Do Narcissists Turn a Blind Eye? A Study on the Willingness to Recognize Performance Cues for Reputable CEOs
Bas van Midde – 11963727
MSc Business Administration – Strategy Track
University of Amsterdam – Amsterdam Business School Thesis supervisor: Hesam Fasaei, MSc
Statement of originality
This document is written by Student Bas van Midde 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.
TABLE OF CONTENTS
Abstract 6
Introduction 7
Literature Review 10
Defining narcissism 10
Reputation of organizations and CEOs 11
Retrospective rationality 13 Performance cues 14 Narcissism as a moderator 15 Methodology 18 Independent variable 19 Dependent variable 20 Moderating variable 20 Control variables 21 Results 23 Normality check 23 Correlation analysis 24 Regression analysis 25 Moderation analysis 27 Discussion 28
Implications of results for theory and practice 28
Limitations 31
Recommendations for further research 32
ABSTRACT
This research examines the top 100 most reputable companies from 2011 to 2017 to examine the effect that reputation has on the willingness to recognize so-called performance cues, signals that firms’ CEOs need to change something in their organizations. Since prior research established ambivalent conclusions about the effect of narcissism on strategic risk-taking, we use narcissism as a moderating variable in an attempt to break through this indecisiveness. We hypothesize that having a high reputation affects the strategic risk-taking variable, the willingness to recognize performance cues, negatively. The main argument behind this proposition is that when CEOs are regarded highly, this leads to a neglection of identifying clues that change is needed. In addition, we propose that the presence of the personality trait of narcissism has a positive moderating effect, since the concept of retrospective rationality argues that narcissists have a tendency to convince themselves that they can do no wrong. Because of this, the willingness to see something that is right in front of them is not there. The statistical research conducted in this study proves both of our hypotheses to be wrong. Moreover, we find a positive, rather than negative, significant effect in the direct relationship of reputation and the willingness to recognize performance cues. Although our propositions do not hold up, many interesting effects are discovered that encourage future research to continue on the path that we have chosen. Moving away from the exhausted, ambivalent, direct relationship between narcissism and strategic-risk taking is the precedent that we set in this research and that we incite others to develop further.
INTRODUCTION
The story of the Greek figure Narcissus has been recited by many academic business-related researches already (Brown, 1997; Carr, 1994; Chatterjee & Hambrick, 2007; Michel & Bowling, 2013). Falling in love with his own reflection in the water, Narcissus does not need anyone else to love him; his self-admiration is completely self-fuelled. Unfortunately, the myth of Narcissus ends badly with him literally drowning in his own self-love. Where Narcissus was satisfied with himself and only himself loving his own person, modern narcissists demand more. Quite recently, it has been found that narcissism also includes a need for confirmation and admiration from other parties (Carr, 1994; Chatterjee & Hambrick, 2007; Wallace & Baumeister, 2002). Narcissists have thus become reliant on external appreciation. The most effective way to get this appreciation is to be in the public eye and build up a favourable reputation. What better way to get into the limelight than being the leader of one of the world’s biggest companies?
Different CEOs possess different magnitudes of narcissism, just like their personalities in general differ from each other. Still, some degree of narcissism, however small, is always there. (Chatterjee & Hambrick, 2007). In other research, it is shown that CEOs have higher needs of being appreciated, standing in the spotlight and building up high reputations (Brown, 1997; Raskin, Novacek, & Hogan, 1991). These personal goals can be classified as ‘self-fuelling needs’. Rather than a CEO fulfilling needs for the organization, a self-‘self-fuelling need is a desire that is beneficial more so for him- or herself. From this, we can derive that there can be a conflict of interest between CEOs’ self-fueling needs and effective firm performance (Chatterjee & Hambrick, 2007; Fralich, 2012). The ability, but also the willingness of CEOs to balance these two demands further research. Moreover, it needs to be researched what such a
self-fuelling need like reputation has on firm performance, specifically in times when heroism is needed; when firm performance is highly dependent on strategic risk-taking.
Throughout previous research, narcissism has mostly been used as an independent variable having either a positive or negative effect on strategic risk-taking. Conflicting points of view have been taken and conflicting conclusions have been made (e.g. Chatterjee & Hambrick, 2007, 2011; Fralich, 2012). This leads us to believe that building such a straightforward relationship may be too direct of an approach. The effect that reputation has on taking risks is something that has not been researched as much. This is where the gap in the existing literature is. In this research, we use narcissism as a moderating variable between the dependent and independent variable. In other words, we hypothesize that narcissism is the driving factor that decides the nature of the relationship between reputation and strategic risk-taking. We believe that attributing a moderating role to narcissism contributes to a further understanding of the overall psychological phenomena that take place.
Strategic risk-taking can only be practiced when there is a timely recognition of the need to take risks. In this research, the term ‘performance cues’ is used to encompass this. Performance cues are signals that change is necessary within the organization. The contribution that this new term brings is that performance cues are indicators that CEOs’ performance is not good enough. Previous research has only gone as far as researching CEOs’ ability cues; indicators that CEOs simply do not have enough ability and that the company is failing for that reason (Chatterjee & Hambrick, 2011). When it comes to performance cues however, company failure is attributed to CEOs’ lacking performance, like the term suggests. Thus, the usage of the term ‘performance cues’ allows for a more heads-on approach to explaining the problem.
Do CEOs want to see that they are lacking performance though? After all, reputable CEOs have earned their reputation for a reason. They cannot fail, unless there is a heavy influence from the external environment, which would not be their fault. These are the kinds of thought processes that prestigious CEOs go through. Unlike in the aforementioned concept of ability cues, untimely recognition of performance cues is not an indicator that ability is lacking (Chatterjee & Hambrick, 2007). Rather, there is another factor that is holding CEOs back from making appropriate changes, namely the willingness to change. This factor of willingness plays a substantial role in this research. For performance cues specifically, it is proposed as the driving force behind the failure of recognition. Consequently, we argue that performance cues are more dependent on the willingness to recognize, rather than the ability to recognize.
In order to fill the existing research gap and realize the proposed contributions that were mentioned before, our research question is: What is the impact of reputation on the willingness to recognize performance cues for CEOs?
