mechanism: The role of
the environment. Strategic Management Journal, 14(7): 485-504.
Williams, K., & O’Reilly, C. A. 1998. Demography and diversity in organizations: A review of 40 years of research.
In B. M. Staw & R. I. Sutton (Eds.), Research in organizational behavior, vol. 20: 77–140: Greenwich, CT:
JAI Press.
Wooldridge JM. 2010. Econometirc Analysis of Cross Section and Panel Data. MIT University Press: Cambridge,
MA.
Wu S, Levitas E, Priem R. 2005. CEO tenure and company invention under differing levels of technological
dynamism. Academy of Management Journal 48(5): 859-873.
Zhang, Y. 2008. Information asymmetry and the dismissal of newly appointed CEOs: An empirical investigation.
Strategic Management Journal 29(8): 859-872.
Zhang Y, Rajagopalan N. 2003. Explaining new CEO origin: Firm versus industry antecedents. Academy of
Management Journal 46(3): 327-338.
Staying Agile in the Saddle
33
TABLE 1
Descriptive Statistics and Correlations
Variable Mean SD Min Max 1 2 3 4 5 6
1 Firm ambidexterity 0.04 0.45 0.00 0.25
1
2 CEO tenure 0.00 4.56 -4.88 24.12
-0.02 1
3 TMT turnover 0.00 2.13 -1.65 14.35
-0.06 -0.11** 1
4 ROE -1.39 38.16 -974.34 6.58
0.01 0.02 0.01 1
5 TMT size 4.87 2.98 0.00 20.00
-0.01 -0.08+ 0.37*** 0.04 1
6 Firm age 95.99 61.40 5.00 187.00
0.02 -0.26*** 0.32*** -0.03 0.32*** 1
+p < .10 * p < .05 ** p < .01 *** p < .001
Staying Agile in the Saddle
34
TABLE 2 CEO Tenure, TMT Turnover, and Firm Ambidexterity
Model 1 Model 2 Model 3 Model 4 CEO tenure 0.0006 0.0018* 0.0017* 0.0019**
(0.00) (0.00) (0.00) (0.00) TMT turnover -0.0001 -0.0003 -0.0002 -0.0014
(0.00) (0.00) (0.00) (0.00) CEO tenure squared -0.0001** -0.0001** -0.0000 (0.00) (0.00) (0.00) CEO tenure*TMT turnover 0.0003 -0.0001 (0.00) (0.00) CEO tenure squared*TMT turnover 0.0001+ (0.00) ROE 0.0078 0.0079 0.0080 0.0089
(0.01) (0.01) (0.01) (0.01) Firm ambidexterity 0.0281*** 0.0277*** 0.0277*** 0.0273*** (0.01) (0.01) (0.01) (0.01) TMT size -0.0011 -0.0012 -0.0012 -0.0011
(0.00) (0.00) (0.00) (0.00) Firm age -0.0000 -0.0000 -0.0001 -0.0000
(0.00) (0.00) (0.00) (0.00) Constant 0.1528*** 0.1559*** 0.1564*** 0.1528***
(0.02) (0.02) (0.02) (0.02)
N 181 181 181 181 r2 0.47 0.50 0.51 0.52 chi2 45.25*** 51.21*** 69.63*** 84.77*** Standard errors in parentheses; two-tailed for all variables. +p < .10 * p < .05 ** p < .01 *** p < .001