CEO Narcissism and The Firm’s Relative Exploration
Orientation: The Moderating Role of Environmental Dynamism
Sanne Posthumus · Supervised by Dr. Philip Steinberg and Prof. Dr. Pedro de Faria ·
Co-assessed by Prof. Dr. Jordi Surroca · Master Thesis · MSc BA Strategic Innovation
Management · Faculty of Economics and Business · University of Groningen · Student
number: S2960605 · Email: s.posthumus.3@student.rug.nl · 20
thof January 2020 · Word
count: 12,865
Keywords: CEO Narcissism | Exploration | Exploitation | Innovation | Environmental
Dynamism
Table of Contents
1. Introduction ... 3
2. Theoretical Background and Hypotheses ... 5
2.1 CEO Influence on Organizational Outcomes ...5
2.2 CEO Narcissism ...7
2.3 Exploration and Exploitation...8
2.4 CEO Narcissism and Innovation ...9
2.5 Hypotheses development ...10
2.5.1 CEO Narcissism, Exploration and Exploitation ...10
2.5.2 Moderating effect of Environmental Dynamism ...12
3. Methodology ...14
3.1 Sample and Data Collection ...14
3.2 Measurements ...15 3.2.1 Dependent Variable ...15 3.2.2 Independent Variable ...16 3.2.3 Moderator ...16 3.2.4 Control Variables ...17 3.3 Analytical Method ...18
4. Results ...18
4.1 Descriptive Statistics and Correlations...19
4.2 Regression Results and Hypothesis Testing ...21
4.3 Robustness Analysis ...22
5. Discussion and Conclusion ...25
5.1 Theoretical Implications ...27
5.2 Practical Implications ...28
5.3 Limitations and Future Research ...29
References...31
Appendices ...40
Appendix A: Variance Inflation Factors ...40
1. Introduction
According to the Upper Echelon Theory, the managerial background of a CEO influences organizational outcomes (Hambrick & Mason, 1984). Among others, the educational background, values, previous experiences and personality of the CEO are in part responsible for the firm’s strategy. Past research focused on the influence of demographic CEO characteristics on the organization, like age, education and tenure (e.g., Barker & Mueller, 2002; Bertrand & Schoar, 2003). Next to the demographic characteristics, the CEOs personality characteristics are also important to research. Because, these characteristics influence how CEOs process information, and are thus essential for the strategic decision making within the firm. An area which the CEOs personality can influence is the firm’s innovation orientation (Kashmiri, Nicol & Arora, 2017). CEOs with certain personality characteristics are more likely to focus on an explorative innovation orientation, whilst others prefer an exploitative innovation orientation.
The literature has identified different types of personality characteristics. One distinction that exists is that between the light triad and dark triad of traits. The triads consist of contrasting traits, where the light triad is positively related, and the dark triad is negatively related to life satisfaction. Positive traits are combined in the light triad. The light triad of traits consists of three facets; Kantianism (treating people as ends unto themselves), Humanism (valuing each individual), and Faith in Humanity (believing the human is inherently good) (Kaufman, Yaden, Hyde & Tsukayama, 2019). The dark triad consists of; Machiavellianism (the manipulative personality), Narcissism (sense of entitlement and need for attention) and Psychopathy (lack of empathy and guilt) (Paulhus & Williams, 2002). These light and dark traits have an influence on the feelings, behaviors and patterns of thoughts in our day-to-day lives (Kaufman et al., 2019). The upper echelon literature states that the dark triad trait narcissism is a personality characteristic often found in CEOs (Chatterjee & Hambrick, 2007). Narcissistic individuals have an ambition to attain influence, power (Kernberg, 1975) and attention (Chatterjee & Hambrick, 2007), which the CEO position offers. Having a CEO with a higher score on the narcissism trait, impacts the way they treat information, which in turn influences the way they make organizational decisions (Chatterjee & Hambrick, 2011).
Burgelman, 1994; Christensen, 1997). More recently, Kashmiri et al. (2017) find that firms with a CEO that scores high on narcissism introduces more radical innovations in their new product portfolio than firms with a CEO that scores lower on narcissism. Their study contradicts the studies of Burgelman (1994) and Christensen (1997) and states that characteristics of top management (in this case the CEO) does have an influence on firm innovation. They measure radicalness by studying new product introductions (NPI) and coding them as either more incremental or more radical. They argue that due to their overconfidence, more narcissistic CEOs believe in the success of their innovations, and are therefore more likely to introduce risky (radical) innovations. From this article, I conclude that CEO narcissism might lead to more risky innovative outcomes. But, it is unknown whether CEO narcissism plays a role in the extent to which firms define their innovation strategy as more explorative over exploitative (hereafter labeled as ‘relative exploration orientation’). Having a clearly defined innovation orientation stimulates alignment, committing the organization to a set of coherent behaviors, designed to realize competitive objectives (Pisano, 2015). Therefore, the goal of this study is to discover whether CEO narcissism influences the degree to which a firm defines its innovation orientation as relatively more explorative.
I argue that CEO narcissism is positively related to the firm’s relative exploration orientation. Because, CEOs that score high on narcissism want to be the center of attention, and are therefore more inclined to take risky actions. Furthermore, because of their feelings of superiority, they believe that they can easily manage risks (Raskin & Hall, 1979), and therefore apt to exploration. Thirdly, due to their over-confidence they are better at pitching their risky ideas and convincing others that exploration is the right direction for the firm. Lastly, CEOs with higher levels of narcissism are less willing to build on knowledge from previous CEOs, because this will not give them the needed attention and because they think this knowledge is inferior to their own. I also look at the moderating effect of environmental dynamism on the relationship between CEO narcissism and the firm’s exploration orientation. I argue that a more dynamic environment strengthens the relationship between CEO narcissism and exploration. Because, in such a dynamic environment, firms have to take more risky actions before the CEO receives the necessary attention. Furthermore, in dynamic environments CEOS are more likely to follow their internal compass, which is set on exploration. Lastly, building on knowledge from previous CEOs or directors (exploitation) is detrimental for firm performance in a dynamic environment.
by regressing time against yearly industry sales. I test my hypotheses using a random effects regression and do not find support for my hypotheses.
This research makes several contributions to research. First, I aim to contribute to the upper echelon literature field by clarifying the relationship between CEO characteristics and firm innovation, as previous research has reached opposing conclusions (e.g. Burgelman, 1994; Kashmiri et al., 2017). Furthermore, using Emmons (1987) four factors of narcissism, I explain the mechanisms that explain why a more narcissistic CEO might be more likely to choose an exploration orientation over an exploitation orientation. Thirdly, in an effort to enrich previous research of CEO narcissism, I examined the moderating effect of environmental dynamism. This has, to my knowledge, not been researched before as a moderator between CEO narcissism and firm innovation. Lastly, this study adds to the improvement of the use of computer-aided-text-analysis (CATA). Next to theoretical contributions, this study also has implications for managers. It will help practitioners understand one of the factors that might influence their firm’s explorative orientation. By understanding the effects of CEO narcissism on the firm’s relative exploration orientation, the board can compensate for this effect, so that the firm’s innovation strategy is optimal.