LITERATURE REVIEW
There are conflicting views as to what the effect of being a highly valued CEO has on driving change in firms. On the one hand there has been previous research that argues that the overall reputation of CEOs has a positive effect on strategic dynamism and risk taking (Chatterjee & Hambrick, 2007; Fralich, 2012; Zhang, Liu, & Jiraporn, 2011). This makes sense from a logical point of view since having a good reputation makes CEOs more likely to adhere to these reputations and undertake dynamic, risky decisions that, when successful, confirm their expertise as managers. However, there is also research that has found that CEOs with high reputations have a harder time seeing that they need to take risky decisions (Brown & Sarma, 2007; Chatterjee & Hambrick, 2011; Menkhoff, Schmidt, & Brozynski, 2006). The general reasoning here is that the praise that CEOs get builds up overconfidence and triggers the neglecting of signals that change is necessary. The phenomenon of narcissism helps bridge the gap between the two conflicting views on how reputation and celebrity have an effect on strategic risk-taking. The interdependencies of reputation, performance cues, and the aforementioned notion of narcissism will be examined more closely using the existing literature.
Defining narcissism
Defining narcissism is a complicated matter. Throughout the years, the definition has been changed and expanded multiple times moving away from the initial meaning of the word as was established by the Greek myth of Narcissus. At this moment in time, the literature has agreed on some major characteristics of narcissism (Chatterjee & Hambrick, 2007; Emmons, 1987; Gerstner, König, Enders, & Hambrick, 2013). First of all, narcissists have a sense of entitlement, feeling that they deserve respect and admiration at all times and have a high need
for applause and appreciation (Morf & Rhodewalt, 2001). In addition, narcissism often coincides with a sense of superiority. To maintain this feeling of superiority, narcissists have an urging need to outperform others (Campbell, Goodie, & Foster, 2004; Judge, LePine, & Rich, 2006). What’s more, peer domination and leadership are also narcissistic traits that allow for increased self-esteem (Raskin et al., 1991). Finally, another common denominator for narcissism is self-admiration, a characteristic that can be traced back to the origin of the concept of narcissism (Emmons, 1987). Rather than forming one perfect, all-covering phrase to define narcissism, these four points of agreement about what narcissism is will act as a definition. Knowing this, we can examine the existing literary relationships with the other main variables of this research.
Reputation of organizations and CEOs
The narcissistic factor of entitlement bring us closer to the concept of reputation. Reputation can be a very valuable asset for organizations, but is also an intangible that is highly dependent on outside (media) coverage and quite circumstantial (Westphal & Deephouse, 2011; Zavyalova, Pfarrer, Reger, & Shapiro, 2012). On the other hand, using forms of impression management, the amount of coverage and the tone of the coverage can be influenced by organizations as well (Westphal, Park, McDonald, & Hayward, 2012). This means that reputation is thus not entirely coincidental. One research notes that: ‘Organizations are expected to be acutely sensitive to what outsiders think of them and to make the sort of strategic decisions and information disclosures that create a favourable impression’ (Brown, 1997, p. 668). Organizations are thus inherently narcissistic because of the need for a favourable reputation. This even shines through in the corporate decisions that are made within the organization.
Reputation is not only a valuable asset for organizations, but also for CEOs. First of all, the monetary rewards of CEOs are dependent on firms’ reputations. Bad media coverage about unrighteous bonusses for top-level managers have shown to have a negative effect on CEO pay, and an aggravated negative effect on CEOs in organizations that value reputation more (Kuhnen & Niessen, 2012). A popular reputation of CEOs establishes a bargaining position that is hard to deny for organizations. Outside-the-organization celebrities and influential spokespersons increase brand recognition and help build up a solid brand image (Driessens, 2013). Although this is true, having a well-recognized CEO helps achieve this without having to pay the highest of endorsements and having to cope with the wishes of famous sports stars or actors and actresses. Therefore popularity of a firm’s CEO is a great resource to possess. This appreciation is expressed through monetary rewards, and is the reason why recognizable CEOs receive higher compensation than unknown CEOs (Wade, Porac, Pollock, & Graffin, 2006).
The monetary aspect is not where the root of the need for favourable reputation lies though. As mentioned before, firms exert self-centred behaviour through the strategic decisions that they make. The reigns to make these strategic decisions are usually handed to the CEO of the organization (with a monitoring role for firms’ directive boards). This makes the firms’ operators, the CEOs, quite susceptible to being narcissistic themselves. More often than not, narcissism has a substantial effect on the structure and decision-making of organizations (Brown, 1997; Chatterjee & Hambrick, 2011; Miller & Droge, 1986). The exact effect of a narcissistic personality on firm performance is rather unpredictable and a narcissist’s behaviour is not very rational in itself. What has been discovered, though, is that narcissism in CEOs is positively related to strategic dynamism and grandiosity (Chatterjee & Hambrick, 2007). In addition, it was found that narcissism has a positive effect on the engagement in technological
discontinuities (Gerstner et al., 2013). The relationships between narcissism and these ego-boosting changes can for some part be explained by the fluctuating, unpredictable behaviours of narcissistic CEOs, but it can additionally be justified by arguing that narcissistic CEOs are more confident in making changes.
Retrospective rationality
These unpredictable behaviours are also where the danger lies with narcissism. It is argued in prior literature that there are several types of irrational behaviour that impair narcissists’ sight of their own performance (Brown, 1997; Chatterjee & Hambrick, 2011; Jordan & Audia, 2012; Ross & Staw, 1993).One of the most relevant irrational behaviours to explain this hazard of unpredictability is the concept of ‘retrospective rationality’ in business. Retrospective rationality is based on the phenomenon where people spend more time thinking about the consequences of their actions than about the impact these consequences have on future behaviours (London, 1983). In other words, retrospective rationality arises when a decision-maker in a narcissistic state is concerned with maintaining consistency in the outcome that the assessment should lead to instead of establishing a constant evaluation process and envisioning what this could mean for similar situations in the future (Jordan & Audia, 2012; Staw, 1981).
