Boundary Conditions for Risk Taking Derived from Performance Feedback:
The Moderating Effect of Top Management Team Characteristics
Master’s Thesis Business Administration - Strategy
Student: Laura van Regenmortel / Student №: 10694404
University of Amsterdam, Faculty of Economics and Business
Supervisor: Bernardo Silveira Barbosa Correia Lima MSc
University of Amsterdam, Faculty of Economics and Business
STATEMENT OF ORGINALITY
This document is written by Student Laura van Regenmortel who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
ABSTRACT
Poor performance is an indicator of the fact the routines of the organization are not well
aligned with the environment and therefore serves as a signal for decision makers that
solutions are needed. However, there are conflicting results regarding how this search for
solutions influences the risk behaviour of firms. Namely, executives in some firms tend to
take more risks when performance falls below their aspirations, while other managers in the
same situation react risk averse. In an attempt to reconcile this contradicting results, the
current study examines characteristics of top managers drawn from the upper echelons theory
that are expected to moderate risk taking when performance falls short. First, I argue that
longer-tenured top managers are less capable of information processing and more committed
to status quo, and hence that performance below aspirations constraints risk taking in these
cases. Second, I suggest that firms with older top managers are less likely to take risks when
performance falls short since they have decreasing cognitive abilities and an increasing
commitment to established ways of doing things. Third, I also argue that female executives
are more risk averse than their male colleagues, and hence that performance shortfalls lead to
less risk taking when firms have a high number of female top managers. The results, however,
showed no significant support for these hypotheses, but in contrast to what was expected,
longer-tenured executives were found to actually take more risks when performance falls
short compared to the firm’s own past results. In this sense, the findings provide new insights and valuable contributions to theory and practice, as well as opportunities for future research.
TABLE OF CONTENTS
ABSTRACT ...3
I. Introduction ...5
II. Literature review ...8
1. Performance feedback and risk-related behaviour ...8
2. Literature gap and research question ... 19
III. Theoretical framework ... 21
1. Performance feedback, top management tenure and risk taking ... 22
2. Performance feedback, top management age and risk taking ... 24
3. Performance feedback, top management gender and risk taking ... 26
IV. Methodology ... 28
1. Sample and data collection ... 28
2. Dependent variable ... 30 3. Independent variables ... 31 4. Moderator variables ... 33 5. Control variables ... 35 6. Data analysis ... 37 V. Results ... 38 1. Descriptive statistics ... 38 2. Correlation analysis ... 42 3. Regression analysis ... 45 VI. Discussion ... 52
1. Performance feedback, top management tenure and risk taking ... 53
2. Performance feedback, top management age and risk taking ... 55
3. Performance feedback, top management gender and risk taking ... 56
4. Additional findings ... 58
5. Limitations ... 60
6. Implications for theory ... 62
7. Implications for practice ... 65
VII. Conclusion ... 66
References ... 69
Appendix A: Descriptives ... 78
I. Introduction
Performance feedback theory has developed into one of the main research streams on strategic
change in organizations over the past twenty years (Argote & Greve, 2007; Gavetti, Greve,
Levinthal, & Ocasio, 2012). The theory is built on the conviction that cognitive limited
decision makers use an aspiration level to evaluate performance and that the relative
difference of performance to aspiration level influences risk taking behaviour within the
organization (Greve, 2003a).
Within the performance feedback research stream there is an ongoing debate whether
firms react to performance below an aspiration level with risk taking or risk aversion (Audia
& Greve, 2006). Scholars focused on research concerning organizational decline argue that
underperforming firms reduce or eliminate risky activities and opt for efficiency and shy away
from innovations (K. Cameron, 1983; Hambrick & Schecter, 1983; Hofer, 1980; Staw,
Sandelands, & Dutton, 1981; Wiseman & Bromiley, 1996). On the other hand, research based
on the traditional behavioural theory of the firm (Cyert & March, 1963; Kahneman &
Tversky, 1979) argues that low performing organizations often take more risk and engage in
more innovative behaviour in their decision making (Bromiley, 1991; Greve, 1998, 2003b). It
can therefore be suggested that there is an apparent conflict in the literature concerning the
relation between organizational performance and risk-related behaviours.
More recently, researchers have begun to investigate boundary conditions under which
low performance is likely to lead to risk taking or risk aversion (Audia & Greve, 2006; Gaba
& Joseph, 2013; Greve, 2003a). This literature shows that (past) slack search (Greve, 2003a),
firm size (Audia & Greve, 2006) and corporate structure (Gaba & Joseph, 2013) affect the
relationship between performance feedback and risk-related behaviour. Although these
researchers contribute to the ongoing debate within the performance feedback literature by
feedback and risk taking, they have mostly ignored the individual and team processes which
are fundamental in the traditional decision making theory (Cyert & March, 1963) and
behavioural theory (Argote & Greve, 2007). Namely, these internal processes are especially
relevant since organizations are considered as the reflection of its employees. In particular,
individuals and teams within firms make decisions based on their personalized interpretations
of the situations they face. Since the choice to engage in risky actions or not is part of the
firm’s strategic decision making (Cyert & March, 1963; Hambrick & Mason, 1984), it is important to consider individual and team level processes in the performance feedback
literature.
However, the current literature is mainly treating the organization as a unitary agent in
assessing performance and making decisions (Argote & Greve, 2007) and consequently falls
short in taking into account the internal processes and structures described by Cyert and
March (1963). Only in Jordan and Audia (2012) and Kacperczyk, Beckman and Moliterno
(2015) the individual is taken into consideration in some part by incorporating self-enhancing
motives and greater risk taking in response to personal concerns of individuals in assessing
performance relative to the aspiration level.
Since the assessment of performance feedback and the specific decisions made within
organizations are affected by the personalized interpretations and perceptions of individuals
and teams, it is important to consider the specific characteristics of these actors that have an
impact on those internal processes (Argote & Greve, 2007). By recognizing that top managers
are the most important decision makers within firms (Hambrick, 2007; Hambrick & Mason,
1984), their individual characteristics can contribute to help explain when low performance
leads to risk taking or risk aversion. In this sense, by also building on the premise of bounded
rational individuals (Cyert & March, 1963; J. G. March & Simon, 1958), based on the notion
merely interpretable (Hambrick, 2007; Mischel, 1977), a promising school of thought is the
upper echelons theory. The fundamental idea of this theory is captured in Hambrick and
Mason's (1984) paper by maintaining that executives act and make decisions on the basis of
their individual interpretations of strategic situations and that these interpretations are the
result of their personal values, personalities and experiences (Hambrick, 2007). Therefore, for
understanding why organizations do the things they do, taking risks or not, it is important to
consider the biases and dispositions of their most powerful actors, namely top managers
(Hambrick, 2007).
