DO CEOS ACT UPON THEIR STEREOTYPES?
GENDER, AGENTIC, COMMUNAL AND THE EFFECT OF CRISIS
Master Thesis, MSc Human Resource Management University of Groningen, Faculty of Economics and Business
June 4, 2016
ASTRID VOS Student number: s2190826
Supervisors:
Prof.dr. J.I. Stoker Dr. T. Vriend
2 DO CEOS ACT UPON THEIR STEREOTYPES?
GENDER, AGENTIC, COMMUNAL AND THE EFFECT OF CRISIS
ABSTRACT
The amount of female chief executive officers (CEO) in organizations increases despite the male leader stereotype. This male leader stereotype brings female CEOs in the ‘double bind’
position in which they are expected to live up to the masculine expectations of their leader role as well as show that they are ‘good’ women. This ‘double bind’ position might affect female CEO communication. Up till now, it is unclear whether male and female CEOs communicate differently to the outside world, and especially whether female CEOs stress feminine (communal) or masculine (agentic) attributes in their written communication. This research investigates the use of communal and agentic words in the communication of both male and female CEOs. Furthermore, it is suggested that crisis affect the communication of CEOs. More specific, theory suggests that an external crisis requires agentic attributes.
Therefore, this research investigates if CEOs in general used more agentic words after an external crisis than before, and whether this effect is stronger for female than male CEOs.
Text analysis is used to research 280 CEO letters written in 2007 and 2009 to answer the research question. Results show that female CEOs used significantly more communal words than male CEOs in 2007. However, there was no significant difference in the degree of agentic communication between male and female CEOs. Also, this research found no significant effect of the external crisis on the use of agentic words, either in male or female CEOs.
Keywords: female leadership, CEO, double bind theory, crisis, agentic, communal
3 INTRODUCTION
In the last couple of years, the amount of female CEOs has increased significantly. At firms listed in the S&P 1500, the amount of female CEOs increased from 1 female CEO in 1992, to 61 female CEOs in 2004 (Adams, Gupta, Haughton, & Leeth, 2007). According to the Credit Suisse Research (2014), the amount of female CEOs continues to increase in the next couple of years. However, although there is growth in the amount of female CEOs, the percentage of total female CEOs is still very low. Currently, 4% of CEOs are women (Catalyst, 2015). This has been explained by the leader stereotype ‘think manager – think male’ (Schein, 1973). This means that people tend to have similar expectations about leaders and men, but dissimilar expectations about leaders and women (Schein, Mueller, & Jacobson, 1989). As a result, it is difficult for women to rise from the ranks to become a leader within an organization, especially to become a CEO (Rosette & Tost, 2010).
Moreover, once a woman becomes a CEO, the ‘think manager – think male’
stereotype might bring female CEOs in the ‘double bind’ position (Appelbaum, Audet, &
Miller, 2002). In this situation, female leaders are expected to live up the masculine expectations of their work as well as preserve their female qualities and values at the same time (Eagly, 1987). According to Eagly (1987), the traditional gender role can be explained by agentic (stereotypically masculine) and communal (stereotypically feminine) attributes.
Agentic and communal are two different scales which are used to indicate stereotyping, and are derived from masculinity and femininity of the Bem Sex-Role Inventory (Bem, 1974).
These agentic and communal gender roles consist of rules regarding the behavior of men and women. This causes people to have different expectations of male and female leaders (Eagly
& Johnson, 1990). According to gender roles, firstly, women, and thus female CEOs, are expected to show communal attributes, such as being affectionate, empathic and cooperative (Williams & Best, 1990; Eagly & Johannesen-Schmidt, 2001; Sterkel, 1986). Secondly, men are expected to show agentic attributes which are generally perceived to show both in their role as ‘good’ men, and in their role as ‘good’ leaders (Schein, 1973). Agentic attributes include goal-achievement and task functioning like controlling, confident tendency, and assertiveness (Eagly & Johannesen-Schmidt, 2001; Abele & Wojciszke, 2014). However, in view of the ‘double bind’ theory (Eagly, 1987), women are also expected to exhibit the agentic attributes to be perceived as ‘good’ leaders (Powell & Butterfield, 1979).
It is unclear how the ‘double bind’ position influences the behavior of female CEOs.
One of the possible influences could be on the style in which female CEOs communicate.
4 Communication of the CEO is important for companies because proper communication with both internal and external stakeholders might increase the firms’ competitive advantage (Ranft, Zinko, Ferris, & Buckley, 2006). A CEO without good communication skills is less likely to be trusted, believed, or followed (Ha & Riffe, 2015). According to Campus (2013), language is a good indicator of the influence of the ‘double bind’.