How this relates to this study is that narcissistic CEOs feel that the outcome of their assessment should always be on par with or above their aspiration level. A recurring phenomenon is that CEOs rationalize incorrect decisions in order to maintain or enhance their reputations (Golden, 1992; Merkl-Davies & Brennan, 2011). Sometimes CEOs may go as far as attempting to hide their faulty decisions to preserve their images (Cannell & Henson, 1974). However, large corporations such as the ones studied in this research are heavily controlled
and audited so that these decisions often come to light anyway. It is therefore that CEOs’ main instrument for defending their reputations is reporting events and occurrences in their firms as reasonable, thought-out and goal-oriented behaviours and actions (Aerts, 2005; Merkl-Davies & Brennan, 2011). This distorted image of CEOs’ abilities causes CEOs to neglect forming an objective opinion of their own performance but rather they put too much focus on finding ways to prove to themselves and others that their results were in fact good enough.
Performance cues
The subsequent logical step is then to apply the concept of retrospective rationality to dynamism in the organization. Changes happening in an organization are often preceded by certain hints or clues that something is not working as effectively or not working anymore entirely. Chatterjee & Hambrick (2011) introduce the concept of capability cues, which they define as ‘contextual signs that decision makers might reasonably interpret as indicators of their current level of overall ability’. The theorization here is that prior research find such cues in external praise and recognition, also adhering to the more modern form of how narcissism expresses itself nowadays. Executives are, however, likely to vary in their reactions to these capability cues as a result of their personalities (R. Brown & Sarma, 2007; Chatterjee & Hambrick, 2011; Menkhoff et al., 2006; Miller & Toulouse, 1986). This will be discussed in more detail in a later section of this literature review, inserting the notion of narcissism into this phenomenon.
In our research this term of capability cues is refrained from. The concept of retrospective rationality tells us that narcissistic, irrational decision-makers overestimate the consistency of their own abilities even when it is clear that they did not perform adequately. Stubborn as they are, narcissistic CEOs do not believe that they or their organizations can fail
and subsequently they do not see cues that strategic risks need to be taken in order to change. Contextual signs as indicators for change are therefore not necessarily related to (cap)ability, but could very well be a result of ignorance of clear cues by reputable CEOs. Misevaluating indicators that change is needed could thus be related more to the willingness to recognize contextual signs. As a consequence, CEOs do not take strategic risks when this is necessary (Sitkin & Weingart, 1995; Van de Ven & Polley, 1992). For that reason the term performance cues is used. Rather than cues that establish that the ability of CEOs is changing, we only say that the performance is changing. We thereby eliminate the assumption that performance changes because of ability. We propose that recognizing performance cues is especially dependent on CEOs’ reputations such that more reputable CEOs are more stubborn and negligent about recognizing performance cues.
Hypothesis 1: CEOs’ reputation has a negative effect on their willingness to recognize
performance cues.
Narcissism as a moderator
Most research trying to establish a relationship involving narcissism and strategic risk-taking approach this by using narcissism as an independent variable affecting the dependent variable that is (some form of) strategic risk-taking (e.g. Gerstner, König, Enders, & Hambrick, 2013; Rijsenbilt & Commandeur, 2013; Wales, Patel, & Lumpkin, 2013; Wallace & Baumeister, 2002). This research establishes part of its novelty by hypothesizing narcissism as a moderator. In an attempt to disentangle the aforementioned conflicting effects of narcissism on risk-taking that were found in prior research we propose a different independent variable. Simultaneously, we introduce the existence or non-existence of narcissism in CEOs as the determinant of what that relationship looks like. Since reputation is something that was found
by other researchers as a strong factor affecting risk-taking propensity, this is the independent variable that we use (e.g. Fralich, 2012; Staw & Epstein, 2000; Wade et al., 2006; Zhang et al., 2011).
The best CEOs are able to recognize signals that the firm needs to change as soon as possible and are subsequently fastest to find a proper response to these signals. Executives’ personalities can have quite an impact on the skills required for sound strategic risk-taking though. One research suggests that locus of control is a strong component of a company’s strategic doings (Miller & Toulouse, 1986). Managers with an internal locus of control attribute consequences to their own actions, while external managers are less self-reflecting (Rotter, 1966). One can easily identify that narcissistic managers would fit into the latter group. Other research finds that experience plays a substantial role in strategic-risk taking. It is argued that less experienced managers tend to take riskier decisions (Menkhoff et al., 2006). Finally, research has been done on CEO overconfidence and dominance, two major characteristics of narcissists, and their effect on corporate acquisitions. Not surprisingly, a positive effect of overconfidence and dominance was found on the propensity to acquire other firms (Brown & Sarma, 2007).
All in all, when we dissect how narcissistic personality traits can have an effect on strategic risk-taking it is clear how narcissism can be hypothesized as a deciding variable for the relationship between two closely related variables, reputation and the willingness to recognize performance cues. The purpose of this research being to attempt to break through the conflicting views of the effect of narcissism on strategic risk-taking, the utilization of reputation as an independent variable acting as a breath of fresh air is also explained. We can
therefore justify using narcissism as a moderating variable looking at it from both sides of the direct relationship that is proposed.
Hypothesis 2: The relationship between reputation and willingness to recognize performance
METHODOLOGY
This paper aims to research the effect that reputation of CEOs has on the willingness to recognize so-called performance cues. Performance cues are signals that change is needed in an organization. The characteristic of narcissism in CEOs plays a moderating role in this relationship. There is a hypothetical moderating effect of narcissism that strengthens the hypothetical negative effect of reputation on the willingness to recognize a need for change (Chatterjee & Hambrick, 2011; Fralich, 2012; Jordan & Audia, 2012). The conceptual model helps visualize this proposed relationship (Fig. 1).
Figure 1 Conceptual model
The empirical setting of this research is worldwide but concentrated on larger corporations. There is much more of a spotlight on multinationals than on regular small and medium-sized enterprises (SMEs), so CEOs of these types of organizations are much more susceptible to being sensitive to their reputation levels. As an effect, these CEOs are, hypothetically, more likely to exert narcissistic behaviour and are thus the focus of this research. Timewise, we wanted this research to be as recent as possible, especially regarding the volatility of reputation today. Therefore this research does not look at historical, outdated data and statistics but focusses on recent numbers.