The remainder of the thesis is structured as follows. Chapter 2 gives a brief overview of the upper echelon theory in which I situate my research. Furthermore, in this chapter I discuss other relevant literature (CEO narcissism, exploration and exploitation). This literature acts as the foundation for the development of my hypotheses. Chapter 3 discusses the methodology of this research. It includes an explanation of the sample, data, measurements of the variables and the analytical method used. Chapter 4 presents the empirical results, as well as the results of the robustness tests. Finally, Chapter 5 concludes with a discussion of the findings, the theoretical and managerial implications and the limitations and suggestions for future research.
2. Theoretical Background and Hypotheses
In order to analyze how the level of CEO narcissism influences the firm’s relative exploration orientation, I integrate the upper echelon theory with March’s (1991) framework of exploration and exploitation. I start with an explanation of why my research belongs to the upper echelon theory. Then I elaborate on the concepts of CEO narcissism, exploration and exploitation, to conclude with the development of my hypotheses.
2.1 CEO Influence on Organizational Outcomes
to them. ‘’The heart of the theory is the portrayal of upper echelon characteristics as determinants of strategic choices and, through these choices, of organizational performance’’ (Hambrick & Mason, 1984, p. 197). From the upper echelon theory, three different streams have developed (Liu, Fisher & Chen, 2018).
The first stream analyzes which characteristics of the Top Management Team (TMT) have an effect on firm performance. Differences in the composition of the TMT lead to different choices and thus different outcomes (Cannella, Park & Lee, 2008). Characteristics that have been analyzed are the member’s functional backgrounds (Cannella et al., 2008), tenure (e.g. Finkelstein & Hambrick, 1990) and team size (Haleblian & Finkelstein, 1993). The findings of this stream state that having a diverse TMT, with for example members with different functional backgrounds or with different tenures, has both benefits and costs. Advantages of greater diversity within the TMT include developing more choice alternatives and a better assessment of the alternatives (Lant, Milliken & Batra, 1992). However, greater diversity also results in slower decision making and communication issues (Hambrick & D’Aveni, 1992).
The second stream examines the effects of CEO characteristics on the firm. Examples of individual characteristics researched are CEO background (e.g. Barker & Mueller, 2002), CEO demographics (e.g. Hitt & Tyler, 1991) and CEO personality characteristics (e.g. Chatterjee & Hambrick, 2007). This stream thus moves the focus from the entire TMT to the individual level of the CEO. Studies find that these different CEO characteristics all influence the firm. For example, younger CEOs with experience in marketing and/or R&D are more likely to take risky actions and spend more money on R&D (Barker & Mueller, 2002; Hitt & Tyler, 1991). Furthermore, bright and dark traits can both influence firm performance positively and negatively (Judge, Piccolo & Kosalka, 2009). Research within this stream either looks at the effect of CEO characteristics on strategic decision making in the firm (e.g. Katz, 1982; Miller & Toulouse, 1986; Lewin & Stephens, 1994), or on the effect of CEO characteristics on firm performance (e.g. Berson, Oreg & Dvir, 2008; Peterson, Galvin & Lange, 2012; Ou, Waldman & Peterson, 2018).
To summarize, each of these three streams have a different focus. The first stream examines how the entire TMT affects firm performance and does not focus on the CEO individually. The second stream, however, discards the TMT and only focuses on the characteristics of the CEO. The last stream combines the effects of the CEO and TMT and analyzes how individual characteristics of the CEO impact the actions of the TMT. This stream thus does not look at the direct effects of the CEO on the firm, but on the indirect effects of CEOs on the firm through the actions of the TMT. The literature stream that is helpful to study the relationship between CEO personality characteristics and the firm’s innovation strategy is the second stream. The second stream of literature analyzes the individual traits of the CEO and its effects on the firm. It is important to study individual CEO traits because the CEO is often seen as the most powerful individual in the organization (Harrison, Torres & Kukalis, 1988; Pearce & Robinson, 1987). They have substantial discretion in the strategic decision-making process (Chatterjee & Hambrick, 2011). Personality traits of CEOs have an effect on their thoughts, which in turn influences the firm’s strategic decision making.
2.2 CEO Narcissism
The CEO characteristic that I will focus on in this research is narcissism. Narcissism is defined as ‘’the degree to which an individual has an inflated sense of self and is preoccupied with having that self-view continually reinforced’’ (Chatterjee & Hambrick, 2007, p. 353). Previously, narcissism was seen as a mental disorder; people were either a narcissist or they were not. This was mainly due to the fact that measures of narcissism only focused on the extreme manifestation of narcissistic behavior, relating to pathological narcissism. Nowadays narcissism is seen as a personality dimension, on which people can score high or low (Emmons, 1987).
Research linking CEO narcissism to the firm has focused on organizational strategy and organizational outcomes. The factors of narcissism of Emmons (1987) explain how CEO narcissism influences organizational strategy and outcomes. The first factor, leadership/authority, states how the narcissist wants to be the center of attention. Because they want to be the center of attention, narcissistic CEOs are more inclined to take bold and risky actions that in their view will draw praise and attention. There are different examples of such bold actions. One is Corporate Social Responsibly (CSR). Petrenko et al. (2016) find that CEO narcissism is positively related to CSR, because CSR helps narcissistic CEOs draw praise and attention. Furthermore, Entrepreneurial Orientation (EO) is also positively related to CEO Narcissism (Wales et al., 2013). This is the extent to which organizations are considered to be entrepreneurial (Covin & Wales, 2012), which relates to the firm’s innovativeness, risk-taking and proactiveness. EO thus helps a firm create new products and services. Next to the positive influence of these bold actions, there are also actions that have destructive effects on the organization. Rijsenbilt and Commandeur (2013), for example, find that CEO narcissism is positively related to fraud. Because they have unrealistically high goals, CEOs with higher levels of narcissism might behave unethically to achieve them (Duchon & Drake, 2009). Furthermore, the superiority/arrogance factor states that narcissists think they are better than others. Due to this feeling, it can again be explained why they are more likely to indulge in fraud; because they think they will not get caught, invest more in CSR; because they are morally superior, and exhibit higher levels of EO; because they think they will not fail. They believe they can take risky actions free of cost, and will succeed anyway, because they are superior to others. Also, Zhu & Chen (2015) find that narcissistic CEOs are more likely to be influenced by corporate strategies they have experienced before in other firms. But, less likely to be influenced by corporate strategies experienced before by their directors. As they want to show their superiority, next to opposing the experience of the director, they select a strategy opposite from the strategy proposed by the director.
The literature thus describes both negative and positive effects of CEO narcissism on the organization. As CEOs with higher levels of narcissism desire attention and feel superior to others, they take bold and risky actions because they think these actions can be operated free of cost. Sometimes these actions are positive for the organization and society, but sometimes they are detrimental.
2.3 Exploration and Exploitation
analyzing new possibilities, which comes with extensive distant knowledge search. It involves risky activities because outcomes are not known in advance. Alternatively, exploitation focuses on the already existing knowledge in the firm and old certainties, to carry out improvements (He & Wong, 2004).