It is currently unclear whether female CEOs use a different amount of agentic and communal attributes in communication than male CEOs. According to their gender role, female CEOs are expected to show more communal communication than male CEOs. From a leader perspective, female CEOs are also expected to show at least the same amount, or perhaps even more agentic communication than male CEOs to show that they are ‘good’
leaders. This is the subject of the first part of this study.
Crises might have impact on the perception of what constitutes a ‘good’ leader and by extent, the communication of a leader, because after a crisis different traits are requested from a leader than before a crisis (Haddon, Loughlin, & McNally, 2015). Different theoretical expectations of leadership attributes as an effect of crisis exist, namely the ‘glass cliff’ theory versus the ‘male warrior’ hypothesis. The ‘glass cliff’ theory stresses the need for communal attributes in times of internal crisis (Ryan & Haslam, 2005), including managing issues in the workforce, coping with failure, and fostering cooperation (Brückmuller, Ryan, Rink, &
Haslam, 2014; Mio, Fasan, & Ros, 2016). However, there is also theoretical and empirical work that would suggest other required attributes due to another type of crisis. That is the
‘male warrior’ hypothesis in which external threat what might be the case after an external crisis, requires agentic attributes such as risk taking, defending the group, and increasing status (Van Vugt, Cremer, & Janssen, 2007; McDonald, Navarrete, & Van Vugt, 2012).
According to Greenglass et al. (2014), the most recent crisis is the financial crisis of 2008, which affected countries all over the world. The financial crisis occurred outside the spheres of influence of organizations which indicates that this is an external crisis (Greenglass et al., 2014; Coombs, 1998). In light of this global crisis, it is interesting to research whether the financial crisis influenced the communication of CEOs. Following the ‘male warrior’
hypothesis (Van Vugt et al., 2007), it is expected that CEOs in general show more agentic communication after than before the crisis. Furthermore, it is expected that this effect will be stronger for female CEOs than male CEOs since female CEOs might use more agentic communication than their male counterparts in order to show that they are ‘good’ leaders.
5 There is little research (Kirtley & Weaver III, 1999; Hupfer, 2002) on possible differences in communication styles between male and female CEOs, and especially on the possible differences in communal and agentic attributes. Since communication is important for organizations, more information on gender-specific differences in communication is important. Therefore, the present research will firstly investigate communication of female CEOs, and compare it with the communication of male CEOs. Following Eagly (1987), we expect that female CEOs communicate in general more communal than male CEOs.
Following Powell and Butterfield (1979) and the female gender role, we also expect that female CEOs will communicate more agentic than male CEOs. Secondly, we expect that crises have an impact on the communication style of CEOs in general. Following the ‘male warrior’ hypothesis (Van Vugt et al., 2007), we expect that CEOs use more agentic communication after a crisis than before. More specific, we expect that this effect will be stronger for female CEOs than for male CEOs.
This study has two theoretical contributions. First of all, this study provides insight into the ‘double bind’ theory because actual use of communal and agentic communication of female and male CEOs is currently unknown. In the second place, this study provides insight into the effect of crises on communication styles. Following the ‘male warrior’ hypothesis (Van Vugt et al., 2007), agentic attributes are required. However, it is currently unknown what the effect of an external crisis is on the use of actual agentic communication.
THEORY SECTION Female CEOs and the ‘double bind’
A CEO is the chief executive officer with the highest rank within a company (Elsaid &
Ursel, 2011). The CEO develops and implements high-level strategies, is responsible and accountable for operational activities, manages both internal and external environment, and has an important role in the communication between internal and external stakeholders (Fanelli & Misangyi, 2006; Kakabadse, Kakabadse, & Barratt, 2006; Mullins & Schoar, 2015). Internal and external stakeholders differ from each other and require different communication styles from the CEO to manage them both (Fanelli & Misangyi, 2006). These differences are based on internal stakeholders who are relatively “captive” to the CEO which means that internal stakeholders are dominated by and dependent on the CEO. Additionally, these differences are based on external stakeholders performing in a competitive environment
6 with multiple actors and multiple firms meaning that external stakeholders are less dominated by the CEO than internal stakeholders (Thompson, 1967; Zuckerman, 2000; Fanelli &
Misangyi, 2006). It is important for the CEO to establish proper communication with all stakeholders to be trusted, believed, and followed (Ha & Riffe, 2015).