In this way this research aims to contribute to the ongoing debate regarding how poor
performance impacts risk taking. To do this, the current study incorporates the upper echelons
theory within the performance feedback theory. In particular, this paper attempts to clarify the
effect of performance on risk-related behaviour by identifying the top management team
(TMT) characteristics as boundary conditions under which performance below aspiration
level leads to risk taking or risk aversion. Where in the current literature the organization is
mainly treated as a unitary agent and falls short in taking into account the internal processes
(Argote & Greve, 2007), this study addresses this gap by examining the effect the specific top
management characteristics have on the internal process of decision making when
performance decreases. Since decision makers use their personal values, personalities and
experiences to assess performance and making their decisions accordingly (Hambrick, 2007),
these heterogeneous characteristics can give the explanation for why some firms react with
risk taking and others with risk aversion when performance falls short. Furthermore, where
Jordan and Audia (2012) and Kacperczyk, Beckman and Moliterno (2015) only focus on
individual self-enhancing motives, the current research is especially unique by also
considering the broader characteristics of the whole TMT which can affect the team process
this sense, this paper attempts to give new insights to the performance feedback literature and
reconcile the contradictory views regarding how low performance affects risk taking by
incorporating the top management characteristics and processes as boundary conditions which
are central to the upper echelons theory.
This research addresses these ideas by examining three different TMT characteristics
that can affect risk-related behaviour when performance is below the aspiration level.
Namely, the TMT characteristics tenure, age, and gender are found to be primary predictors of
the cognitive base of executives and have been shown to lead to heterogeneous ways of
interpreting information and subsequently making risky strategic decisions (Finkelstein &
Hambrick, 1990; Finkelstein, Hambrick, & Cannella, 2009; Katz, 1982). The examination of
the effects of these specific characteristics can give new insights to the performance feedback
literature and can lead to valuable implications for practitioners when aligning the
composition of their TMT with the firm’s targeted strategy.
II. Literature review
In the following paragraphs the main insights of the existing literature on the topic will be
discussed. I first introduce the role that aspiration level has on the evaluation of performance
within the firm. Subsequently, I outline the main findings of the existing literature with regard
to the influence that performance feedback has on risk-related behaviour. Finally, I introduce
the role TMT characteristics can play as a boundary condition and guiding determinant in
organizational decision making derived from performance feedback.
1. Performance feedback and risk-related behaviour
Performance feedback is based on the notion that organizations learn from their past
experiences (Cyert & March, 1963; Greve, 2003a). Firms, with cognitive limited actors,
their organizational goals (Cyert & March, 1963; Gavetti e.a., 2012), where aspiration level is
“the smallest outcome that would be satisfactory by the decision maker” (Schneider, 1992, p. 1053). The levels of aspiration function therefore as an important benchmark and facilitate in
the interpretation and evaluation of organizational performance, which in turn influence the
decision making process and inclination to make changes and take risks (Audia & Greve,
2006; Cyert & March, 1963; Kim, Finkelstein, & Haleblian, 2014).
In the decision making process aspiration levels represent the borderline between
perceived success and failure (Greve, 1998; Schneider, 1992). However, since in many cases
the exact acceptable performance level is unclear, organizations may have different aspiration
levels to evaluate if their performance is perceived as a success or failure (Greve, 1998).
These variant levels could emerge from two different sources of performance feedback –
historical and social (Argote & Greve, 2007; Greve, 1998; Kim e.a., 2014).
Historical performance feedback is based on the firm’s own performance history
(Cyert & March, 1963; Kim e.a., 2014; Levinthal & March, 1981). Greve (1998) suggests that
this type of feedback can be used for setting levels of aspiration that retain the differences
between firms constant. The usefulness of these historical aspirations lies especially in the
fact that it can be used to evaluate the effectiveness of the current capabilities of the
organization (Kim e.a., 2014; Levinthal & March, 1981). This further results in the fact that
managers have access to private knowledge within the firm, which they can use to identify
those factors that contribute to the experienced performances or not (Menon & Pfeffer, 2003).
In this light, historical aspirations can be seen as a relatively credible predictor of how well
the organization could perform given their current resources and capabilities (Greve, 2003a;
Kim e.a., 2014) and are particularly suited for firms acting in stable and imperfect markets
(Greve, 2003b). In these markets historical aspiration levels may help organizations with
and competitive markets historical aspirations, due to their idiosyncratic nature, can be a poor
reflection of the organizational environment (Greve, 2003b). In these kinds of markets social
performance feedback may be better suited for setting levels of aspiration.
In case of social performance feedback, aspiration levels are determined by the
performance of similar others (Cyert & March, 1963; Festinger, 1954; Greve, 1998; Kim e.a.,
2014). In this sense, firms set their performance goals by observing the performance of the
specific reference group (Kim e.a., 2014). Comparable organizations become a particular
useful reference point for executives to benchmark and evaluate their own performance
(Fiegenbaum, Hart, & Schendel, 1996) and therefore their social aspiration levels. Although
benchmarking against others may improve judgements when it is based on an analysis of both
the value chains and markets of the focal organization and others, intuitive judgements tend to
have a preference for market characteristics due to their easy availability (Clark &
Montgomery, 1999). This may result in concealment of differences in the underlying
capabilities (Clark & Montgomery, 1999) and, due to simplicity of intuitive judgements, the
ignorance of many characteristics that distinguish organizations (Greve, 2003b; Porac &
Rosa, 1996).
Following, the existing literature is inconclusive with regard to whether organizations
need to use historical aspirations, social aspirations, or both due to the biases the sources
contain. Where some researchers (Audia & Brion, 2007; Audia & Greve, 2006; Greve, 2003b)
found more significant effect for historical aspirations, others found more evidence for social
aspirations (Harris & Bromiley, 2007; Mishina, Dykes, Block, & Pollock, 2010). However
Iyer and Miller (2008) also report the importance of equivalence like the traditional
behavioural theory of Cyert and March (1963), which emphasizes that the aspirations of an
organization are based on the past goals, past performance (historical aspirations) and past
mainly used to evaluate organizational performance (Cyert & March, 1963; Greve, 1998,
2003a; Kim e.a., 2014). Although decision making becomes a heuristic by assessing
performance based on historical and social levels of aspiration, it resembles the most formal
way of forming expectations (Greve, 1998), which determines if organizations engage in
taking risks or not.