Exploration improves a firm’s innovative behavior and long-term performance, whilst exploitation benefits a firm’s efficiency and short-term performance. Both are important for the firm, exploration to ensure long-term survival and exploitation for current prosperity. Therefore, it is the firm’s ultimate goal to balance exploration and exploitation. This activity is known as organizational ambidexterity (e.g. Duncan, 1976; Tushman & O’Reilly, 1996; Jansen, Van Den Bosch & Volberda, 2005). However, firms often have difficulties excelling in both. Tensions arise when exploration and exploitation activities compete for scarce resources. Furthermore, the processes and routines that are essential for exploration activities are different from those essential for exploitation activities. This is why the past processes and routines of a firm tend to keep developing in the same direction (Miller & Friesen, 1980), either in the direction of exploration or of exploitation. These tensions mainly arise due to the diverse knowledge management processes of the two strategies (March, 1991). It is thus considered that exploration and exploitation are two ends of a continuum (Gupta, Smith & Shalley, 2006; Lavie, Stettner & Tushman, 2010). Following this premise, this study focuses on the firm’s innovation orientation concentrated on exploration relative to exploitation. Firms with a high relative exploration orientation mainly focus on explorative activities, which means that they concentrate their resources, processes and routines on new knowledge search, products and services. Whilst firms with a low relative exploration orientation mainly focus on exploitative activities, such as further developing already known knowledge, products and services. As the choice between either focusing on exploration or on exploitation is a strategic decision, the CEO of a firm plays a major role. Therefore, it is important to look at the influence of his or her personality.
2.4 CEO Narcissism and Innovation
Only a few studies link CEO narcissism with firm innovation (e.g. Gerstner et al., 2013). Usually, these studies focus on the effects of narcissism on creativity, which is considered as the starting point of innovation (West, 2002). Individuals that possess the creativity trait can help spur innovation. Therefore, the creative process of the CEO is very important to an organization.
Slevin, 1989). ‘’Narcissism is a spur in risk taking and innovation; when others are timid, the narcissist acts on his or her supreme confidence and craving for applause’’ (Gerstner et al., 2013, p. 281). Narcissistic CEOs are perceived as more creative, and as creativity is considered as the starting point of innovation, they are thus seen as more innovative.
The few studies that link CEO narcissism to innovation find that narcissistic CEOs tend to overinvest in R&D (Ham, Seybert & Wang, 2018), due to their tendency to take risky actions and feelings of superiority. Furthermore, they are aggressive in adopting technological discontinuities (Gerstner et al., 2013) because they expect that these actions will lead to admiration and attention from others. However, most of these studies tend to neglect the existence of different types of innovation. One study that does look at different types of innovation is that of Visser & Faems (2015). They find that the cognitive characteristics of CEOs influence strategic decision making on dividing resources amongst explorative and exploitative activities and firm performance. My research differs from that of Visser & Faems (2015) in that I analyze a specific personality characteristic, narcissism, and its influence on the firm’s relative exploration orientation.
2.5 Hypotheses development
2.5.1 CEO Narcissism, Exploration and Exploitation
More narcissistic CEOs want to be the center of attention, as this will give them approval and admiration from others. In order to attain this attention, they need to take bold and risky actions that are noticed by others (Emmons, 1987). Exploration, the search for new knowledge and information from outside the company, will help attain this attention. With the help of this new knowledge and information, the firm is better able to create new and innovative products and services that are noticed by others. Next to narcissistic CEOs taking these bold actions to receive the attention from others, they also need a personal sense of surprise and excitement, which these bold moves offer (Jaussi & Dionne, 2003). Exploitative, incremental behaviors (like improvements and refinements) on the other hand, seem uninteresting and ordinary to the narcissist (Drucker, 1992). Thus, exploration is more attractive to the narcissistic CEO than exploitation (Zhang, Ou, Tsui & Wang, 2017), because exploration gives them the needed praise and attention from others.
especially concerned with the risks associated with such discontinuous innovations. They are constantly driven to improve products and processes incrementally (Vera & Rodriguez-Lopez, 2004), and therefore prefer an exploitative innovation orientation (Owens & Hekman, 2012; Kashmiri et al., 2017). The reason why CEOs with higher levels of narcissism have such a different view of risk is because they are well equipped to quickly recover firm performance after risky situations, like an economic crisis or organizational failure (Chatterjee & Hambrick, 2007; Zhu & Chen, 2015). For example, Patel and Cooper (2014) find that in the pre-crisis period, firm performance of firms with a more narcissistic CEO declines, whilst firm performance increases in the post-crisis period. More narcissistic CEOs are less likely to recognize potential threats, but they are skilled in helping their firm overcome such threats. Exploration concerns more risky activities, and therefore the narcissistic CEO is more interested in exploration than in exploitation.
Thirdly, more narcissistic CEOs are better at convincing others of their ideas concerning the firm. CEOs cannot make strategic decisions all by themselves. They are restricted by institutional and environmental forces (Chatterjee & Hambrick, 2011). For example, they need approval from their board of directors. More narcissistic CEOs are better able at convincing their creative and innovative ideas to others and pitching them, than others (Goncalo et al., 2010; Wisse et al., 2015). Certain characteristics of narcissism explain this mechanism. First, as CEOs admire themselves and feel a sense of entitlement, they are extremely confident about their ideas. They really believe that it is the best for the firm. Secondly, narcissists are more extraverted. Leaders that are more extraverted are more forceful and convincing in communicating their opinions (Judge, Bono, Ilies & Gerhardt, 2002). More narcissistic CEOs can thus use their convincing power to influence strategic decisions. Lastly, people with higher levels of narcissism have high charisma and compelling rhetoric with which they can inspire others (Eisenmann, 2013). As more narcissistic CEOs favor exploration over exploitation, they will use their convincing power to encourage explorative activities.
directors are not important to the CEO with higher levels of narcissism. Thus, CEOs with higher levels of narcissism are less inclined to focus on a relative exploitation orientation.
CEOs with higher levels of narcissism have certain traits that explain why they are more likely to define their innovation orientation as more explorative than exploitative. First of all, more narcissistic CEOs want to be the center of attention, which is why they take more risky actions than less narcissistic CEOs. Secondly, more narcissistic CEOs feel superior to others, which is why they think their risky and bold actions can be operated free of cost. Thirdly, due to their self-admiration, combined with their extraversion and openness, more narcissistic CEOs are better at pitching their risky actions and convincing others that it is the right direction for the firm. Lastly, CEOS with higher levels of narcissism are less willing to build on knowledge from previous CEOs, because this will not give them the needed attention and because they feel like this knowledge and ideas are inferior to their own.
H1: CEO narcissism is positively related to the firm’s relative exploration orientation.
2.5.2 Moderating effect of Environmental Dynamism
As stated above, there are both theoretical and empirical grounds that suggest that CEO narcissism enhances the firm’s relative exploration orientation. However, it is important to note that these effects can be influenced by opportunities and threats from the environment (Wang & Li, 2008). The environment provides top management with information, necessary for strategic decisions (Duncan, 1972). In this way, the environment can influence the actions taken by top management. Therefore, this study will analyze the potential moderating effect of environmental dynamism.
Environmental dynamism is the extent to which environments are subject to high levels of uncertainty. It concerns both the degree of instability as the rate of change of the environment (Dess & Beard, 1984). The more dynamic the environment in which a firm operates is, the greater the change in customer preferences and behavior (Child, 1972; Dess & Beard, 1984). This makes it more difficult to anticipate future events and actions (Khandwalla, 1977).
organizations that pursue exploitation are likely to fall behind competition. These organizations continue to improve routines that are less valued by the environment (Sørenson & Stuart, 2000). This is detrimental for their firm’s performance.