Oakley (2000) mentioned that the traditional network of CEOs is an ‘old boy’s network’.
This refers to the traditional all-male network in which men generate impediments toward women’s advances in organizations, foster solidarity between men, and threaten women (Rigg
& Sparrow, 1994). The leader stereotype is male, ‘think manager, think male’, in which managerial successes are more likely to be held by men than by women (Schein, 1975). To indicate gender based stereotyping, Bem (1974) constructed and tested (Bem Sex Role Inventory) two scales, namely femininity and masculinity instead of an unidimensional view of masculinity-femininity. This means that before, masculinity and femininity were seen as one dimension, where Bem used these two as separate dimensions (Bem, 1974; Brannon, 2005). Both stereotypes femininity and masculinity are beliefs and attitudes which are generally associated with women and men (Bem, 1974). These dimensions were also researched by Spence, Helmreich and Stapp (1974) in the Personal Attributes Questionnaire.
However, one of these researchers (Spence, 1985) acknowledged the weaknesses of the conceptualization of femininity and masculinity. Therefore, agentic and communal scales (Bakan, 1966) were more adopted. Spence (1985) suggested that agentic and communal refer to scores in terms of instrumentality and expressiveness rather than masculinity and femininity. Agentic and communal are the underlying concepts of masculinity and femininity in which agentic refers to assertive which is associated with masculinity, and communal refers to concern with others which is associated with women (Eagly & Wood, 1985). Agentic and communal are two different scales to indicate stereotyping (Bakan, 1966; Eagly & Wood, 1991). People can score high or low on both agentic and communal to explain gender roles since these agentic and communal scales are complementary, not the opposite of each other (Bem, 1974; Powell & Butterfield, 1979; Eagly & Wood, 1985). To research the aspects of gender roles, it is relevant to understand how leadership pertains to agentic and communal attributes (Eagly, Wood, & Diekman, 2000).
In the 1970s, Schein mentioned sex role stereotyping in which she showed that women are less qualified than men for management positions (Schein, 1978). Additional research showed the content of leader stereotype. In general, self-confidence, forcefulness, analytic, emotional stability and ambition are attributes which are viewed as both leadership and male qualities
7 (Heilman, Block, Simon, & Martell, 1989). These male attributes are labeled as agentic which are included in the nature of leadership roles (Eagly & Wood, 1985; Eagly & Karau, 1991).
This means that the leader stereotype includes agentic attributes such as being independent, assertive, competent, and self-confident (Eagly & Karau, 1991). From a gender perspective, men are expected to show these agentic attributes (Eagly, 1987).
This male leader stereotype brings female CEOs in the ‘double bind’ position. This means that, for female CEOs to be considered as stereotypical ‘good’ leaders, they have to possess agentic qualities, but on the other hand also have to show that they are ‘good’ women, that is showing communal characteristics expected from a woman (Eagly, 1987; Eagly & Karau, 1991). This implies that female leaders are considered as not being tough enough or not taking charge if they lack the stereotypical agentic qualities of ‘good’ leaders. Conversely, people dislike female CEOs who display the leader stereotype agentic qualities, because such women are too aggressive, called a ‘bitch’, and seem unfeminine (Eagly, 2007). The ‘double bind’ position historically constrained women because of the impossible combination of caring and understanding people’s worries, and of looking tough enough to take risk for the organization (Campus, 2013). From a stereotypical point of view, communal aspects include caring, expressing emotions, and concerning for other’s wellbeing (Eagly, 1987; Eagly &
Karau, 1991). Additionally, communal also includes traits as being emotional, dependent, friendly, loyal, and cooperative (Bem, 1974; Spence, Helmreich, & Stapp, 1974; Eagly, 1987;
Markus & Kitayama 1991).
However, women are also expected to be ‘good’ leaders by showing agentic attributes (Eagly, 1987; Powell & Butterfield, 1979). Powell and Butterfield (1979, 1989) showed that both men and women described a ‘good’ manager as agentic which were showed by both male and female applicants. This means that leadership is associated with agentic which can be possessed by both men and women (Powell & Butterfield, 1989). According to Powell (1990), there are aspects in management, such as behavior and motivation, in which there are not always gender based differences between male and female managers. This derives from the goal of women to strive for a career as a manager in which they reject the communal stereotype, and have values and needs similar to those men who also strive for a manager position. To become a manager, it is needed for both men and women to see themselves fitting the prevailing stereotype of a manager (Powell, Butterfield, & Parent, 2002).