Hence, performance targets or aspiration levels are set to assess the results of the
organization (Cyert & March, 1963; Desai, 2008). When performance is below these levels of
aspiration this is a signal that the practices and routines of the firm may not be aligned with
the requirements of the environment (Boeker & Goodstein, 1991). Although most studies are
conclusive that performance above the aspiration level decreases innovation and risky
decision making (Bromiley, Miller, & Rau, 2001; Cyert & March, 1963; Greve, 1998, 2003b;
Nickel & Rodriguez, 2002), there is an ongoing debate in the literature about how
organizations react on the notion that performance is below their set aspiration level (Audia &
Greve, 2006; Desai, 2008). In this line of discussion, the behavioural theory maintains that
when actual or expected performance of the firm falls below their aspiration level, also
referred to as ‘performance-aspiration gap’ (Joseph & Gaba, 2014) or ‘attainment
discrepancy’ (Lant, 1992), decision makers engage in ‘problemistic search’ (Audia & Greve,
2006; Cyert & March, 1963; Greve, 2003a, 2003b; Shinkle, 2012). Cyert and March (1963, p.
121) define problemistic search as “search that is stimulated by a problem and is directed towards finding a solution to that problem” and maintain that this results in organizational
change. Illustrative, it has been suggested that organizations may become more likely to
explore new opportunities and to learn from other organizations when performance falls short,
instead of relying only on the internal available knowledge (Desai, 2008).
Where the behavioural theory suggests that firms reduce their search activity and
organizational decline maintains that firms performing below aspiration level may also reduce
or eliminate risky activities and shy away from innovations (K. Cameron, 1983; Hambrick &
Schecter, 1983; Hofer, 1980; Staw e.a., 1981; Wiseman & Bromiley, 1996). Organizational
decline is “a condition in which a substantial, absolute decrease in an organization’s resource base occurs over a specified period of time” (K. S. Cameron, Kim, & Whetten, 1987, p. 224). Firms facing low performance within the organizational decline literature, also referred to as
the threat-rigidity theory (K. Cameron, 1983; McKinley, Latham, & Braun, 2014; Staw e.a.,
1981), tend to focus their attention more on efficiency, cost cutting, and tighter budgets,
which impede risk taking. While the behavioural theory in general argues that increased risk
aversion will cause further organizational decline because it fails to solve the problems
concerning organization’s strategic misfit, Staw e.a. (1981) suggest that under some conditions firms can succeed through this approach. In the case that changing environments
make established routines effective again, firms can reverse organizational decline by
avoiding innovations and risky activities that fail to generate revenues and hold high costs
(McKinley e.a., 2014). This idea is consistent with other findings that moderate rather than
high risk tolerance maximizes return on assets (ROA) (Walls & Dyer, 1996), that strategic
inertia (low innovation) is related to increased future performance (D. Miller & Chen, 1994),
and that Red Queen survivors, which are firms that have developed through the reinforcement
of competition and learning, that diversify into new market niches run high risks of failure
(Barnett & Pontikes, 2008).
Although the behavioural theory and the threat-rigidity theory do not agree as to
whether firms facing performance below aspiration level engage in risky behaviour or not,
both streams of literature in general treat firms in their decision making as a unitary agent
(Argote & Greve, 2007). In this sense performance feedback is interpreted within the
interpretation occurs or not, various studies (Beckman, Haunschild, & Phillips, 2004; Hill,
Hitt, & Hoskisson, 1992; Sanders & Hambrick, 2007) maintain that risk-related behaviour of
firms can be expressed in research and development (R&D) expenditures. R&D expenditures,
also indicated as R&D intensity, can be channelled to the development of new products and to
increased support of existing projects (Greve, 2003a). Although Greve (2003a) suggests that
R&D can reduce organizational risk because it allows executives to base decisions for
innovation on the expected returns and risks, R&D expenditures are in the vast majority of the
literature (Lim & McCann, 2013; Vissa, Greve, & Chen, 2010) acknowledged to be related to
firm’s risk taking when performance is below the aspiration level. Namely, this stream of literature characterizes R&D expenditures by its highly uncertain returns, especially when
these investments are large and in unrelated businesses (Lim & McCann, 2013; Vissa e.a.,
2010). This means that when performance is below the social and historical aspiration level,
decision makers within organizations can choose to make risky investments in R&D or not in
order to close the performance-aspiration gap.
In this sense, there can be a difference in the organizational decisions makers’
attention on risk taking or risk aversion as response to performance shortfalls (Audia &
Greve, 2006; Desai, 2008). Namely, March and Shapira (1987) argue that decision makers
shift their attention between aspiration and a point of survival at which the firm fails due to
extremely low performance. This means that when the focus is upward to exceed the
aspirations in the future, risk taking will ensue when performance decreases below the
aspiration level, since outcomes of these decisions are uncertain and the variability in
potential performance may lead to improvements (Audia & Greve, 2006; Desai, 2008).
However, when the focus is on the survival point, performance shortfalls induce risk aversion
since the uncertain outcomes of risky decisions, like R&D expenditures, could lead to actual
Therefore, whether firms react with risk taking or risk aversion when performance is
low depends on the shifting attention of decision makers’ within firms on aspirations or survival. This means that risk preferences are not stable over time and can be influenced by
various determinants of attention focus (J. G. March & Shapira, 1987; K. D. Miller & Chen,
2004). Namely, it has been found that when performance is extremely low, such as
bankruptcy, it appears to restrict risk taking by making the decision makers more attentive for
the firm’s survival (J. G. March & Shapira, 1987; K. D. Miller & Chen, 2004; Sitkin, 1992). However, since these kinds of major shortfalls are rare, also a wider range of contextual
factors are examined in the literature that influence the attention focus and can act as a
boundary conditions under which low performance is likely to lead to risk taking or not
(Audia & Greve, 2006; Gaba & Joseph, 2013; Greve, 2003a).