More narcissistic CEOs want to be the center of attention, which is the reason why they take grandiose actions. Narcissistic leaders are inclined to be even more motivated to pursue certain strategies in situations in which they are known to achieve success (Wallace & Baumeister, 2002). Thus, they are more inclined to pursue an explorative innovation orientation in those environments where this strategy leads to success. A dynamic environment is such an environment. As in dynamic environments there is a high risk of obsolescence due to high competition, explorative innovation is seen as an indication of success. Only with new products, services and knowledge, a firm is able to stay ahead of competition. This environment thus forces a CEO with higher levels of narcissism to show even more explorative than exploitative innovative behavior, as this is the only strategy that will give the CEO the necessary attention and applause.
CEOs with higher levels of narcissism feel superior to others, and are therefore less likely to notice risks, especially in dynamic environments. First, they evaluate themselves very positively and feel like they are able to control their environments, instead of their environments controlling them (Campbell et al., 2004). Second, when environmental dynamism is high, CEOs with these positive self-evaluations are more likely to take risks (Simsek, Heavey & Veiga, 2010). They are more likely to depend on their own internal compass to make decisions in uncertain and fast-changing environments, because they believe they know what is best for the firm. Thus, these CEOs have a preference towards more innovative and risk-taking behavior, especially in dynamic environments. Furthermore, CEOs with high scores of self-efficacy tend to feel less threatened by dynamic environments (Westerberg, Singh & Häckner, 1997). Therefore, in dynamic environments, narcissistic CEOs are more likely to take risky actions, and thus choose exploration over exploitation.
Lastly, CEOs that score higher on narcissism are less willing to build on knowledge from previous CEOs or directors, especially in dynamic environments, because doing so is detrimental for the firm. As stated above, exploitative actions, like improving routines, make the firm fall behind competition (Sørenson & Stuart, 2000) and will not give the CEO praise and applause. Furthermore, in such uncertain environments, CEOs with higher narcissism scores become more focused on their own internal compass. This is because, overconfidence is most likely to appear when decision alternatives are ambiguous (as is the case in dynamic environments) versus clear (as is the case in stable environments) (Klayman, Soll, Gonzalez-Vallejo & Barlas, 1999).
explorative actions. Lastly, building on previous knowledge (exploitation) in dynamic environments is detrimental for firm performance, and therefore the CEO with higher levels of narcissism is even less likely to pursue it.
H2: Environmental dynamism moderates the relationship between CEO narcissism and the firm’s relative exploration orientation, such that this relationship is stronger in dynamic than in stable environments.
3. Methodology
This chapter explains the choices concerning the methodology of this research. First, I describe the sample of the study. Then, I give a detailed description of the variables and their measurements. I conclude this section with an explanation of the analytical method I used to test my hypotheses.
3.1 Sample and Data Collection
The sample consists of the Standard & Poor’s (S&P) 100 companies list from 2008 to 2018, consisting of 150 companies in total. I already deleted three companies in the beginning of data collection as they went bankrupt or were acquired by another firm in 2008. Therefore, there was no data available of these companies. The S&P 100 is a subset of the S&P 500 and is comprised of the 100 largest public firms in the US stock market. Research on CEO characteristics asks for a sample of companies that have information readily available. As the S&P 100 consists of the largest public companies in the US, a wide array of information is available on both the CEOs and the company. Like for example, elaborate annual reports with a letter from the CEO and the transcripts of earnings calls. Therefore, the S&P 100 is a suitable sample for this research.
used the company websites. We tried to collect annual reports for all the companies, as these include photos and graphics. This was necessary to collect data on the CEO photograph (which I used as a robustness test). When annual reports were not available, we collected the 10-K reports instead. In total, we collected 1,661 annual reports/10-K reports. We had to rename these files as well, to be able to trace them back to the right company and year. As mentioned above, from these reports we collected data on the size of the CEO photograph and whether the CEO was also the chairman of the board (CEO duality). Lastly, for firm age we googled the date of foundation of the firm.
When all the data were collected, I merged the different excel files in Stata. After running the final regression, the final sample consists of 130 publicly listed companies with 1,067 observations.
3.2 Measurements
3.2.1 Dependent Variable
In order to measure the degree to which the company has an explorative or an exploitative innovation orientation, I use the adapted measure of Uotila et al. (2009), following previous research (Heyden, Oehmichen, Nichting & Volberda, 2015; McKenny, Aguinis, Short & Anglin, 2018). Uotila et al. (2009) created a measure of exploration and exploitation using computer-aided-text-analysis (CATA). CATA counts how frequent specific words belonging to a construct are used in the text. This creates a quantitative measurement of the construct of exploration and exploitation (Short, Broberg, Cogliser & Brigham, 2010).
I chose the measure of Uotila et al. (2009) for different reasons. First, the search terms that are used in such a content analysis include more of the firm’s exploration and exploitation orientation than survey or patent data. This makes the measure more generalizable (Uotila et al., 2009). Secondly, content analysis is useful for a large-scale longitudinal research, because news documents are widely available over time (Uotila et al., 2009). Thirdly, CATA has been suggested to be a valuable measurement approach with the potential to lead to meaningful advancements in theory (McKenny et al., 2018).
The textual source I used for the content analysis are the firm’s annual reports from 2008 to 2018. I used the annual reports for the exploration and exploitation measure because annual reports address the companies’ activities and views of the past and upcoming year. In these reports, the firm’s innovation orientation is discussed, which contains information on exploration and exploitation activities.
process the annual reports and analyze the number of words concerning exploration and exploitation. I measured the firm’s relative exploration orientation, by dividing the data on exploration by that of exploration plus exploitation (Uotila et al., 2009).
3.2.2 Independent Variable
Measuring the personality of an executive is extremely difficult to conduct. Executives are not willing to do psychological tests or fill out surveys about their personalities (Cycyota & Harrison, 2006). Therefore, I use a promising unobtrusive indicator of personality in this research. A text analysis of the transcripts of quarterly earnings calls (Raskin & Shaw, 1988).
Raskin and Shaw (1988) state that an individual’s relative use of first-person singular pronouns to first-person plural pronouns is correlated with the Narcissistic Personality Inventory (NPI) (Raskin & Hall, 1979). The NPI score is the most established test to identify narcissistic individuals (Rhodewalt & Morf, 1995). Therefore, the measure of Raskin and Shaw (1988) is a valid measure of CEO narcissism. The measure relies on unplanned and spontaneous speech, instead of on predetermined speech. Therefore, the documents that we used for the text analysis of CEO narcissism are the transcripts of the companies’ quarterly earnings calls. We selected these transcripts as the CEO is able to speak freely during these calls, especially during the question-and-answer sessions with investors. These sessions allow a degree of spontaneity of response of CEOs, which reflects their personalities. We used the LIWC program again to process the transcripts of earnings calls and to analyze the number of first-person singular pronouns and first-first-person plural pronouns. Following previous research, the final narcissism score is measured as the use of first-person singular pronouns (I, me, my, mine, myself) divided by the use of both first-person singular pronouns and first-person plural pronouns (I, me, my, mine, myself, we, us, our, ours, ourselves) (Aktas, Bodt, Bollaert & Roll, 2016; Raskin & Shaw, 1988). In order to obtain an annual measure, we used the mean of the data of the four quarters.