Furthermore, since top management positions are valued agentic, all individuals who enter the steps at any level towards this top management position will be expected to act according to
8 the agentic leader stereotype (Powell et al., 2002). Thus, female leaders who want to succeed in a top position have to demonstrate at all levels that they could fulfill the agentic expectations associated with the leadership position (Powell et al., 2002; Rosette & Tost, 2010). Concluding, women in any leadership position are expected to show agentic attributes according to their leadership position.
As aforementioned, female CEOs have to deal with the ‘double bind’ position (Eagly, 1987) including their gender stereotype as women and the agentic ‘think manager, think male’
principle (Schein, 1973). In order to be a ‘good’ woman, it is expected that female CEOs communicate more communal attributes than male CEOs. Further, following research of Powell and Butterfield (1979), to be a ‘good’ leader as well, it is expected that female CEOs communicate at least the same amount, or perhaps even more agentic attributes than male CEOs. It is expected that these communal and agentic attributes are found in communication of CEOs. This suggests that it is expected that female CEOs show more communal communication than male CEOs. Therefore, the following hypothesis is proposed:
H1a: female CEOs communicate more communal than male CEOs.
Moreover, it is expected that female CEOs show the same amount or perhaps even more agentic communication than male CEOs. This leads to the following hypothesis:
H1b: female CEOs communicate more agentic than male CEOs.
Gender, leadership and crisis
To understand the prototype ‘good’ leader, it is important to understand the context in which it exists (Haddon et al., 2015). Researches show that the requirements for effective leadership and being a ‘good’ leader vary across different contexts (Lord, Foti, & De Vader, 1984; Lord, Brown, Harvey, & Hall, 2001; Osborn, Hunt, & Jauch, 2002). For instance, what is considered as being a ‘good’ leader during ‘normal’ times might be different than being a
‘good’ leader in times of crisis (Haslam et al., 2001; Hunt, Boal, & Dodge, 1999; Meindl, 1993). To research the effect of crisis, it is important to distinguish between an internal and an external crisis. Firstly, an internal crisis is a precarious circumstance caused within a particular organization which has been caused by organizational failures (Coombs, 1998;
Pearson & Clair, 1998). An external crisis means that the crisis has been caused and occurred
9 outside the spheres of influence of an organization, thus resulting from external contingencies (Coombs, 1998; Pearson & Clair, 1998).
Two stands of literature explain the expected role of a leader within an organization, depending on the nature of the crisis, namely the ‘glass cliff’ theory and the ‘male warrior’
hypothesis.
Firstly, the ‘glass cliff’ theory explains that communal attributes are favored in times of internal crisis (Brückmuller et al., 2014; (Ryan, Haslam, Hersby, & Bongiorna, 2011).
Female leaders are expected to bring communal communication to the table, which is assumed to be helpful in times of uncertainty which is labeled as ‘think crisis – think female’
(Ryan & Haslam, 2005; Brückmuller, et al., 2014). ‘Think crisis – think female’ means that women are favored in times of internal crisis because they are expected to be good people managers and they can take blame for failure (Ryan et al., 2011). According to Brückmuller and colleagues (2014), communal attributes are associated with the ‘ideal’ manager for an unsuccessful organization which might be the case in times of an internal crisis.
Another stand of literature is the ‘male warrior’ hypothesis (Van Vugt et al., 2007).
According to the ‘male warrior’ hypothesis, agentic attributes may be favored in times of external threats. Based on history of male-to-male conflict, men have the cognitive instruments to make coalitions with other men to emerge triumphant in intergroup conflicts to protect resources (Van Vugt, 2011). Participating in intergroup conflict is related to taking risk because of the risk of ‘death’. According to Van Vugt et al. (2007), men are aggressive in intergroup conflicts because they want to save face. Consequentially, people show stronger preference for male than female leaders in times of intergroup competition. Intergroup competition might be the effect of external threats which occurs in all modern and traditional cultures (McDonald et al., 2012; Van Vugt, 2011). In times of external threats, leaders in general are expected to show male-stereotypic attributes because people look for someone who possesses capabilities of planning, initiating and executing to protect resources. Also, these precarious circumstances create the opportunity to increase status and to gain prestige (McDonald et al., 2012; Van Vugt, et al., 2007).