In this sense, this stream of literature (Audia & Greve, 2006; Gaba & Joseph, 2013;
Greve, 2003a) attempts to reconcile the contradictory views within the performance feedback
theory concerning whether low performance leads to risk taking or risk aversion. By
examining the contextual factors of organizational decision making, this literature shows that
(past) slack search (Greve, 2003a), firm size (Audia & Greve, 2006) and corporate structure
(Gaba & Joseph, 2013) affect the relationship between performance feedback and risk-related
behaviour. As suggested by Greve (2003a), slack resources, defined as “the pool of resources in an organization that is in excess of the minimum necessary to produce a certain level of
organizational output” (Nohria & Gulati, 1996), serve as an important condition for organizational risk taking since it has a positive impact on the engagement in experimentation
resulting in opportunities for innovation and change for the firm (Cyert & March, 1963;
Levinthal & March, 1981). Also the firm size is found by Audia and Greve (2006) as an
important contextual factor that has an impact on the risk-related behaviour of the
they found that performance decreases within big firms with large resource endowments are
viewed as repairable and lead to risk taking, where small firms with limited resources react
more conservatively when performance falls short. Furthermore, Gaba and Joseph (2013) also
examined the organizational structure as a boundary condition that might affect risk taking
when performance decreases. They show that due to the fact that the corporate structure
vertically differentiates performance goals and problemistic search, solutions to performance
shortfalls vary between corporate and business unit levels, which means that there is a
variability in risky strategic decisions made by different organizational decision makers in
various contexts.
Although these researchers (Audia & Greve, 2006; Gaba & Joseph, 2013; Greve,
2003a) contribute to the debate within the performance feedback theory by examining the
boundary conditions under which low performance leads to either risk taking or risk aversion,
this studies tend to fall short in taking into account the individual and team processes which
are fundamental to both the traditional decision making theory (Cyert & March, 1963) and
behavioural theory (Argote & Greve, 2007). In this sense, the organization is mainly treated
as a unitary actor in assessing performance and making decisions, which means that internal
processes are currently not adequately covered in the literature (Argote & Greve, 2007).
Following, the current literature is limited since especially those individual and team level
processes serve as an important factor in strategic decision making. Namely, by typifying the
firm as a reflection of its employees, individuals and teams within organizations are
considered to make decisions based on their personalized interpretations of the context in
which they act. Since the choice to act with risk seeking or risk aversion is part of the firm’s strategic decision making process and can vary for each person and group (Cyert & March,
processes on the individual and team level as boundary conditions for risk taking into the
performance feedback literature.
Only Jordan and Audia (2012) and Kacperczyk e.a. (2015) take the individual into
account in some part by incorporating self-enhancing motives and greater risk taking in
response to personal concerns of individuals in assessing performance relative to aspiration
level. In this line of thought, Jordan and Audia (2012) argue that performance assessment
processes may become more complex when performance is below the aspiration level than
suggested by the problem-solving mode assumed in the performance feedback theory. This
means that decision makers not only conclude that a performance gap exists, but that the
self-enhancing modes of performance assessment can play an important role in setting and, later,
retrospectively revising performance evaluation standards to minimize the
performance-aspiration gap (Staw, 1980). Due to this self-enhancement modes individual decision makers
can reduce or even eliminate the perception of performance problems and in that way
diminish the urge to take risky actions (Jordan & Audia, 2012). Besides that, Kacperczyk e.a.
(2015) state that individual managers are actually more likely to take risky decisions in case
of low performance relative to internal social comparison. In this sense, social and spatial
proximity of internal comparisons trigger personal concerns and fear of negative individual
consequences (e.g. job loss). On the other hand, managers consider a performance shortfall
relative to external benchmarks more as organizational concern, which reduces their urge to
take risky decisions (Kacperczyk e.a., 2015).
Hence, this stream of literature emphasizes the important role internal processes and
individuals can play in the assessment of performance, decision making and risk taking within
firms. In this sense, these studies attempt to reconcile the conflicting views in the performance
feedback literature between the traditional behavioural theory and the threat-rigidity theory by
to risk taking or risk aversion of organizational decision makers. Although both
self-enhancement modes and internal social comparison are found to be important factors that
explain why some firms react with risk seeking and others with risk aversion when
performance decreases below the aspiration level, Jordan and Audia (2012) and Kacperczyk,
Beckman and Moliterno (2015) studies are limited by only focusing on these individual
motives in general.
Therefore, this paper aims to give new insights and contribute to the current debate
within the performance feedback literature by incorporating the upper echelons theory which
specifically emphasizes the relevance of top managers in the organization. Where both the
decision making theory and the upper echelons theory acknowledge the importance of
individuals within organizations, the upper echelons theory namely suggests that the firm’s
top managers (upper echelons) in particular have a major impact on the organization as a
whole (Hambrick, 2007; Hambrick & Mason, 1984). This stream of literature maintains that
the individual cognitive base of top managers evolving from experiences, influence their
personalized ways of interpreting information, like performance feedback, and therefore can
serve as an important boundary condition for risky decision making within organizations
(Finkelstein e.a., 2009). In this sense, previous research has found that those specific
experiences and demographic characteristics are predictors of the beliefs, values and abilities
of executives that have an important impact on risk-related behaviours within the firm. For
example, Carlsson and Karlsson (1970) found that age is negatively related to risky decision
making due to the diminishing cognitive abilities and the commitment to status quo of older
executives. It is therefore that some TMT’s tend to engage more in risky behaviour (e.g. R&D
expenditures (Sanders & Hambrick, 2007)) in case performance is below aspiration level,
whereas others tend to react in a more risk averse way. Hence, these differences between
functional background, education, age, gender, etc.) (Hambrick & Mason, 1984). Especially
since TMT characteristics can affect the process of interpretation and subsequently the risky
decision making of the top managers through their combined personal backgrounds and
experiences (Hambrick, 2007; Hambrick & Mason, 1984). In this sense, this paper attempts to
give new insights and reconcile the contradictory views of the traditional behavioural theory
and the theory of threat-rigidity regarding how performance shortfalls influence risk taking by
incorporating the top management characteristics and processes as boundary conditions drawn
from the upper echelons theory.
Hence, this research examines whether and how these TMT characteristics affect the
organizational behaviour of risk taking through the personalized interpretation of performance
feedback. Although in the upper echelons theory a variety of top managers’ characteristics are
outlined (Hambrick & Mason, 1984), for the purpose of this study I examine how executive
tenure, age, and gender can each influence the attention focus of the managers and whether
this leads to risk taking or risk aversion as response to organizational performance decreases.
Previous research has namely shown that these three specific characteristics are indicators of
interpretations and attitudes about change and risk taking of top managers (Hambrick &
Mason, 1984; Homberg & Bui, 2013; Rost & Osterloh, 2010; Wahid, 2012; Wiersema &
Bantel, 1992) and therefore can act as boundary conditions for risky decision making based
on the personalized interpretation of performance feedback. In this sense, these kinds of
conditions can explain why some firms take more risks than others when performance is
below the social and historical aspiration level and therefore needs more attention and
examination for the reconciliation of the contradictory views within the performance feedback
2. Literature gap and research question
Performance below and above the aspiration level are found to have an influence on firms’
risk taking behaviour (Greve, 1998, 2003a; Staw e.a., 1981; Wiseman & Bromiley, 1996).