Willingness to recognize
performance cues
Narcissism
Reputation
Independent variable
Taking into regard the empirical setting that was established, the data collection started with finding reputation statistics of these larger corporations. In previous research CEO certifications were assessed through a CEO of the Year competition by Financial World (Wade et al., 2006). Unfortunately, this contest was discontinued in 1996, but it did give us an idea what kind of data to look for to find this reputation statistic. In the end, the Reputation Institute has provided us with each year’s top 100 most reputable companies in the world from 2011-2017. This gives us an initial sample of 700 reputation scores. The used measurement for reputation is the ‘pulse’ of a brand, which is explained as the emotional connection consumers have to a brand (Reputation Institute, 2018). Companies are then scored on seven different dimensions. Although the Reputation Institute does not want to give away their entire methodology, their research has been validated by other publications (Fombrun, Ponzi, & Newburry, 2015), making it a representative way to quantify reputation.
Both the rank of reputation in each year as well as the actual score that was given on the basis of the brand’s pulse are entered as separate variables into the initial dataset. This is done so that both the rank and the score can be assessed on their significance related to the other variables. Although CEO reputation and firm reputation are not exactly the same, it can be argued that CEOs’ behaviours and appearance make up a large share of what the outside public thinks of a corporation. One paper found that there is a significant positive relationship between CEOs that use popular management techniques and firm reputation (Staw & Epstein, 2000). Firm leaders that make use of practices that the outside public approves of thus improve the reputation of their firms. This proves that the reputation of companies can be used as a valid proxy for the reputation of CEOs.
Dependent variable
As for the dependent variable, the willingness to recognize performance cues, the primary data that is being used is research and development expenses, extracted from the Orbis database. This data gives us an indication of how much risk is being taken by a firm and has been used in previous research as a proxy for strategic risk-taking as well. Several reasons for using this variable as a proxy have been given, the first being that it is a sound indicator for strategic resource allocation and management (Fralich, 2012; Geletkanycz & Hambrick, 1997). What has also been used as justification is that research and development expenses are to some extent controllable by CEOs (Finkelstein & Hambrick, 1990).
To make this statistic more representative, it may be useful to calculate the research and development expenses relative to total expenses. This would also make research and development expenses more comparable across firms and industries, which is one of the benefits that this indicator can have (Finkelstein & Hambrick, 1990). Whether there is more significant than the absolute form of research and development expenses will come forward in the statistical analysis.
Moderating variable
A suitable proxy for narcissism has been the most challenging to find data for. Previous research has recognized this difficulty as well, the reason being that it is difficult to conduct interviews with CEO’s of highly regarded companies about their personalities. However, previous research did find a manner to get proxies for CEO narcissism using unobtrusive indicators. The two main proxies that were used in previous research are CEO compensation and the usage of CEO photographs in companies’ annual reports (Chatterjee & Hambrick,
2007; Gerstner et al., 2013; Rijsenbilt & Commandeur, 2013). These are therefore the indicators that will be used in this research as well.
CEO compensation for US-based firms was retrieved from COMPUSTAT while non-US firms’ information on CEO pay was manually retrieved from annual reports, 10-K filings, compensation reports, sustainability reports and other firm document releases. This was a time-consuming and difficult process that also led to the inability to find data on CEO pay for some firms. This increased the number of missing values in our dataset and thus reduced our sample size. The prevalence of a CEO’s photograph in the annual report was also manually researched, and given a score from 1 to 4, similar to how Chatterjee & Hambrick (2007) assessed this. 1 = no photograph on annual report, 2 = photograph with one or more other executives, 3 = photograph of CEO encompassing less than half a page, 4 = photograph of CEO encompassing half a page or more. Both of these unobtrusive indicators will be handled separately as moderating variables in the data analysis.
Control variables
It is quite important for this research to be able to isolate the effect of reputation on recognizing performance cues. In order to achieve this, control variables are used. We used two CEO controls, CEO age and CEO tenure. The age of CEOs and the duration of their functioning as CEO were researched manually, utilizing many different news articles and general resources to find out who was CEO when and for how long. In addition, we use a time control, the calendar year of when each entry earned a place on the list of top 100 most reputable companies. This gives us three total control variables that were also used in many previous studies on the same subject matter (Chatterjee & Hambrick, 2007, 2011; Fralich, 2012; Gerstner et al., 2013).
After the examination of missing values we finalize our dataset at a sample of N=285 entries. Although this is quite a cut from the 100 most reputable firms across seven years, this was to be expected following the extensive manual data collection that had to be done. In the next section our hypotheses will be tested using this sample.
RESULTS
Normality check
Starting off with the statistical analysis, the histogram plots showed the existence of kurtosis and skewness for some variables. Especially for CEO compensation and CEO compensation relative to operating revenue, the numbers were astronomically high, and violated the standard of numbers between -2 and 2 set by statistical research (Curran, West, & Finch, 1996). After assessing the outliers we saw that there were multiple Z-scores that were over 3, implying that these needed to be disregarded in order to get a normal distribution (Pukelsheim, 1994). Using the proper SPSS commands we omitted these observations. In total there were 17 different variables that crossed the Z-score of 3, so we move from an N of 285 to 268.
After doing this we still had high kurtosis and skewness numbers for three variables, even though they were generally much lower now that the outliers were disregarded. There was still high kurtosis and skewness for the moderating variables CEO compensation and CEO compensation relative to operating revenue. Also, disregarding some of the variables increased the skewness and kurtosis statistic for the control variable CEO tenure. The common solution for this is to move to log-based variables (Keene, 1995). This was necessary for these three variables, since they all showed kurtosis and skewness over 2. After changing the variables to log-variables this was completely fixed. The skewness and kurtosis variables for all three variables now fit the threshold of being between -2 and 2 and even the more conservative criterium of being between -1 and 1 (Groeneveld & Meeden, 1984).
Correlation analysis
Next, a correlation analysis of all variables was conducted in order to check for the significance of the variables in relation to each other. In addition the means and standard deviations of all variables were computed (Fig. 2).