In this paper, the financial crisis of 2008 will be used to research the effect of an crisis on the communication by CEOs, since this is the most recent global crisis (Greenglass et al., 2014). The financial crisis is the global crisis caused by the collapse of the Lehman Brothers Holding INC in September 2008. This was a result of the explosion of the United States real
10 estate bubble (Mio et al., 2016). The financial crisis spread quickly around the world through financial systems with dramatic, unforeseen consequences for the global economy (Raithel, Wilczynski, Schloderer, & Schwaiger, 2010). According to Mio et al. (2016), the financial crisis lead to precarious circumstances, and damaged trust in top executives and their ability to manage their companies. Mio et al. (2016) also mentioned that the main contributing factors for the financial crisis were greed, bad regulation in the financial sector, and lack of transparency. Since the financial crisis has been caused by external contingencies, outside the spheres of influence of a particular organization (Raithel et al., 2010), the financial crisis is an external crisis as mentioned in the ‘male warrior’ hypothesis (Van Vugt et al., 2007).
According to the ‘male warrior’ hypothesis (Van Vugt et al., 2007) including the rise for agentic attributes as an effect of an external crisis, the following hypothesis is proposed:
H2a: CEOs will communicate more agentic after the crisis than before the crisis.
As aforementioned, women are expected to be ‘good’ leaders by showing agentic attributes (Powell & Butterfield, 1979). Therefore, female leaders who want to succeed in a top position have to demonstrate at all levels that they could show the agentic expectations associated with the leader position (Powell et al., 2002; Rosette & Tost, 2010). In times of an external crisis like the financial crisis, it might be that women are disadvantaged because men are preferred (Van Vugt, 2011). To overcome this disadvantage, female CEOs might have to show that they are also ‘good’ leaders despite their sex female. Due to this, it might be that female CEOs use more agentic words than male CEOs.
According to the ‘male warrior’ hypothesis (Van Vugt et al., 2007) and gender expectations, we propose the following hypothesis:
H2b: Female CEOs will show more agentic communication after crisis than before compared to male CEOs.
METHOD Procedure
As aforementioned, CEO linguistic communication is a good indicator of the influence of the ‘double bind’ (Campus, 2013). Linguistic communication is the process of using words to express or exchange information (Luhmann, 1992). CEO letters can be seen as sources of data on such linguistic communication (Hyland, 1998; Kendall, 1993; Palmer, King, & Kelleher, 2004; Prasad & Mir, 2002). CEO letters are one of the most widely read and most important
11 parts of an annual report (Costa, Oliveira, & Rodrigues, 2013). CEO letters are important for reflection of managerial attributions, and are highly indicative of framing strategies (Westphal
& Zajac, 1998). Furthermore, CEOs ideologies, aspirations, and strategic thinking are elements that form a CEO letter, and thus elements for the choice of specific written words in the CEO letter (Amernic & Craig, 2007). CEO letters can be written by the CEO themselves or these can be written in name of the CEO. However, the reader can take the words in the CEO letter as the CEOs own since these words are emblematic and typical for the CEO (Amernic & Craig, 2007). Consequently, this research used CEO letters to investigate the hypotheses.
CEO letters are a voluntary disclosure that have been found to be richer in large firms (Ahmed & Courtis, 1999). The starting point to select CEOs was to first find female CEOs and their CEO letters. Therefore, this research started focusing on large publicly traded firms from all over the world with a female CEO since there were less female CEOs than male CEOs (Mohan & Ruggiero, 2003). Data were collected from the Fortune1000 (USA), S&P1500 (USA), list of female CEOs created by the Catalyst (USA), FTSE100 (UK), Women 1st Top 100 (UK), The CEO Magazine (AUS), Forbes list of Most Powerful Women (worldwide), and a list of female top executives Wikipedia (worldwide). After these sources, we were able to include in total 52 CEO letters from female CEOs. There were not enough CEO letters from female CEOs since there were not many female CEOs, and not all companies publish these CEO letters online. Female chairmen of the Board of Directors created new opportunities since the chairman of the Board of Directors also has an important leadership position within the organization (Kakabadse et al., 2006; Mullins & Schoar, 2015), so we included chairmen as well but in this research they are all labeled as CEO (n=65).
However, after including chairmen, there were still not enough CEO letters from female CEOs. The non-profit sector created new opportunities since performance is not a variable in this paper, only agentic and communal communication of leaders are the dependent variables.
Furthermore, non-profit organizations also have a CEO and chairman (Drucker, 2011). Non- profit organizations which are used in this research were hospitals, universities and cultural organizations since these are female-dominated fields (Mastekaasa & Smeby, 2008). Data were collected from lists of hospitals Wikipedia (UK), list of universities by the government (UK), and list of popular cultural organizations including dancing schools, art councils, and museums (worldwide). In these fields, snowball sampling is used whereby the female sample built up till enough data were gathered to be useful for research (n=140).