The existing literature has tested both the effects of performance below and above the
aspiration level on risk taking under various conditions. Although it is generally
acknowledged that performance above the aspiration level leads to less risk taking (Bromiley
e.a., 2001; Cyert & March, 1963; Greve, 1998, 2003b; Nickel & Rodriguez, 2002), the
literature is inconclusive about whether decreases in performance below the aspiration level
lead to risk seeking or risk aversion (Audia & Greve, 2006). Where research within the
traditional behavioural theory (Cyert & March, 1963; Kahneman & Tversky, 1979) maintains
that performance decreases lead to more innovations and risk taking, the threat-rigidity
literature emphasizes that organizations take less risks and shy away from innovations when
performance declines (K. Cameron, 1983; Hambrick & Schecter, 1983; Hofer, 1980; Staw
e.a., 1981; Wiseman & Bromiley, 1996). This indicates the current conflict and the ongoing
debate in the performance feedback literature about whether firms react to performance
decreases with risk taking or risk aversion.
Therefore, some studies have examined the boundary conditions under which
performance shortfalls lead to risk taking or not (Audia & Greve, 2006; Gaba & Joseph, 2013;
Greve, 2003b). However, most articles on this topic are mainly treating the organization as a
unitary agent when assessing performance and making decisions (Argote & Greve, 2007). In
particular, existing research falls short in considering the individual and team processes which
are fundamental in the traditional decision making theory (Cyert & March, 1963) and the
behavioural theory (Argote & Greve, 2007). Namely, in reality strategic decisions within
firms are made by cognitive limited decision makers (top managers) who use an aspiration
1963). Characteristics of top managers (e.g. tenure, age, gender) have been shown to guide the
way executives make these individual interpretations of the context they face as a result of
their personal values, beliefs and experiences (Hambrick, 2007; Hambrick & Mason, 1984).
In this sense, representatives of the upper echelons theory state that the specific characteristics
and experiences of top managers can lead to biases and specific dispositions which can play a
major role in the way they process information and make decisions accordingly (Finkelstein
& Hambrick, 1990; Finkelstein e.a., 2009; Katz, 1982). Since interpreting performance
feedback and subsequently taking risks or not by executives are important parts of strategic
decision making when performance decreases, it is reasonable to suggest that TMT
characteristics, as executive tenure, age and gender, can act as an important boundary
condition. In particular, heterogeneous characteristics of executives can explain why some
firms react with risk taking and others with risk aversion when performance falls short.
In this sense, the upper echelons theory can contribute to the ongoing debate between
the traditional behavioural theory (Cyert & March, 1963; Kahneman & Tversky, 1979) and
the threat-rigidity theory (Staw e.a., 1981; Wiseman & Bromiley, 1996) about the way
organizations react on performance shortfalls. It provides one way to better understand
whether performance decreases lead to risk taking or risk aversion by examining the effect
that TMT characteristics have on this relationship (Sanders & Hambrick, 2007). Hence, this
study aims to close the gap within the performance feedback literature by giving an in-depth
understanding of the moderating effect that characteristics of top managers have on the firm’s
risk behaviour resulting from their personalized ways of interpreting performance feedback.
Hence, the research question that this study aims to answer is the following:
What is the moderating effect of top management team characteristics on the relationship between performance below the aspiration level and risk taking behaviour?
III. Theoretical framework
This academic paper aims to contribute to the debate in the performance feedback literature
by theorizing that risk taking behaviour is influenced by three different TMT characteristics
central in the upper echelons theory – executive tenure, age, and gender (Hambrick & Mason,
1984; Homberg & Bui, 2013; Rost & Osterloh, 2010; Wahid, 2012; Wiersema & Bantel,
1992). There is strong evidence that heterogeneous characteristics of top managers lead to
differences in the personal interpretation of information and subsequently strategic decision
making (Finkelstein e.a., 2009; Hambrick & Mason, 1984). Namely, differences in the
cognitive base and attention focus evolving from the executives’ own personal experiences,
values and abilities can explain why some top managers are more likely to take risks in
organizational decision making than others (Hambrick, 2007). This is also illustrated by
research of Finkelstein and Hambrick (1990) who found in the computer, chemical, and
natural-gas distribution industries that managerial teams with longer-tenured executives
followed more persistent, risk averse strategies than teams with shorter-tenured managers
when assessing performance.
Hence, in congruence with the assumptions that performance feedback needs to be
interpreted by top managers, as most powerful actors, and that these managers are accordingly
key in the decision making process of the firm (Hambrick & Mason, 1984), I expect that
executive tenure, age and gender will explain differences in organizational risk taking derived
from the personalized interpretations of performance decreases. In this way, the incorporation
of the upper echelons theory within the performance feedback theory can provide new
insights and contributions in the reconciliation of the contrasting views between the
traditional behavioural theory (Cyert & March, 1963; Kahneman & Tversky, 1979) and the
1. Performance feedback, top management tenure and risk taking
Research within the upper echelons theory maintains that the amount of years executives have
spent in the organization (i.e. tenure) is generally acknowledged for affecting the cognitive
base, the information processing and the decision making by managers within firms
(Finkelstein & Hambrick, 1990; Finkelstein e.a., 2009; Katz, 1982). Hence, executive tenure
(in years) is likely to have an impact on the individual interpretation of performance feedback
and subsequently the risk taking behaviour of the firm (Finkelstein & Hambrick, 1990) and
therefore can contribute to the ongoing debate in the performance feedback literature.
Namely, differences in TMT’s executive tenure is expected to explain why some firms react
on performance below aspiration level with risk taking and others with risk aversion (Katz,
1982).