Figure 2 Correlation matrix
Mean Std. Deviation 1 2 3 4 5 6 7 8 9 1. R&D Expenses 3 341 159 3 198 516 2. R&D Expenses / Operating Revenue 0.0594 0.0497 .644** 3. Prominence CEO Photograph 2.55 1.127 0.046 -0.008 4. Reputation Rank 46 29 -.291** 0.046 -0.115 5. Reputation Score 71.71 3.60 .305** -0.028 .137* -.939** 6. Year 2014 1.8 0.0 -0.022 -0.071 0.054 -0.005 7. CEO Age 56.69 5.22 -.129* -.351** 0.107 -0.099 0.109 -0.009 8.LogCEO Tenure 0.6690 0.3378 -0.0489 -0.049 -.124* 0.066 -0.072 -0.003 .298** 9. Log CEO Comp 3.8783 0.3614 .203** .235** -0.084 .218** -.179** .185** -0.016 0.089 10. Log Relative CEO Compensation -3.7247 0.5174 -.378** .130* -0.106 .464** -.437** .157* -.136* .123* .624**
We find many significant effects from this table, of which some will be highlighted here. First of all we find that there is a negative significant effect of reputation rank on R&D expenses while there is a positive effect of reputation score on R&D expenses. These might seem contradictory at first, but it must be kept in mind that a high reputation rank means that a firm is further down the list in terms of reputation score. For example, a rank of 75 is thus a ‘higher’ rank than rank 10. Furthermore, we find that the control variable CEO age has significant negative effects on both R&D expenditures as well as R&D expenditures relative to operating revenue. Finally, it can be pointed out that there are significant negative effects of reputation score on (relative) CEO compensation.
A note that should be taken into consideration with this correlation is that the means and standard deviations for variables 8-10 lack meaning now that they have been computed to log-based indicators. What’s more, one might look at the mean of reputation rank and see that this is not 50 which would be the expectation. However, missing variables and the aforementioned data cleaning have caused this number to move away from the original mean. This is not in any way problematic for the analysis.
Regression analysis
Up until the correlation analysis we had used two indicators for the independent variable of reputation: reputation rank and reputation score. For the regression analysis however, we have to choose one. Since reputation score has proven to provide more significant relationships, as is evident from Figure 2, we will be using that variable in this section. Hierarchical multiple regression was used to identify the effect of reputation score on R&D expenses, which acts as a proxy for recognizing performance cues. During the first step only the control variables year, age and tenure were entered. For the second step the independent
variable, reputation score was added so the difference in R2 could be observed (Fig. 3).
Figure 3 Multiple Regression Analysis
From the multiple regression analysis we find that the move from step 1 to step 2 causes an increase of 10.3% in the model’s ability to explain the variance in the dependent variable. In other words, adding the variable of reputation score to the control variables causes a significant increase at the p < 0.001 level so that the model better explains what happens to R&D expenses. In the second model we see that the Beta value of age as a predictor increases from being significant at p < 0.05 to p < 0.01. In addition, reputation score is significant at p <
R R2 Change in R2 B Standard error Beta t Step 1 .131 .017
Log CEO Tenure -106 093.5 606 425.4 -0.011 -0.175
Year 22 345.5 106 512.0 0.013 0.210
Age at year’s end -77 264.8 39 273.7 -0.126* -1.967
Step 2 .347 .120 .103***
Log CEO Tenure 246 578.0 578 231.3 0.026 0.426
Year 24 627.1 100 945.9 0.014 0.244
Age at year’s end -105 797.8 37 574.3 -0.173** -2.816
Reputation score 289 535.6 52 163.7 0.325*** 5.551
*. Correlation is significant at the 0.05 level (2-tailed); **. Correlation is significant at the 0.01
0.001 as a predictor for R&D expenses. The beta is positive, signalling that there is a positive effect of reputation score on R&D expenses, contradicting hypothesis 1.
Moderation analysis
Using Andrew Hayes’ PROCESS program in SPSS, we analyse if there is indeed a moderating effect of the narcissism proxies on the relationship between the dependent and independent variable. Because our input variables, R&D expenses, reputation score, prominence of CEO photograph and the log of relative CEO compensation are already numerical variables, there is no need to standardize the variables. Running the commands twice, for the moderating proxy of CEO photograph’s prominence as well as the log of relative CEO compensation, we find that neither has a significant moderating effect. We can therefore reject hypothesis 2 of the research.
DISCUSSION
In this research we tried to prove a negative relationship between reputation and the willingness to recognize performance cues. We proposed that this relationship was enforced in a moderating manner by the existence of the personality trait of narcissism. The purpose was to move away from the rigidities that the contrasting findings of the direct relationship between narcissism and strategic risk-taking had caused while still holding close to the undeniable effect that several narcissistic personality traits have on being able to react to change.
From the statistical analyses that were done we can say that both of our hypotheses have been proven untrue. Rather than finding a negative direct relationship between reputation and the willingness to recognize performance cues, we found a positive relationship. This was directly against our expectations, making it all the more interesting of a result to analyse in this section. Furthermore, there was no moderating effect found of narcissism on the relationship between reputation and the willingness to recognize performance cues. Again, this does not mean that there are not meaningful takeaways that result because of this finding.
Implications of results for theory and practice
What we can easily say about the statistical research is that there are many significant effects arising from the correlation matrix, signalling that there are multiple relationships taking place among the variables. Most important for our research is the direct positive relationship that reputation score has on R&D expenses, disproving our primary hypothesis. It turns out that CEOs do not turn a blind eye, like our research’s title suggests. It does seem worthwhile to also examine the insignificance between the relative form of R&D expenses and reputation score and rank. This means that when we correct for a firm’s operating revenue there is not a
contradiction of our primary hypothesis. For theoretical purposes, we conclude that the effect of reputation on the willingness to recognize performance cues is rather inconclusive and deserves more attention. Practically, firms should still closely monitor the effect of reputation on their risk-taking policies set out by CEOs even though there is the suggestion here that there is a positive relationship between the two.
Although the usage of it was merely to control and thus isolate other effects, the significant effect that CEO age has on research and development expenses is another interesting finding. It was mentioned in the literature reviews that there was certain research that proved that less experienced managers were more inclined to take strategic risks (Menkhoff et al., 2006). This is what we also find in this correlation matrix. For theory in this field of research, this is something that deserves more research. This study as well as prior literature has used CEO age as a control variable (e.g. Chatterjee & Hambrick, 2007, 2011; Fralich, 2012). However, our results suggest that CEO age may be able to fulfil a bigger role in the statistical research in this field and may be able to shine more light on the interplays between reputation, narcissism and the willingness to recognize performance cues. In firms, evaluating this variable more closely would help organizations pinpoint the different attitudes that younger and older managers have on taking strategic risks.