3.2.3 Moderator
industry sales, I summed the total sales of each industry category, by groupings based on the first two digits of the NAICS codes. These first two digits symbolize the top-level industry classifications.
For the regression, I entered time (2008-2018) as the independent variable and annual sales as the dependent variable for each industry category. Then, I divided the standard error of the regression coefficient by the mean sales value of the 11 years. This measure shows the dynamism of sales in each of the different industries. Because the raw data for environmental dynamism were skewed to the right, I took the natural logarithm. This measure was also used in past research (e.g. Hmieleski & Baron, 2009, Baron & Tang, 2011).
3.2.4 Control Variables
To control for potential confounding factors, I included different CEO control variables and firm control variables based on prior research. For variables that were subject to skewness, I used the natural logarithms.
CEO age. The first control variable is CEO age. The degree to which a CEO takes risky,
explorative strategic decisions, depends on age. Younger CEOs still have the energy to perform, are less attached to the firm and are more likely to take risks than older CEOs (Serfling, 2014). Older CEOs tend to be more conservative (Hambrick & Mason, 1984), risk-averse (MacCrimmon & Wehrung, 1986) and tend to follow lower-growth strategies (Child, 1974). Thus, younger CEOs are more likely to focus on exploration, whilst older CEOs are more likely to follow an exploitative orientation.
CEO tenure. CEOs tend to make fewer adjustments in the firm’s strategy as their tenure
increases (Grimm & Smith, 1991), which may lead to more exploitation than exploration over time. Furthermore, more tenure CEOs are less inclined to take risky actions (Wales et al., 2013). As the novelty of the CEO function decreases, the CEO loses his or her interest in organizational change (Hambrick & Fukutomi, 1991). CEO tenure is measured as the number of years the CEO is in function.
CEO structural power. With this variable I code whether the CEO is also a board chairman
(Chatterjee & Hambrick, 2007). CEO duality gives a CEO greater power which they can use for opportunistic behavior and pursuing their own agenda (Rijsenbilt & Commandeur, 2013; Adams, Almeida & Ferreira, 2005). Thus, they can use this power to pursue their orientation of exploration or exploitation.
CEO gender. Previous research found that female leaders make different strategic decisions
than male leaders (Chen, Crossland & Huang, 2016). For example, women tend to be more risk averse than men (Zhang et al., 2017), and are therefore less inclined to focus on exploration.
Firm age. Firm age can have an effect on risk-related decisions (Finkelstein, Hambrick &
Firm size. Firm size is included as a control variable and is measured as number of employees.
Firm size is stated to have an effect on the firm’s innovation performance (Zhang et al., 2017). Larger firms have for example access to more financial resources, but are also more bureaucratic, which has a negative effect on innovation (Kashmiri et al., 2017). A larger amount of resources can be beneficial to firm exploration, but bureaucracy can on the other hand lead to greater exploitation. The variable is measured as the natural logarithm of the number of employees (Cao, Gedaljlovic & Zhang, 2009).
Year dummies. Year dummies are included to control for unobserved time effects (Uotila et al.,
2009). This creates 11-year dummies.
Industry dummies. To control for potential industry effects I included industry dummies, as the
2-digit NAICS code (Visser & Faems, 2015). This creates 17 industry dummies.
3.3 Analytical Method
When a sample consists of panel data, two types of analysis are often used. These are; the fixed-effects panel model and the random-fixed-effects panel model. In the fixed-fixed-effects model, the individual-specific effect is a random variable, permitted to be correlated with the explanatory variables. This model studies the causes of changes within an entity (in this case, a firm). In the random-effects model, the individual-specific effect is also a random variable, uncorrelated with the explanatory variables. This model studies the causes of changes across an entity (in this case, a firm) (Angrist & Pischke, 2008).
As my explanatory variable (CEO narcissism) is very stable over time for some firms that have the same CEO for a large number of years, I was prevented from using the fixed-effects approach (Buyl, Boone & Wade, 2017). In such a case, a random-effects approach is more suitable. As explained above, the effects model shows the causes of change across firms. Therefore, in my study the random-effects approach shows how different levels of CEO narcissism lead to different levels of exploration across the different firms in my sample.
I also performed a Hausman Test (Hausman & Taylor, 1981) in Stata to check whether the fixed effects or the random-effects panel regression was the most suitable. The null hypothesis was that the preferred model is random-effects versus the alternative fixed-effects. The result was not significant which confirms that the random-effects model is the preferred analysis.
4. Results
4.1 Descriptive Statistics and Correlations
Table 1 shows the summary statistics, including the means, standard deviations, minimum values and maximum values per variable. The final sample consists of 130 firms with 1,067 observations for all of the variables, except for CEO Picture (used in the robustness analysis). For the variable CEO picture, there were only 766 observations. From the summary statistics for the relative
exploration orientation, CEO narcissism and environmental dynamism, I do not observe noticeable
results. From the mean of CEO picture (μ = 2.688) I can state that most CEOs have a bigger picture of them alone in the annual reports. From the control variables I conclude that CEOs are on average older males, with some years of tenure. Furthermore, most CEOs are also the chairman of the board.
Table 2 presents the correlations of the measures. Overall, I do not observe high correlations (r < 0.5). The highest correlation is that between CEO tenure and CEO age (r = 0.240), which is still a relatively weak relationship. Furthermore, it can be easily explained as the tenure of a CEO increases with his or her age. The second highest correlation is an unexpected one, that between CEO age and
environmental dynamism (r = 0.192). But, as it is a weak relationship as well, it is of no concern. None
of the correlations between the measures exceed the benchmark of 0.7 (Luger, Raisch & Schimmer, 2018).
Table 1: Descriptive Statistics
Variable N Mean S.D. Min. Max.
(1) Relative Exploration Orientation 1,067 0.495 0.171 0 0.891 (2) CEO Narcissism 1,067 0.169 0.038 0.037 0.315 (3) CEO Picture 766 2.688 0.848 1 4 (4) Environmental Dynamism 1,067 0.012 0.026 0.004 0.195 (5) CEO Age 1,067 57.246 5.673 28 77 (6) CEO Tenure 1,067 6.323 5.122 1 36 (7) CEO Duality 1,067 0.654 0.476 0 1 (8) CEO Gender 1,067 0.071 0.257 0 1 (9) Firm Age (ln) 1,067 3.961 0.956 0 5.420 (10) Firm Size (ln) 1,067 10.734 1.236 6.745 13.381 Notes: ln = natural logarithm
Table 2: Correlations
Variable (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
(1) Relative Exploration Orientation 1
4.2 Regression Results and Hypothesis Testing
To test my hypotheses, I performed a random effects model. The results are reported in Table 3. Hypothesis 1 (Model 2) suggests that firms that have a CEO with higher levels of narcissism have a more explorative than an exploitative innovation orientation. Therefore, I expect that when CEO narcissism is higher, the relative exploration orientation of that firm is also significantly higher. Hypothesis 2 (Model 3-4) suggests that environmental dynamism positively moderates the relationship between CEO narcissism and the firm’s relative bigger focus on exploration. Therefore, I expect that when the dynamism in the environment is higher, higher levels of CEO narcissism lead to even more exploration.