Although some studies (Simsek, 2007; Zahra, 2005) maintain that long-tenured
managers are more likely to take risks due to the fact that their experience will let them
perceive less uncertainty regarding the possible outcomes of risky decisions, it is generally
acknowledged in the upper echelons theory that executive tenure leads to commitment to the
status quo, less informational diversity, and increased risk aversion (Finkelstein e.a., 2009;
Pfeffer, 1985). This stream of literature suggests that as individuals spend more time in an
organization, especially when they have success and climb the hierarchy, they become
committed to established policies and practices (Finkelstein & Hambrick, 1990; Finkelstein
e.a., 2009; J. C. March & March, 1977). This means that once people are committed to a
course of action, they resist changing their behaviour, whether these actions are a success or
not (Katz, 1982). Therefore, executive tenure is in this line of thought associated with
restricted information processing and interpretation through the executives’ establishments of routines, the use of familiar information sources, and relying more and more on past
e.a., 2009; Katz, 1982). Logically following on their commitment to stability, organizational
tenure has an impact on the risk taking behaviour of top managers. Since these individuals
have invested years to achieve their position in the top of the organization and are well
established in their work environment, it is found that they have more to lose than to gain by
engaging in risk-related behaviour (Finkelstein & Hambrick, 1990).
For example, research by Katz (1982) shows that groups with longer-tenured members
process less information and are more committed to common sets of stable beliefs about their
work setting, which restrict their tendencies to innovate and take risks. Similarly, Wiersema
and Bantel (1992) offer additional evidence that executive tenure reduces risky decision
making. In their study of 87 firms, they found that the tenure of top managers has a highly
positive influence on strategic persistence and conformity to the general strategic tendencies
in the industry. Moreover, when long-tenured executives need to interpret performance
feedback, they are perceived to do this through their established routines and by relying
especially on their past experiences (Finkelstein & Hambrick, 1990; Finkelstein e.a., 2009;
Katz, 1982).
Following, in situations where the firm’s performance is below the aspiration level,
TMT members with longer tenures are expected to engage in less problemistic search and
react in a more risk averse way than their shorter-tenured colleagues (Finkelstein &
Hambrick, 1990). This means that they are less likely to engage in risk taking and
innovations, which can be expressed by investments in R&D. In line with these assumptions
of the traditional upper echelons theory (Finkelstein & Hambrick, 1990; Finkelstein e.a.,
2009; Katz, 1982), it is reasonable to expect that having a substantial amount of
longer-tenured executives within the TMT leads to less risk taking when performance is below the
Hypothesis 1: When performance is below the aspiration level, performance decreases lead to less risk taking among firms with higher levels of TMT tenure than among firms with lower levels of TMT tenure.
2. Performance feedback, top management age and risk taking
Another TMT characteristic that is found to have an influence on the cognition and decision
making within the firm is the age of executives (Hambrick & Mason, 1984; Hitt, Hoskisson,
Johnson, & Moesel, 1996; Wiersema & Bantel, 1992; Yang & Wang, 2014). Namely, the top
manager’s age (in years) indicates the cognitive framework and skills on which their actions within the firm are based (Wiersema & Bantel, 1992). Therefore, the examination of the
executive’s age can be of value in the discussion about the way organizations react on performance feedback. Namely, the differences in age of top managers can determine why
some are more likely to engage in innovative and risky behaviour (e.g. R&D expenditures)
than others when performance declines (Wiersema & Bantel, 1992).
Although older executives tend to seek more information before making decisions, are
more likely to accurately evaluate this information, and take sufficient time to make the
eventual decisions (Hambrick & Mason, 1984; Taylor, 1975), most of the scholars recognize
that as people become older, they experience a decrease in cognitive abilities (e.g. memory,
learning ability, and reasoning) and therefore the ability to process and integrate information,
like performance feedback (Bantel & Jackson, 1989; Hambrick & Mason, 1984).
Furthermore, older top managers tend to be more committed to the organizational status quo
(Grimm & Smith, 1991; Stevens, Beyer, & Trice, 1978). In this sense, older executives within
the TMT are more likely to hold-on to their established ways of information processing and
decision making when performance declines. Namely, it has been found that when people
(Wiersema & Bantel, 1992). Besides that, older executives tend to place more value on
security, in both career and financial terms, compared to their younger colleagues who find
having a challenging job, promotions and growth more important (Hambrick & Mason, 1984;
Wiersema & Bantel, 1992). Hence, since younger executives are more flexible and oriented
towards growth, these managers are expected to engage more in innovative and risky
strategies to seize opportunities than their older colleagues when performance falls short
(Grimm & Smith, 1991; Hambrick & Mason, 1984). Contrarily, older executives are, due to
their commitment to stability, more likely to react inflexible to change and avoid risky
decisions in case of performance decreases (Bantel & Jackson, 1989; Carlsson & Karlsson,
1970; Goll, Sambharya, & Tucci, 2001; Vroom & Pahl, 1971), which could have a
considerable impact on the strategic direction of the firm (Hambrick & Mason, 1984).
For example, early research of Vroom and Pahl (1971) administrated to managers
from over 200 different companies found evidence that there is a negative relationship
between age and both risk taking and the value placed upon risk. Additionally, empirical
support is found by Thomas, Litschert and Ramaswamy (1991) who also show that executive
age is negatively associated with risky product or market innovation strategies when
examining the impact of the fit between top management characteristics and strategic
orientation on firm performance. Therefore, having a considerable amount of older executives
within the TMT is expected to lead to less problemistic search and risk taking, expressed in
R&D intensity, when performance is below the aspiration level (Carlsson & Karlsson, 1970;
Goll e.a., 2001; Vroom & Pahl, 1971; Wiersema & Bantel, 1992). This results in the
following hypothesis:
Hypothesis 2: When performance is below the aspiration level, performance decreases lead to less risk taking among firms with higher levels of TMT age than among firms with lower levels of TMT age.
3. Performance feedback, top management gender and risk taking
Gender of top managers is also acknowledged in the TMT literature as a characteristic that
has an influence on the information processing and decision making of executives within
organizations (Johnson & Powell, 1994; Klenke, 2003; Schubert, Brown, Gysler, &
Brachinger, 1999; Yang & Wang, 2014). Since it is generally found in the literature that there
is clear distinction in the way males and females process and interpret information (e.g.
performance feedback) and the type of decisions they make (Johnson & Powell, 1994; Rost &
Osterloh, 2010; Yang & Wang, 2014), executive gender can be of value in the ongoing debate
in the performance feedback literature. Namely, the differences in interpretation of
performance feedback and subsequently taking risky decisions or not is expected to be
explained by the discrepancy in risk propensity between the two sexes (Levin, Snyder, &
Chapman, 1988; Levin e.a., 1988; Powell & Ansic, 1997).