12 After gathering CEO letters from female CEOs, CEO letters from male CEOs were gathered as well. The Fortune500 was used for profit organizations, and the same list of hospitals, universities and cultural organizations for non-profit organizations were used because not all of these non-profit organizations were led by women. We tried to match the amount of profit and non-profit organizations with the amount of female CEOs. Again, we used snowball sampling for non-profit organizations to build up the male CEO sample till enough data were gathered to be useful for research (n=140).
To analyze CEO letters, each CEO letter was transcribed to an individual Microsoft Word document. Linguistic In Word Counts (LIWC) (Pennebaker, Booth, Boyd, & Francis, 2015) was used which performed an extensive linguistic analysis on each individual Word- file. Furthermore, LIWC analyzed the Word-files and calculated the percentage of words into agentic and communal linguistic categories. The percentages of agentic and communal words of all CEO texts were analyzed by the LIWC program. This output was viewed in an Excel- file. The statistical analyses were conducted by using SPSS 20.0 statistical software for Windows. An ANOVA test was conducted to test the hypotheses.
Measures Communal
The list of words that are considered to be communal were originating from the Bem Sex-Role Inventory (Bem, 1974) and the Personal Attributes Questionnaire (Spence, Helmreich, & Stapp, 1974). These are widely used instruments to measure gender role perceptions. Also, articles from Pennebaker, Mehl and Niederhoffer (2003), Gaucher, Friesen and Kay (2011), Madera, Hebl and Martin (2009), and Bartz and Lydon (2004) were used to compose the list. To measure communal words, a dictionary of 107 communal words was created including caring, commitment, communication, cooperation, connection, empathy, engagement, and loyal. The complete list of communal words is added in Appendix A.
Agentic
To measure agentic words, there was also a list of words created that are considered to be agentic deriving from Bem (1974), Spence et al. (1974), Pennebaker et al. (2003), Gaucher et al. (2011), Madera et al. (2009), and Bart and Lydon (2004). This final list of 107 agentic words includes words as ambitious, assertiveness, autonomous, analytic, competitive,
13 independent, leader, and opportunities. The complete list of agentic words is added in Appendix A.
Sex
In our sample in 2007, there were 50% female CEOs and 50% male CEOs, and in 2009, there were also 50% female CEOs and 50% male CEOs. All groups were independent groups, meaning that the CEOs before the crisis were different from the CEOs after the crisis.
Sex was coded as men (0) and women (1).
Year
In order to test the hypotheses, data were collected from male and female CEOs in 2007 and 2009. Due to the collapse of Lehman Brothers Holdings, INC in September 2008, after crisis was measured in 2009, a year right after the financial crisis. To avoid any influences of the financial crisis, hypotheses 1a and 1b were measured by CEO letters of 2007. In 2007, the full sample was 140 CEOs, and in 2009, there were also in total 140 CEOs.
Years were coded as 2007 (0) and 2009 (1).
Control variables
Age is a control variable since age has an effect on communal and agentic in self- representation of adults (Diehl, Owen, & Youngblade, 2004). These authors (2004) found that the use of communal words increases with age, and younger adults include more agentic attributes than older adults. As a result, it is possible that age has an effect on the use of communal and agentic words of CEOs.
Sector is another control variable used in this paper divided into profit and non-profit sector. The profit sector strives for making profits, where non-profit organizations does not strive for making money but strives for adding social value. Both have an impact on the style of management and the type of leaders of an organization (Osula & Ng, 2014). Due to the differences between the profit and non-profit sector, it might have an impact on communal and agentic words of CEOs. In 2007, 140 organizations were represented with a concentration of 71 organizations in the non-profit and 69 organizations in the profit sector. In 2009, also organizations were represented with a concentration of 68 organizations in the non-profit and 72 organizations in the profit sector for both men and women.
14 RESULTS
Descriptive statistics
The descriptive statistics are summarized in Table 1.
TABLE 1: Descriptive Statistics
M SD
Total Male Female Total Male Female Variable 2007 2009 2007 2009 2007 2009 2007 2009 Communal (%) 6.88 6.43 6.53 7.30 7.24 2.22 2.23 2.25 2.24 2.14 Agentic (%) 3.91 3.95 3.97 3.83 3.90 1.20 1.22 1.31 1.09 1.16 Age (year) 55.23 57.04 57.33 53.10 53.43 7.62 7.54 7.40 7.29 7.30
Hypotheses testing Hypothesis 1
Hypothesis 1a predicts that women use more communal communication than men. To test the hypothesis, an ANOVA test was conducted. There was a significant difference in the percentage of communal words between men and women at the p < .05 level for the two conditions [F (1,138) = 5.30, p = 0.023]. Comparison of the mean indicated that women use more communal words than men in a CEO letter in 2007, thus the results supported Hypothesis 1a.