Model 1 only includes the control variables and the dummies for industry and years. The results
provide an R2 of 0.481. Furthermore, the control variable CEO tenure (β = -0.002, p =0.004) exhibits a
significant result. Meaning that for each increase of CEO tenure, a decrease equivalent to the coefficient is observed in the firm’s relative exploration orientation. Next, the control variable firm age (β = 0.015, p =0.072) also exhibits a significant result. Meaning that for each increase in firm age, an increase equivalent to the coefficient is observed in the firm’s relative exploration orientation, which is a different result than expected for firm age.
Model 2 is used to test hypothesis 1. It includes the independent variable CEO narcissism, the
control variables and the dummies for industry and years. The results provide an R2 of 0.481, which is
the same R2 as in Model 1. This means that Model 2, which includes the independent variable – CEO
narcissism – does not explain more of the variability in the dependent variable, than the model with only control variables. The coefficient for CEO narcissism is negative, where I expected it to be positive, and insignificant (β = -0.095, p >0.1). Therefore, the model does not provide empirical support for the first hypothesis. Again, the control variables CEO tenure (β = -0.002, p =0.005) is significant and firm age (β = 0.015, p =0.067) exhibits a somewhat significant result.
Model 3 adds the moderator, environmental dynamism, to the analysis. But, not yet as an
interaction effect, just as another explanatory variable. This model has an R2 of 0.481, the same as in
Model 1. By adding environmental dynamism as a variable, the results for CEO narcissism do not
change (β = -0.095, p >0.1), compared to Model 2. Environmental dynamism exhibits significant results (β = 0.330, p =0.042), which means that for each increase in environmental dynamism, an increase equivalent to the coefficient is observed in the firm’s relative exploration orientation. Thus, environmental dynamism has a direct effect on the firm’s relative exploration orientation. Furthermore, the results of the control variables remain the same as in Model 2, where CEO tenure (β = -0.002, p =0.005) and firm age (β = 0.015, p =0.067) exhibit significant results.
Model 4 adds the interaction effect between CEO narcissism and environmental dynamism to
the model, which I use to test hypothesis 2. This model again has an R2 of 0.481, just like the previous
models. Model 4 is the model which includes all of the variables. But, as the R2 did not increase whilst
do not explain a larger percentage of the variance in the dependent variables. The coefficient of CEO narcissism becomes positive, but remains insignificant (β = 0.022, p >0.1). Environmental dynamism still shows a positive and significant coefficient (β = 0.326, p =0.048). The interaction effect shows the moderating role of environmental dynamism on the relationship between CEO narcissism and relative exploration orientation. It shows a positive but insignificant coefficient (β = 0.024, p >0.1). Therefore, the model does not provide empirical support for the second hypothesis. The control variables CEO tenure (β = -0.002, p =0.005) and firm age (β = 0.016 p =0.067) remain significant.
4.3 Robustness Analysis
In order to test the robustness of my analysis I perform two additional robustness checks.
For the first robustness check, I use a different unobtrusive measure for CEO narcissism than the text analysis from the transcripts of the earnings calls. Instead I used the prominence of the CEOs photograph in the firm’s annual reports (Chatterjee & Hambrick, 2007). CEOs use their power to impact and control how they are portrayed in annual reports (Chatterjee & Hambrick, 2007). More narcissistic CEOs want to have a larger picture of themselves in the annual report, as this draws attention to them. We measured this construct using a 4-point scale, as also used in the study by Chatterjee and Hambrick (2007). We assigned 4 points when the CEOs photo was of him or her alone and occupied more than half a page; we assigned 3 points when the CEOs photo was of him or her alone and occupied less than half a page; we assigned 2 points when the CEO was photographed together with one or more fellow executives; we assigned 1 point when there was no photograph of the CEO.
This measure can be found in Model 5, together with the control variables and dummies for industry and years. The model has an R2 of 0.532. The coefficient for the CEO picture is negative and
insignificant (β = -0.020, p >0.1), and therefore does not provide empirical support for hypothesis 1. Next, environmental dynamism remains positive and significant (β = 0.354, p =0.045). The interaction effect of environmental dynamism is negative and insignificant (β = -0.005, p >0.1), and therefore does not provide empirical support for hypothesis 2. Next to the control variables CEO tenure (β = -0.002, p =0.007) and firm age (β = 0.019, p =0.084) that were already significant in the previous models, in
Model 5, CEO gender is also significant (β = 0.039, p =0.033).
If I compare the results of Model 5 with the results of Model 4, I find that in both models the independent variable CEO narcissism is insignificant. Furthermore, CEO narcissism has the predicted positive coefficient in Model 4, but has a negative coefficient in Model 5. Additionally, the interaction effect of environmental dynamism is insignificant in both Models. In Model 4, the interaction effect has the predicted positive coefficient whilst in Model 5 the interaction effect has a negative coefficient.
interaction between CEO narcissism and environmental dynamism. The results can be found in Appendix B.
Table 3: Regression Results Hypothesis 1 and Hypothesis 2
Variable Model 1 Model 2 Model 3 Model 4 Model 5
5. Discussion and Conclusion
This final chapter discusses the conclusions of the empirical results. I start with a brief summary of the research and results. Then, I give both theoretical and methodological reasoning that can explain my insignificant results. An explanation of the theoretical and practical implications will follow. I conclude my research with the limitations of the study and suggestions for future researchers.
Within the upper echelon literature, researchers have become interested in the influence of CEO characteristics on firm decisions and outcomes. Initial research analyzed the effects of demographic characteristics of the CEO on the firm (e.g. Barker & Mueller, 2002; Bertrand & Schoar, 2003). Nowadays scholars are also examining the effects of personality characteristics of the CEO (e.g. Chatterjee & Hambrick, 2007). One of the personality characteristics that is the subject to a growing body of research is narcissism. Because narcissists want to attain influence and power (Kernberg, 1975), they are more likely to aspire for the CEO position. Therefore, many CEOs of large firms are said to have higher levels of narcissism than average people (Chatterjee & Hambrick, 2011). Even though the amount of research on CEO narcissism is increasing, the literature lacks an understanding of the effects that higher levels of CEO narcissism have on the firm’s relative exploration orientation. Therefore, the goal of this research was to discover this effect.
I argue that firms that have a CEO with higher levels of narcissism orientate their innovation strategy more towards exploration than to exploitation. There are different reasons that explain this effect. More narcissistic CEOs want to be the center of attention, and therefore take bold and risky actions. Furthermore, they feel superior to others and therefore believe their risky actions can be taken free of cost. Next, they are very skilled at convincing others that their risky actions are the right path for the company. Lastly, CEOs with higher levels of narcissism are less willing to build on knowledge from previous CEOs. Furthermore, I argue that environmental dynamism positively moderates the relationship between CEO narcissism and exploration. Such that when environmental dynamism is high, CEO narcissism leads to more exploration.