We also find significant effects for the log of absolute CEO compensation both on absolute research and development expenses as well as on relative research and development expenses. Additionally, this positive significant effect can be found for the log of relative CEO compensation on relative research and development expenses. In short, we see the much-researched effect of narcissism on strategic risk-taking taking place. Even though we wanted to establish a different approach to this field of business by using narcissism as a moderator,
we can still see the direct effects between narcissism and strategic risk-taking appearing. We can suggest from this research that there is a positive relationship between the two that is being brought forward in this research. However, there is still some ambiguity to this relationship that was discovered in this paper, since relative CEO compensation has a negative significant correlation coefficient in relation with absolute research and development expenses. The theoretical ambivalent effects of narcissism on strategic risk-taking are thus also taking place in this paper.
Finally, there are more contradicting direct effects of narcissism on reputation. On the one hand, we found a positive effect between our independent variable of reputation and the prominence of a CEO’s photograph, which is a proxy for narcissism. This effect would call for the suggestion for firms to monitor the usage of CEO portraits in their annual reports more closely to optimize reputation. However, the results also suggest that there is a negative relationship taking place with regards to CEO compensation and reputation. A higher log of absolute CEO compensation and relative CEO compensation, respectively, leads to a lower reputation score. As it has been recommended in prior research, the amount of money that a CEO is rewarded with has a significant effect on the way that a firm is viewed by the outside public (Kuhnen & Niessen, 2012; Mccall, 2005). Although there is already much emphasis being put on keeping bonusses and compensation acceptable to the outside public, the notion that the outside public is critically watching needs to be stipulated for real-life practices once more.
The many differing, contradictory effects that seem to be taking place in this research are typical for this field of research. A conclusive set of proof and argumentation with regards to narcissism, reputation and strategic risk-taking is yet to be formed, which leaves it as
something that needs to be researched in the future. For practice this means that close monitoring is recommended as long as there is little conclusive advice that can be given on the matter of how to handle narcissistic CEOs most optimally. Careful examination in practice as well as theory may lead to additional insights being discovered and may eventually result in a final stance being taken on the matter that will explain the interplays between the variables altogether.
Limitations
We have attempted to make this paper as representative for the real world as possible, even though this is a field of research that is quite difficult to measure and quantify. Factors like our main variables, reputation, narcissism and the willingness to see something are in their essence rather intangible concepts. The usage of proxies was a way to get around this obstacle for researching these fascinating phenomena. The most extreme example of this was using unobtrusive indicators in order to establish a measurement for narcissism, for which we also have prior research to thank (Chatterjee & Hambrick, 2007; Gerstner et al., 2013; Rijsenbilt & Commandeur, 2013). The fact remains that there are some limitations and boundaries that result because of this. In this section these will be outlined thoroughly.
Let us first resume with what we mentioned last, the unobtrusive indicators for narcissism. In this research only the main indicators that were used in the aforementioned prior research were researched. There were other indicators used in prior research (e.g. Chatterjee & Hambrick, 2007) that would have been useful to include in the dataset as well, however the limited timeframe that was available led us to the decision to keep to two unobtrusive indicators. In hindsight, this was a wise decision, since the manual gathering of the data that we did was already so time-consuming that even if we were able to complete the data gathering
process for the other unobtrusive indicators this would have been at the expense of the overall quality of this thesis. An example of another factor that would have been interesting to include as a proxy for narcissism would have been the language in press releases and other media publications. One can however imagine how costly this would have been from a time-based perspective.
Something quite similar can be mentioned for the proxies used for strategic risk-taking. We chose research and development costs as the primary measure for this, in part because it can be argued that it is a cost that CEOs have the most freedom in allocating (Finkelstein & Hambrick, 1990). However to get to a more representative figure for strategic risk-taking, including other costs into this equation may have helped this cause. In prior research, examples of other expenditures that were used are advertising costs, financial leverage data and capital intensity numbers. These were all collected in order to create a composite score for research and development intensity rather than just research and development costs (e.g. Fralich, 2012; Geletkanycz & Hambrick, 1997; Zhang et al., 2011). Like with the proxies used for narcissism, this would have been too time consuming of a process. We did collect the operating revenue statistics for the firms in the dataset so that a difference could be made between absolute and relative research and development costs, and figured that we would grab the essence of being able to express strategic risk-taking in numerical terms in that way. Adding factors such as advertising intensity and capital intensity would have made that expression more complete though, and is thus a limitation of this research.
Recommendations for future research
Although many direct effects that narcissism has on the dependent and independent variables were outlined, there was no moderating effect found. Maybe partly because of the
unexpected positive effect that reputation has on the willingness to recognize performance cues, the chance to find a moderating effect of narcissism was absent. Although this is true, there may be independent variables other than reputation that do have narcissism as a moderator and can help this area of research further. It might be the case that the direct hypothesis of the relationship of another independent variable on the willingness to recognize performance cues holds true and the moderating role that narcissism may subsequently play would then be more likely to be significant as well. This is therefore our first recommendation for future research. This paper was merely a first step in trying to look at narcissism from a different angle and untangle the concept from the ambiguous effects that other researchers had found of narcissism on strategic risk-taking. We therefore encourage future research to find many different independent variables and test the presence of a moderating effect of narcissism that can help explain when something is either positive or negative for the willingness to recognize performance cues in a firm.