To test the interaction effect Sex X Age, an ANOVA test was conducted. There was no significant effect at the p < .05 level [F (21,85) = 0.82, p = 0.690]. This means that the effect of more communal words of women than men is not related with CEOs age.
Secondly, an ANOVA test was also conducted to test the interaction effect of Sex X Sector. There was no significant effect at the p < .05 level [F (1,136) = 1.53, p = 0.218]. This means that the effect of more communal words of women than men is not related with the sector of the organization.
Hypothesis 1b predicts that women use more agentic communication than men. There was no significant difference in the percentage of agentic words between men and women at the p < .05 level of the two conditions [F (1,138) = 0.34, p = 0.563]. These results suggest that women do not use more agentic words than men in a CEO letter in 2007. Thus, Hypothesis 1b was not supported.
15 To control for age, an ANOVA test was conducted to test the interaction effect Sex X Age. There was no significant effect at the p < .05 level [F (21,85) = 0.49, p = 0.968].
Therefore, age does not affect the use of agentic words differently in men and women.
Secondly, an ANOVA test was also conducted to test the interaction effect of Sex X Sector. There was no a significant effect at the p < .05 level [F (1,136) = 0.62, p = 0.432].
Therefore, the sector does not affect the use of agentic words differently in men and women.
Hypothesis 2
Hypothesis 2a predicts that CEOs communicate in general more agentic words after than before crisis. To test the hypothesis, an ANOVA test was conducted. There was no significant difference in the percentage of agentic words between 2007 and 2009 at the p <
.05 level [F (1,278) = 0.110, p = 0.740]. These results suggest that by male and female CEOs, in 2009 there are no more agentic words used than in 2007 in a CEO letter. Thus, Hypothesis 2a was not supported.
Hypothesis 2b predicts that female CEOs will show more agentic communication after crisis than before compared to male CEOs. For this, an ANOVA test was conducted to test the interaction effect of Year X Sex. There was no significant effect at the p < .05 level [F (1,276) = 0.018, p = 0.894]. This means that female CEOs do not show more agentic communication than male CEOs after than before crisis, thus Hypothesis 2b was not supported.
In order to test the robustness of our findings, we also investigated H2a and H2b for communal words. First of all, an ANOVA test was conducted to measure if CEOs communicate in general more communal words in CEO letters after than before the crisis.
There was no significant difference in the percentage of communal words between 2007 and 2009 at the p < .05 level [F (1,278) = 0.007, p = 0.934]. These results suggest that by male and female CEOs, in 2009 there are no more communal words used than in 2007.
Secondly, an ANOVA test was conducted to test the interaction effect of Year X Sex for communal words. There was no significant effect at the p < .05 level [F (1,276) = 0.092, p
= 0.762]. This means that the effect of crisis on communal words is not stronger for women.
16 DISCUSSION
The goal of this research was to address gender stereotypes, the leader prototype, the actual way of communication of female and male CEOs, and the effect of crisis on the actual way of communication. This paper has been divided into two main topics. First of all, the
‘double bind’ theory. We assumed that female CEOs communicate more communal than male CEOs. Also, we assumed that female CEOs communicate more agentic than male CEOs. Our results confirmed the expectation that female CEOs communicate more communal words than male CEOs. However, the expectation that female CEOs communicate more agentic words than male CEOs was not supported by the results.
Secondly, the effect of crisis is another main topic. We assumed that CEOs in general use more agentic communication after than before crisis. Furthermore, we assumed that female CEOs will show more agentic communication after crisis than before compared to male CEOs. The results did not confirm either expectation, since there were for CEOs in general no differences in agentic communication after crisis than before crisis. Also, there are no differences between male and female CEOs in the use of agentic words after crisis than before crisis.
Theoretical implications
Our findings have several theoretical implications. First of all, the results of this research support the ‘double bind’ theory (Eagly, 1987). In 2007, female CEOs show more communal words than male CEOs, which is in line with the expectations inherent to their female gender role mentioned by Eagly (1987). Furthermore, in spite of the ‘think manager, think male’ principle (Schein, 1973), female CEOs and male CEOs do not differ in agentic words. Regarding to the ‘double bind’ theory (Eagly, 1987), it seems that female CEOs in 2007 are both ‘good’ women and ‘good’ leaders since there is no significant difference in agentic communication between female and male CEOs which suggests that female CEOs might be ‘good’ leaders as long as men are seen as ‘good’ leaders. Thus, due to more communal and equal agentic words of female CEOs compared to male CEOs, it seems that female CEOs are both ‘good’ women and ‘good’ leaders in 2007.