By using a random-effects model, I empirically test my hypotheses using a panel dataset over a time period of 2008 to 2018, with a sample of 130 firms, resulting in 1,067 observations. Counter to my expectations, my data did not provide support for my hypotheses. There are both theoretical and empirical reasons that can explain these results. I start with the possible explanations of the insignificant results of my first hypothesis.
arguments that imply that more narcissistic CEOs might be more likely to have a relative exploitation orientation. First, CEOs with higher levels of narcissism feel like they are very good at correctly evaluating situations and environments (Klayman et al., 1999). This drives them to stick to a certain route concerning the firm’s strategy (Picone, Dagnino & Minà, 2014), which is thus more exploitative in nature than explorative. Furthermore, Hiller & Hambrick (2005) find that CEOs that score high on narcissism tend to have very positive self-evaluations and therefore favor repetition of previous successful decisions. Secondly, overconfident CEOs tend to prefer a quick decision-making process (Hambrick, Cho & Chen, 1996) based on previous successful experiences.
A second possible explanation for the insignificant results of the first hypothesis is that as CEOs with higher scores of narcissism gain more attention from the outside they become fearful of losing their success. The successful narcissist sees everything as a threat (Maccoby, 2004). Therefore, as CEOs with high levels of narcissism grow more successful, they might reduce their risk-taking and prefer the repetition of prior successful strategies.
Third, CEOs with higher levels of narcissism believe in their own superiority. Therefore, they are less inclined to listen to the ideas of others in the firm. This may lead to an outflow of talent, ideas and best practices, which can have a detrimental effect on the firm (Engelen, Neumann & Schmidt, 2016). Instead of acknowledging different points of view and discovering new sources of knowledge, which drives exploration (Levinthal & March, 1993), they surround themselves with a group of yes-men (Maccoby, 2004), that will always agree with them. This does not encourage the explorative orientation of the firm.
After theoretically reconsidering the first hypothesis, the arguments above give reasons why higher levels of CEO narcissism might not necessarily lead to a higher relative exploration orientation of the firm. But, that it could also in some cases lead to a higher relative exploitation orientation. However, some explanation for the insignificance of the results might also be found in the measurement of the independent variable, CEO narcissism. To measure the level of CEO narcissism, I used the individuals relative use of first-personal singular pronouns (Raskin & Shaw, 1988) in the transcripts of the firm’s earnings calls. Because, during these earnings calls, the CEO is able to speak freely. However, due to time constraints we were unable to filter out only the part of the transcript where the CEO was speaking. Therefore, our transcripts might have also included the speech of additional executive officers, which makes this a less accurate measure of CEO narcissism.
Secondly, with high levels of environmental dynamism, the CEO experiences that their environment becomes more uncertain and difficult to predict. Research found that overconfident CEOs are more likely to stubbornly stick to strategic decisions and processes that have worked well in the past (exploitation) in dynamic environments (Hmieleski & Baron, 2008). They undermine the importance of new information (Kroll, Toombs & Wright, 2000). Following this research, environmental dynamism may again weaken the relationship between CEO narcissism and relative exploration. However, Ilieva, Brudermann and Drakulevski (2018) find no significant relationship between environmental dynamism and overconfidence, meaning that according to their study, overconfidence does not increase in a dynamic environment. In another way, this study also explains why environmental dynamism might weaken the relationship between CEO narcissism and the firm’s relative exploration orientation. As the confidence of narcissistic CEOs does not necessarily increase in dynamic environments, they might be less likely to take risky actions in these environments. This decreases their firm’s relative exploration orientation.
Whereas the moderation effect of environmental dynamism was not found, it is found that environmental dynamism has a direct positive effect on the firm’s relative exploration orientation.
To conclude, after an extensive literature research I came to the hypotheses which I presented in this research. But, after further consideration, there is also evidence that suggests that the relationship might not exist, or that environmental dynamism does not strengthen, but weaken the relationship between CEO narcissism and relative exploration orientation.
5.1 Theoretical Implications
relative exploitation or using different measures for the variables. More elaborate suggestions for future research are given in the last section of this research.
Furthermore, this paper argues that CEOs with higher levels of narcissism lead to a higher relative exploration orientation of the firm. This is mainly because more narcissistic CEOs are more likely to take risky actions. Using the four factors of Emmons (1987) as the foundation, I conceptualize four mechanisms that clarify why more narcissistic CEOs take more risky firm actions.
Thirdly, in an effort to develop previous research that linked CEO narcissism to the firm, I also analyzed the moderating effect of environmental dynamism. Previous research has examined the moderating effect of environmental dynamism on the relationship between exploration and performance (Jansen et al., 2006; Kim & Rhee, 2009), or the moderating effect of marketing department power and customer orientation on the relationship between CEO narcissism and firm exploration (Kashmiri et al., 2017). But, this study is the first that applies environmental dynamism as a moderator between CEO narcissism and the firm’s relative exploration orientation. As the environment supplies the CEO with information necessary for strategic decisions (Duncan, 1972), the dynamism thereof can influence the process of how the CEO analyzes this information. The results are inconsistent with previous research that found that in a dynamic environment exploration is a better innovation orientation than exploitation (e.g. Sørenson & Stuart, 2000). This means that even though environmental dynamism might lead to more explorative innovation outcomes for firms that have a more narcissistic CEO, this does not translate to the firm’s innovation orientation.
Furthermore, my research adds to the improvement of the use of CATA. In this research I built on the work of Uotila et al. (2009) and Raskin and Shaw (1988). In my research I used CATA for my dependent variable, using annual reports, and also for my independent variable, using transcripts of quarterly earnings calls. To my knowledge, there are no studies that use text analysis for both of these variables simultaneously. Furthermore, in order to omit the problems associated with using the same data source for both the dependent variable as the dependent variable, I used another unobtrusive measure of narcissism (CEO picture) in the robustness tests.
5.2 Practical Implications
From my research I conclude that having a more narcissistic CEO does not lead to a higher relative exploration orientation of the firm. This means that the individual with the highest position of the firm does not necessarily influence how the firm’s innovation orientation is defined and presented to the public. This is thus influenced by other factors, which practitioners should take into account. Knowing this can help them take actions to ensure that the firm is presented in their preferred way.
Being aware of the personality of the CEO helps recognize the behaviors that are either beneficial or detrimental to firm performance. With this knowledge, practitioners are able to compensate for the effect CEO personality has on the firm, so that firm performance is optimal and firm strategy is as desired.
Furthermore, this research refrains from concluding whether a relative exploration orientation is better for firm performance than a relative exploitation orientation. This depends on the firm and the environment in which the firm operates. Therefore, this research does not offer any conclusions to practitioners whether having a CEO with higher levels of narcissism is better for firm performance or not.
5.3 Limitations and Future Research
In my research, where I tried to connect CEO narcissism to the firm’s relative exploration orientation, several limitations emerged. These limitations should be taken into account in future research.
A first limitation concerns the measure of the independent variable, CEO narcissism. This study focused on a promising unobtrusive indicator of narcissism, text analysis (Raskin & Shaw, 1988). As described above, due to time constraints we were unable to filter out only the part of the transcript where the CEO was speaking. Therefore, our measure of CEO narcissism might also have included the level of narcissism of other executive officers speaking during the earnings calls. Next to the measure I used, there are many other measures of CEO narcissism, like the NPI questionnaire (Raskin & Hall, 1979); the CEOs prominence in the firm’s press releases, measures of relative pay (Chatterjee & Hambrick, 2007); and the size of the CEOs signature (Ham et al., 2018). Future research could replicate the research by filtering only the section of the transcript where the CEO is speaking or by focusing on one of the alternative measures of CEO narcissism.