In addition, something that could explain in more detail the underlying mechanisms that are involved in this field of research is to make a difference between the type of performance cue that is to be recognized. It is not per se a ground-breaking, company-wide discontinuity that needs to be executed. It may be incremental change that might be more effective and is therefore needed to be recognized by CEOs as a required strategy. Whereas prior research established the positive effect of narcissism in CEOs on the adoption of technological discontinuities (Gerstner et al., 2013), the implications of narcissism for incremental, slower change are equally important, yet mostly left unexplored. Truthfully, it might be harder to acquire data on this type of change but it does provide an opportunity for future research. This paper only examined the overall variable of strategic decision-making. One is left to wonder what the differences in the effect of reputation on the usage of radical
change vis-à-vis incremental change. The significance of this finds its root in the fact that radical change is often publicized and brought out to the outside world, likely garnering outside media coverage. Incremental change is much more under the radar though. It can then subsequently be suggested that firms that have high reputations could use radical change as an instrument to maintain this high reputation, while slower, smaller, incremental change would not help this purpose. Not to mention the far-reaching effects that the concept of narcissism could have on this difference between radical and incremental change, future research should look into the possibility to make a difference between the two
Finally, it might be interesting to look at how narcissism grows throughout the years for CEOs. While we used CEO age as a control variable in this research and not as an explaining variable, the data did provide us with significant statistical outcomes. In prior research, the effects of CEO age over time has been argued as a recommendation for future research as well (Chatterjee & Hambrick, 2007). Including the effects that were found in this research only adds to the suggestion that there may be something happening over time with the variables of reputation, narcissism and the willingness to recognize performance cues. Firms and CEOs that acquire steady, prolonged periods of reputability and express narcissistic behaviours throughout this period may approach strategic decision-making in a different manner or may affect it with a different magnitude. We advocated throughout this research for using a different path in order to try to break up the conflicting views about the effect of narcissism on being able to see that a firm needs to change. We did this by using narcissism as a moderator rather than an independent variable in an attempt to achieve this purpose. Conducting a longitudinal study on narcissism and strategic decision-making may be another approach to think differently about narcissism and its effects. It is therefore that we recommend exploring this option in future research.
CONCLUSION
The purpose of this paper was to disentangle the concept of narcissism from the ambivalent ideas about the effects that this personality trait has on strategic risk-taking. Since previous research had yet to reach a definitive answer as to the nature of the relationship between narcissism and strategic-risk taking, we decided on a very different approach, namely using narcissism as a moderator. This left an empty space for the independent variable in our proposed framework, which we filled using the variable of reputation. The intention was to find a direct relationship between reputation and strategic risk-taking. Secondly, we hypothesized that this relationship was then moderated by narcissism.
Another novelty that we introduced was the concept of ‘performance cues’. We argued that it was often CEOs not recognizing mediocre performance that was the problematic part of making sound strategic decisions. Hypothetically, this issue gets worse for reputable or narcissistic CEOs. We argue that it is not about the ability to recognize these performance cues, rather the willingness for CEOs to identify that they are doing something wrong. This is why we isolate the dependent variable as the willingness to recognize performance cues.
Both the hypotheses of the direct relationship between reputation and the willingness to recognize performance cues as well as the moderated relationship came out insignificantly. Moreover, the direct relationship between reputation and the willingness to recognize performance cues came out positively rather than negative, which was our expectation. Although this is the case, this research still lays a precedent for future research. The novelties that we established in this research are important for the field of research as we show that there is much opportunity in approaching the concept of narcissism in business in a different way.
In addition, the results show that there are many other interplays between variables for which an opening has been found. In short, there is much to be discovered about narcissism and its role in business and approaching it from many different perspectives will only add to the understanding of its impact.
REFERENCE LIST
Aerts, W. (2005). Picking up the pieces: impression management in the retrospective
attributional framing of accounting outcomes. Accounting, Organizations and Society,
30(6), 493–517. https://doi.org/10.1016/j.aos.2004.07.001
Brown, A. D. (1997). Narcissism, Identity, and Legitimacy. The Academy of Management
Review, 22(3), 643. https://doi.org/10.2307/259409
Brown, R., & Sarma, N. (2007). CEO overconfidence, CEO dominance and corporate acquisitions. Journal of Economics and Business, 59(5), 358–379.
https://doi.org/10.1016/j.jeconbus.2007.04.002
Campbell, W. K., Goodie, A. S., & Foster, J. D. (2004). Narcissism, confidence, and risk attitude. Journal of Behavioral Decision Making, 17(4), 297–311.
https://doi.org/10.1002/bdm.475
Cannell, C., & Henson, R. (1974). Incentives, motives, and response bias. Annals of Economic and Social Measurement, 3(2), 307-317.
Carr, A. (1994). For Self or Others?: The Quest for Narcissism and the Ego-Ideal in Work Organizations. Administrative Theory & Praxis, 16(2), 208–222.
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(3), 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.
Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of test statistics to
nonnormality and specification error in confirmatory factor analysis. Psychological
methods, 1(1), 16.
Driessens, O. (2013). Celebrity capital: redefining celebrity using field theory. Theory and
Society, 42(5), 543–560.
Emmons, R. A. (1987). Narcissism: theory and measurement. Journal of Personality and
Social Psychology, 52(1), 11.
Finkelstein, S., & Hambrick, D. C. (1990). Top-Management-Team Tenure and Organizational Outcomes: The Moderating Role of Managerial Discretion.
Administrative Science Quarterly, 35(3), 484–503. https://doi.org/10.2307/2393314
Fombrun, C. J., Ponzi, L. J., & Newburry, W. (2015). Stakeholder Tracking and Analysis: The RepTrak® System for Measuring Corporate Reputation. Corporate Reputation
Review, 18(1), 3–24. https://doi.org/10.1057/crr.2014.21
Fralich, R. (2012). Do Prestigious CEOs Take More Risks? Journal of Managerial Issues, 232–249.
Geletkanycz, M. A., & Hambrick, D. C. (1997). The External Ties of Top Executives:
Implications for Strategic Choice and Performance. Administrative Science Quarterly,
42(4), 654–681. https://doi.org/10.2307/2393653
Gerstner, W.-C., König, A., Enders, A., & Hambrick, D. C. (2013). CEO Narcissism, Audience Engagement, and Organizational Adoption of Technological
Discontinuities. Administrative Science Quarterly, 58(2), 257–291.