Although there are some studies that claim that there are no gender differences in risk
taking (Arenson, 1978; Noe, McDonald, & Hammitt, 1983), it is suggested that this lack of
significant differences is due to the fact that these studies were mainly conducted on children,
which means that a person’s gender and their risk propensity is in these cases more related to their age (Coet & McDermott, 1979). Hence, it is generally acknowledged that there is a
discrepancy between the two sexes following from findings that females are more risk averse
than males (Ginsburg & Miller, 1982; Johnson & Powell, 1994; Levin e.a., 1988). In general
this means that women are less willing to take extreme risks and therefore are more likely to
prefer making decisions that have a high probably of some, even low returns. This
commitment to stability and organizational status quo of female executives, it is expected to
be further strengthened by the fact that women are especially more risk sensitive in situations
of losses, like performance decreases, compared to gains (Fehr-Duda, De Gennaro, &
tolerance and are therefore expected to have a preference for more risky decisions
characterized by having a lower probability on higher returns in case of performance
shortfalls (Coombs & Pruitt, 1960; Johnson & Powell, 1994; Rost & Osterloh, 2010; Yang &
Wang, 2014).
Illustrative, Johnson and Powell (1994) found in a non-managerial setting, by
examining betting behaviour and investment decisions of management students, empirical
support that in general women exhibit lower preference for risk in the population. These
findings are further supported by research in business specific literature, as the study of
Sexton and Bowman-Upton (1990) shows that female entrepreneurs score lower for risk
taking, with the strongest effect for monetary risks, than male entrepreneurs by using
self-administered psychometric measures. Hence, it is expected that female executives are more
risk averse compared to their male colleagues and are consequently less likely to engage in
risky decisions when performance decreases (Coombs & Pruitt, 1960; Johnson & Powell,
1994; Yang & Wang, 2014). Therefore, having a larger amount of female executives within
the TMT is expected to result in less risk taking, expressed in R&D expenditures, when
performance is below the social and historical aspiration level. These assumptions result in
the following hypothesis:
Hypothesis 3: When performance is below the aspiration level, performance decreases lead to less risk taking among firms with higher levels of female representation within the TMT than among firms with lower levels of female representation within the TMT.
Based on the three hypotheses developed in the theoretical framework section by
incorporating the upper echelons theory within the performance feedback theory, the
IV. Methodology
In this chapter the research approach and design of this study are explained. Firstly, a
description of the sample and the data collection methods are presented. The section continues
with a detailed operationalization of the variables and a description of the way these variables
are measured. This chapter ends with a brief explanation of the methods used to analyse the
panel data in this study.
1. Sample and data collection
In order to answer the research question and examine the possible moderating effect of top
management characteristics a quantitative, empirical research design is used. Following
existing literature, information on the necessary variables is retrieved from two databases:
Compustat is used for the financial data (Cannella, Park, & Lee, 2008; Carpenter, 2002; Iyer
& Miller, 2008; Wiseman & Bromiley, 1996) and ExecuComp, supplemented with company
sources (e.g. reports, corporate websites) and internet sources (e.g. LinkedIn, Bloomberg), is
used for the data about executives (Adams, Almeida, & Ferreira, 2009; Cain & McKeon,
2014; Hambrick, Humphrey, & Gupta, 2015; Wahid, 2012). These databases were primarily Performance feedback
Performance below aspiration level (social and historical)
TMT characteristics Executive tenure Executive age Executive gender
Risk taking behaviour R&D expenditures
Control variables TMT size
Firm size Slack resources Figure 1 - Conceptual Framework
used because they are easy to access and possess the information that is needed to answer the
research question (Saunders, Lewis, & Thornhill, 2012, pp. 318–2123).
Similar to samples used in previous studies (Fligstein, 1987; Koyuncu, Firfiray, Claes,
& Hamori, 2010; Ocasio & Kim, 1999), the sample of this research includes the largest
companies in the United States (U.S.) as listed in the Standard & Poor’s 500 (the S&P 500),
with 2010 as the base line year. These are the firms that are the largest 500 publicly held
companies that trade on either of the following two American stock markets: the NYSO
Eurnonext and the NASDAQ OMX (Koyuncu e.a., 2010). Because of the size of these S&P
500 firms there is a bigger change on disclosure quality and completeness of the financial and
executives’ data needed for the sake of this research. Concerning the S&P 500 firms, this research is restricted to the companies in the manufacturing industries (Standard Industrial
Classification (SIC) codes 2000 to 3999), because this approach prevents that the obtained
confounding results can be owned to major disparities in industries’ activities (Brush, 1996;
Hitt e.a., 1996; Iyer & Miller, 2008). By restricting the sample to the manufacturing firms in
the S&P 500, I ended up with a sample of 197 firms.
In order to examine the moderating effect of top management characteristics on the
relationship between performance feedback and risk taking a longitudinal (panel) research is
needed (Finkelstein & Hambrick, 1990; Greve, 1998). Therefore, data from all the S&P 500
firms in the manufacturing industry were gathered from Compustat and ExecuComp for the
years from 2010 to 2014, which resulted in a total of 951 firm-year observations. This specific
time horizon is chosen by following those used in the current upper echelons literature (e.g.
Cannella et al., 2008; Hambrick, Humphrey, & Gupta, 2015) and by taking into account the
time available for the execution of this research.
The data retrieved from both databases (Compustat and ExecuComp) were merged by
match the TMT data (e.g. tenure, age, gender, TMT size) with the financial data (e.g. ROA,
aspiration level, total assets) of this research. By merging the two datasets, 910 out of the 951
firm-year observations were matched, which means that for 41 of the observation I was not
able to find a match. The next sections describe the operationalization and the specific
measurements used for the dependent, independent, moderator and control variables.
2. Dependent variable
Consistent with prior studies (Coles, Daniel, & Naveen, 2006; Lim & McCann, 2013; Sanders
& Hambrick, 2007; Vissa e.a., 2010), I used research and development (R&D) expenditures
as the proxy for firm’s risk taking in this study, which is reported as XRD in Compustat.
R&D expenditures include all costs incurred during the year that relate to the development of
new products or services (Beckman e.a., 2004; Hill e.a., 1992; Sanders & Hambrick, 2007). I
specifically focused on R&D expenditures because it is an indicator of risky and uncertain
long-term investment behaviour (Beckman & Haunschild, 2002; Hoskisson, Hitt, & Hill,
1993). By following Greve (2003a), I operationalized R&D expenditures by using a
measurement of intensity. In this sense, I computed R&D intensity as R&D expenditures
(XRD) scaled by sales (Chen & Miller, 2007; Greve, 2003a). By scaling the expenses by the
total of sales, I could use the intensity measurement of R&D in order to test my hypotheses.
The definition and operationalization of the dependent variable are outlined in Table 1
– Definition of the Dependent Variable and Table 2 – Dependent Variable – Operationalization.