A second limitation concerns the measure of the dependent variable, the firm’s relative exploration orientation. I used the adapted measure of Uotila et al. (2009) for my text analysis of the annual reports. However, there is some criticism on the use of annual reports as a source of text analysis. Firms could for example exclude certain information about their strategy in order to stay ahead of competition (Crawford, 2003). Firms try to distinguish themselves from competition with the help of innovation, thus it could be a possibility that they withhold information about their innovation orientation. This makes the use of annual reports a less valid measure for a text analysis on the firm’s innovation orientation. Furthermore, next to a text analysis based on annual reports, there are also other measures for the firm’s relative exploration orientation. For example, patent count and citations (Ahuja & Lampert, 2001) could be valuable measures. Unfortunately, these measures were not possible in this research due to time constraints, but future research can use them to replicate the research.
high power and influence (Hambrick & Mason, 1984). Furthermore, it is difficult to measure the degree of narcissism of other executives. Future researchers can however analyze how the level of narcissism of other top executives impacts the firm’s relative exploration orientation. They can also analyze how the level of narcissism of other top executives interacts with the level of narcissism of the CEO.
Furthermore, only two robustness tests were used in this research. One of these tests recreated the results, but with a different measure for the independent variable, namely the prominence of the CEO picture in annual reports. Future research should perform additional robustness tests, such as using a different narcissism dictionary like for example, the dictionary of Heyden et al. (2015), which can be used to test the reliability of the dictionary of Uotila et al. (2009). Future research could also adjust the Uotila et al. (2009) narcissism dictionary to the guidelines of McKenny et al. (2018).
This research was further restricted by some sample limitations. First, our sample consisted of the S&P 100 firms, which includes the largest public firms on the US stock market. However, in these bigger firms, the relative exploration orientation is likely to be influenced by a larger set of factors, besides the personality of the CEO. It can for example be influenced by complicated organizational systems, which create more complex decision making (Visser & Faems, 2015). Future research could therefore replicate the results on a sample of small and medium sized enterprises (SMEs). Next, in my research I did not distinguish between the length of tenure of the CEOs. Therefore, my sample consisted of CEOs with a very long tenure, but also with a shorter tenure. However, CEOs with a longer tenure (for example, 5 years and more) could have a larger influence on the firm’s strategy than CEOs with a shorter tenure (for example, less than 5 years) (Kashmiri et al., 2017). Because, when CEOs have been in the position for some time, they might feel more confident expressing their opinions, influencing decisions and directing the firm in a certain direction. Therefore, future research could replicate the results of this research by restricting the sample with CEOs with a tenure of at least 5 years.
References
Adams, R. B., Almeida, H., & Ferreira, D. (2005). Powerful CEOs and their impact on corporate performance. Review of Financial Studies, 18, 1403-1432.
Ahuja, G., & Lampert, C. M. (2001). Entrepreneurship in the large corporation: A longitudinal study of how established firms create breakthrough inventions. Strategic Management Journal, 22, 521-543.
Aktas, N., Bodt, E., Bollaert, H., & Roll, R. (2016). CEO narcissism and the takeover process: From private initiation to deal completion. Journal of Financial and Quantitative Analysis, 51, 113-137.
Angrist, J. D., Pischke, J. (2009). Mostly harmless econometrics: An empiricist’s companion. Princeton: Princeton University Press.
Barker, V. L., & Mueller, G. C. (2002). CEO characteristics and firm R&D spending. Management
Science, 48, 782-801.
Baron, R. A., & Tang, J. (2011). The role of entrepreneurs in firm-level innovation: Joint effects of positive affect, creativity, and environmental dynamism. Journal of Business Venturing, 26, 49-60.
Berson, Y., Oreg, S., & Dvir, T. (2008). CEO values, organizational culture and firm outcomes.
Journal of Organizational Behavior, 29, 615-644.
Bertrand, M., & Schoar, A. (2003). Managing with style: The effect of managers on firm policies.
Quarterly Journal of Economics, 118(4), 1169-1208.
Burgelman, R. A. (1994). Fading memories: A process theory of strategic business exit in dynamic environments. Administrative Science Quarterly, 39(1), 24-56.
Buyl, T., Boone, C., & Wade, J. B. (2019). CEO narcissism, risk-taking, and resilience: An empirical analysis in US commercial banks. Journal of Management, 45(4), 1372-1400.
Campbell, W. K., Goodie A. S., & Foster, J. D. (2004). Narcissism, confidence, and risk attitude.
Journal of Behavioral Decision Making, 17, 297–311.
Campbell, W. K., & Miller, J. D. (2011). The handbook of narcissism and narcissistic personality
disorder: Theoretical approaches, empirical findings, and treatments. Hoboken, NJ: Wiley.
Cannella, A. A., Jr., Park, J. H., & Lee, H. U. (2008). Top management team functional background diversity and firm performance: Examining the roles of team member colocation and environmental uncertainty. Academy of Management Journal, 51(4), 768-784.
Chatterjee, A., & Hambrick, D. C. (2007). It’s all about me: Narcissistic Chief Executive Officers and their effects on company strategy and performance. Administrative Science Quarterly, 52, 351-386.
Chatterjee, A., & Hambrick, D. C. (2011). Executive personality capability cues, and risk taking: How narcissistic CEOs react to their successes and stumbles. Administrative Science Quarterly,
56(2), 202-237.
Chen, G., Crossland, C., & Huang, S. (2016). Female board representation and corporate acquisition intensity. Strategic Management Journal, 37, 303-313.
Child, J. (1972). Organization structure, environment, and performance: The role of strategic choice.
Sociology, 6, 1-22.
Child, J. (1974). Managerial and organizational factors associated with company performance.
Journal of Management Studies, 11, 13-27.
Christensen, C. (1997). The innovator’s dilemma: When new technologies cause great firms to fail. Boston: Harvard Business School Press.
Covin, J. G., & Slevin, D. P. (1989). Strategic management of small firms in hostile and benign environments. Strategic Management Journal, 10, 75-87.
Covin, J. G., & Wales, W. J. (2012). The measurement of entrepreneurial orientation.
Entrepreneurship: Theory & Practice, 36, 677-702.
Crawford, V. P. (2003). Lying for strategic advantage: Rational and boundedly rational misinterpretation of intentions. American Economic Review, 93(1), 133-149.
Cycyota, C. S., & Harrison, D. A. (2006). What (not) to expect when surveying executives: A meta- analysis of top manager response rates and techniques over time. Organizational Research
Methods, 9, 133-160.
Deluga, R. J. (1997). Relationship among American presidential charismatic leadership, narcissism, and rated performance. Leadership Quarterly, 8(1), 49-65.
Dess, G. G., & Beard, D. (1984). Dimensions of organizational task environments. Administrative
Science Quarterly, 29, 52-73.
Dougherty, D., & Hardy, C. (1996). Sustained product innovation in large, mature organizations: Overcoming innovation-to-organization problems. Academy of Management Journal, 39, 1120-1153.
Drucker, P. F. (1992). Managing for the future – the 1990s and beyond. New Delhi, India: Tata McGraw-Hill.
Duchon, D., & Drake, B. (2009). Organizational narcissism and virtuous behavior. Journal of