Golden, B. R. (1992). The Past Is the Past--Or Is It? The Use of Retrospective Accounts as Indicators of past Strategy. The Academy of Management Journal, 35(4), 848–860. https://doi.org/10.2307/256318
Groeneveld, R. A., & Meeden, G. (1984). Measuring Skewness and Kurtosis. Journal of the
Royal Statistical Society. Series D (The Statistician), 33(4), 391–399.
https://doi.org/10.2307/2987742
Jordan, A. H., & Audia, P. G. (2012). Self-Enhancement and Learning from Performance Feedback. Academy of Management Review, 37(2), 211–231.
https://doi.org/10.5465/amr.2010.0108
Judge, T. A., LePine, J. A., & Rich, B. L. (2006). Loving yourself abundantly: relationship of the narcissistic personality to self-and other perceptions of workplace deviance, leadership, and task and contextual performance. Journal of Applied Psychology,
91(4), 762.
Keene, O. N. (1995). The log transformation is special. Statistics in Medicine, 14(8), 811– 819. https://doi.org/10.1002/sim.4780140810
Kuhnen, C. M., & Niessen, A. (2012). Public Opinion and Executive Compensation.
Management Science, 58(7), 1249–1272.
London, M. (1983). Toward a Theory of Career Motivation. The Academy of Management
Review, 8(4), 620–630. https://doi.org/10.2307/258263
McCall, L. (2005). Do they know and do they care? Americans’ awareness of rising inequality. Work. Pap., Russell Sage Found.
Menkhoff, L., Schmidt, U., & Brozynski, T. (2006). The impact of experience on risk taking, overconfidence, and herding of fund managers: Complementary survey evidence.
European Economic Review, 50(7), 1753–1766.
https://doi.org/10.1016/j.euroecorev.2005.08.001
Merkl-Davies, D. M., & Brennan, N. M. (2011). A conceptual framework of impression management: new insights from psychology, sociology and critical perspectives.
Accounting and Business Research, 41(5), 415–437.
https://doi.org/10.1080/00014788.2011.574222
Michel, J. S., & Bowling, N. A. (2013). Does Dispositional Aggression Feed the Narcissistic Response? The Role of Narcissism and Aggression in the Prediction of Job Attitudes and Counterproductive Work Behaviors. Journal of Business and Psychology, 28(1), 93–105.
Miller, D., & Droge, C. (1986). Psychological and Traditional Determinants of Structure.
Administrative Science Quarterly, 31(4), 539. https://doi.org/10.2307/2392963
Miller, D., & Toulouse, J.-M. (1986). Chief Executive Personality and Corporate Strategy and Structure in Small Firms. Management Science, 32(11), 1389–1409.
Morf, C. C., & Rhodewalt, F. (2001). Unraveling the paradoxes of narcissism: A dynamic self-regulatory processing model. Psychological Inquiry, 12(4), 177–196.
Pukelsheim, F. (1994). The Three Sigma Rule. The American Statistician, 48(2), 88–91. https://doi.org/10.2307/2684253
Raskin, R., Novacek, J., & Hogan, R. (1991). Narcissistic self-esteem management. Journal
of Personality and Social Psychology, 60(6), 911.
Reputation Institute. (2018). RepTrak® Methodology. Retrieved March 28, 2018, from https://www.reputationinstitute.com/reputation-measurement-services/reptrak-framework
Rijsenbilt, A., & Commandeur, H. (2013). Narcissus Enters the Courtroom: CEO Narcissism and Fraud. Journal of Business Ethics, 117(2), 413–429.
Ross, J., & Staw, B. M. (1993). Organizational Escalation and Exit: Lessons from the Shoreham Nuclear Power Plant. The Academy of Management Journal, 36(4), 701– 732. https://doi.org/10.2307/256756
Rotter, J. (1966). Generalized Expectancies for Internal Versus External Control of Reinforcement. Psychological Monographs: General and Applied, 80(1).
Sitkin, S. B., & Weingart, L. R. (1995). Determinants of Risky Decision-Making Behavior: A Test of the Mediating Role of Risk Perceptions and Propensity. The Academy of
Management Journal, 38(6), 1573–1592. https://doi.org/10.2307/256844
Staw, B. M. (1981). The Escalation of Commitment to a Course of Action. The Academy of
Management Review, 6(4), 577–587. https://doi.org/10.2307/257636
Staw, B. M., & Epstein, L. D. (2000). What Bandwagons Bring: Effects of Popular Management Techniques on Corporate Performance, Reputation, and CEO Pay.
Administrative Science Quarterly, 45(3), 523–556. https://doi.org/10.2307/2667108
Van de Ven, A. H., & Polley, D. (1992). Learning While Innovating. Organization Science,
3(1), 92–116.
Wade, J. B., Porac, J. F., Pollock, T. G., & Graffin, S. D. (2006). The Burden of Celebrity: The Impact of Ceo Certification Contests on Ceo Pay and Performance. The Academy
of Management Journal, 49(4), 643–660. https://doi.org/10.2307/20159790
Wales, W. J., Patel, P. C., & Lumpkin, G. T. (2013). In Pursuit of Greatness: CEO
Narcissism, Entrepreneurial Orientation, and Firm Performance Variance: In Pursuit of Greatness. Journal of Management Studies, 50(6), 1041–1069.
https://doi.org/10.1111/joms.12034
Wallace, H. M., & Baumeister, R. F. (2002). The performance of narcissists rises and falls with perceived opportunity for glory. Journal of Personality and Social Psychology,
82(5), 819–834. https://doi.org/10.1037//0022-3514.82.5.819
Westphal, J. D., & Deephouse, D. L. (2011). Avoiding Bad Press: Interpersonal Influence in Relations Between CEOs and Journalists and the Consequences for Press Reporting
About Firms and Their Leadership. Organization Science, 22(4), 1061–1086. https://doi.org/10.1287/orsc.1100.0563
Westphal, J. D., Park, S. H., McDonald, M. L., & Hayward, M. L. A. (2012). Helping Other CEOs Avoid Bad Press: Social Exchange and Impression Management Support among CEOs in Communications with Journalists. Administrative Science Quarterly,
57(2), 217–268.
Zavyalova, A., Pfarrer, M. D., Reger, R. K., & Shapiro, D. L. (2012). Managing the message: the effects of firm actions and industry spillovers on media coverage following
wrongdoing. The Academy of Management Journal, 55(5), 1079–1101.
Zhang, Y., Liu, Y., & Jiraporn, P. (2011). CEO Reputation and Corporate Risk Taking. SSRN