Table 1 - Definition of the Dependent Variable
No Dependent variable Definition Key reference
1. Research & Development expenditures
Investments related to the development of new products or services
Beckman, et al.(2004) Hill, et al. (1992)
Table 2 - Dependent Variable - Operationalization
No Dependent variable Operationalization Key reference 1. Research & Development
expenditures (R&D intensity)
XRD scaled by sales Greve (2003a)
Chen and Miller (2007)
3. Independent variables
In order to examine performance feedback within firms, the relative difference between the
current performance and the aspiration level needed to be measured. Therefore, I
operationalized performance feedback by computing three related performance variables: a)
the firm performance, b) the aspiration level, and c) the performance adjusted by the
aspiration level.
Firm performance. The most popular measures of firm performance are return on equity (ROE), return on assets (ROA) and return on sales (ROS). Since ROE is affected by an
organization’s mix of equity and debt, it makes it difficult to compare it across firms. This makes using ROA or ROS the most preferable (Greve, 2003a). By following a large amount
of previous studies, I chose to use ROA as this proxy is used as the main accounting-based
proxy for firm profitability within the manufacturing industry (Bromiley, 1991; Greve, 2003a;
Lim & McCann, 2013).
Aspiration level. Based on prior research (Chen & Miller, 2007; Iyer & Miller, 2008; K. D. Miller & Chen, 2004), I used two complementary measures of aspiration level: historical
aspiration level and social aspiration level. Social aspiration level is measured by the median
profitability of companies within each industry (defined at the four-digit SIC level). Besides
that, the historical aspiration level is computed as mixture of past-period historical aspiration
level and the previous performance of the firm (Greve, 2003a) by taking an exponentially
weighted average of past values of the performance variable (Greve, 1998, Lant, 1992,
Performance adjusted by aspiration level. Eventually, performance feedback is measured by
the difference between the firm’s performance (ROA) and aspiration level (social and historical) (Iyer & Miller, 2008). To examine the effect of performance on risk taking when
performance is below the aspiration level, I split the performance variables into two
categories. Performance below the aspiration level equals 0 when performance is above the
aspiration level and equals performance minus the aspiration level when performance is
below the aspiration level. Similarly, performance above the aspiration level equals 0 when
performance is below the aspiration level and equals performance minus aspiration level
when performance is above the aspiration level (Audia & Greve, 2006).
To test for the influence performance feedback has on risk taking, moderated by TMT
characteristics, the independent variables were lagged by one year. The overview of the
definitions and operationalization of the independent variables representing performance
below the aspiration levels are outlined in Table 3 – Definitions of the Independent Variables
and Table 4 – Independent Variables – Operationalization.
Table 3 - Definitions of the Independent Variables
No Independent variable Definition Key reference
1. Performance below historical aspiration level
Performance below the smallest outcome, based on the firms’ own history, that would be satisfactory by the decision maker
Cyert and March (1963) Kim et al. (2014)
Levinthal and March (1981)
2. Performance below social aspiration level
Performance below the smallest outcome, determined by the performance of similar others, that would be satisfactory by the decision maker
Cyert and March (1963) Festinger (1954) Greve (1998b) Kim et al. (2014)
Table 4 - Independent Variables - Operationalization
No Independent variable Operationalization Key reference 1. Performance below historical
aspiration level
ROA minus the exponentially weighted average of past values of the performance variable, lagged by one year
Greve (1998)
Greve and Audia (2006)
2. Performance below social aspiration level
ROA minus the median profitability of companies within each industry (defined at the four-digit SIC level), lagged by one year
Greve (1998)
Greve and Audia (2006)
4. Moderator variables
For the purpose of this academic paper, TMT characteristics are used as moderating variables.
Following the existing literature, the sample of the firm’s TMT is defined as all the executives
with titles above the vice president or those who serve on the firm’s board of directors
(Carpenter, 2002; Finkelstein, Hambrick, & Cannella, 1996; Wiersema & Bantel, 1992). For
each TMT three different demographic characteristics, namely executive tenure, age and
gender, were expected to moderate the relationship between performance feedback and risk
taking within the firm (Hambrick & Mason, 1984; Homberg & Bui, 2013; Rost & Osterloh,
2010; Wahid, 2012; Wiersema & Bantel, 1992).
TMT tenure. Executive tenure represents the total of years a top manager has spent within
the focal organization (Finkelstein e.a., 2009; Pfeffer, 1985). Following Herrmann and Datta
(2005), I operationalized the tenure of top managers by subtracting the year of entrance from
the specific fiscal year focused on. The overall TMT tenure is operationalized by computing
the average tenure in years of all the members within the team (Finkelstein & Hambrick,
1990; Hambrick e.a., 2015; Herrmann & Datta, 2005; Heyden, Reimer, & Van Doorn, 2015).
TMT age. Furthermore, the age of each executive of the firm was taken within the specific fiscal years as reported in ExecuComp (Wahid, 2012). Following the existing
making within organizations (Goll e.a., 2001; Hambrick e.a., 2015; Herrmann & Datta, 2005;
Heyden e.a., 2015), I operationalized the overall TMT age by taking the mean age of all the
executives within the team.
TMT gender. Besides that, the gender of top managers within the firms was classified in
being ’male’ (0) or ‘female’ (1) (Hambrick e.a., 2015). I operationalized the overall gender
composition of the TMT as the proportion of females on the team, which is the number of
female executives divided by the total number of board members (Hambrick e.a., 2015;
Wahid, 2012).
To test for the moderating effect of TMT characteristics on the relationship between
performance below the aspiration level and risk taking, the moderator variables were lagged
by one year. The list of the definitions and operationalization of the moderator variables
concerning TMT characteristics are outlined in Table 5 – Definitions of the Moderator
Variables and Table 6 – Moderator Variables – Operationalization.
Table 5 - Definitions of the Moderator Variables
No Moderator variable Definition Key reference
1. TMT tenure The average years the TMT members has spent within the focal organization
Finkelstein et al. (2009) Pfeffer (1985)
2. TMT age The average age (in years) within the TMT Hambrick et al. (2015) Herrmann and Datta (2005) 3. TMT gender The gender composition within the TMT Hambrick et al. (2015)
Wahid (2012)
Table 6 - Moderator Variables - Operationalization
No Moderator variable Operationalization Key reference 1. TMT tenure The average tenure in years of all the
members within the TMT, lagged by one year
Finkelstein and Hambrick (1990)
Hambrick et al. (2015) Herrmann and Datta (2005)