This research adds also value to the ‘think manager, think male’ principle. Related to this principle, Schein (1973) mentioned that agentic attributes are leadership qualities. Results of the current research show that both male and female CEOs use on average more communal than agentic words in both 2007 and 2009. This might imply that ‘think manager, think male’
17 is associated with sex and not with gender meaning that men are traditionally more favored to be manager than women because of their sex men, and not because of favored agentic attributes. This is proven by the current research since the results showed that both men and women communicate in general more communal than agentic.
Practical implications
This study adds value to courses about leadership and gender. Nowadays, since the importance of diversity within organizations, courses about leadership and gender are upcoming and widely used. These findings give input for courses about leadership and gender, more specific about female CEOs and their use of communal and agentic attributes.
These findings can be used to specify courses about leadership and gender.
Limitations and suggestions for future research
This study is not without limitations. First of all, the ‘male warrior’ hypothesis was not supported in this research. This research only focused on 2007 and 2009 to measure the impact of the financial crisis. Possibly, the effect of the financial crisis does not occur significantly on the short-term. For the future, we suggest to extent the time-period to measure the impact of the financial crisis.
Another limitation is the inclusion of female dominated fields. We have chosen for non-profit organizations in female dominated fields since female CEOs were difficult to find.
Actually, this can influence the amount of communal words. On average, both male and female CEOs communicate in general more communal than agentic. This can be caused by the female dominated fields within the non-profit sector. One clear area for further research is to explore organizations of multiple, different areas, not only from female dominated fields.
In addition, the current study provides input for more research into the ‘double bind’
theory. In this study, in 2007 it seems that women are ‘good’ women and ‘good’ leaders at the same time looking at the amount of communal and agentic words. It is interesting to investigate whether female leaders seem to be still both ‘good’ women and ‘good’ leaders.
CONCLUSION
Findings from this research show that female CEOs in 2007 use more communal words than male CEOs. Thereby, it shows that male and female CEOs do not differ in the
18 communication of agentic words. These findings are important for the ‘double bind’ theory because it seems that, in 2007, female CEOs are both ‘good’ women and ‘good’ leaders at the same time. There were no significant results found for the impact of crisis on communication, this might be caused by the limited amount of years investigated in this paper. Future research is needed to further investigate the ‘double bind’ theory. Also, it is interesting to extent the time-period to understand whether crisis has a long-term impact on communal and agentic communication.
19 REFERENCES
Abele, A., & Wojciszke, B. (2014). Communal and Agentic Content in Social Cognition: A Dual Perspective Model. Advances in Experimental Social Psychology, 50, 195-255.
Adams, S., Gupta, A., Haughton, D., & Leeth, J. (2007). Gender differences in CEO compensation: evidence from the U.S.. Women in Management Review, 22, 208–224.
Ahmed, K., & Courtis, J. (1999). Association between corporate characteristics an disclosure levels in annual reports: a meta-analysis. The British Accounting Review, 31(1), 35-61.
Amernic, J., & Craig, R. (2007). Making CEO-speak more potent: editing the language of corporate leadership. Strategy & Leadership, 35(3), 25-31.
Appelbaum, S., Audet, L, & Miller, J. (2003). Gender and leadership? Leadership and gender? A journey through the landscape of theories. Leadership & Organization Development Journal, 24(1), 43-51.
Bakan, D. (1966). The duality of human existence. Reading, PA: Addison Wesley.
Bartz, J., & Lydon, J. (2004). Close Relationships and the Working Self-Concept:
Implicit and Explicit Effects of Priming Attachment on Agency and Communion.
Personality and Social Psychology Bulletin, 30(11), 1389-1401.
Bem, S. (1974). The measurement of psychological androgyny. Journal of Clinical and Consulting Psychology, 42, 155-162.
Brannon, L. (2005). Gender: Psychological Perspectives. Boston, MA: Allyn & Bacon.
Brückmuller, S., Ryan, M., Rink, F., & Haslam, S. (2014) Beyond the Glass Ceiling: The Glass Cliff and Its Lessons for Organizational Policy. Social Issues and Policy Review, 8 (1), 202-232.
Campus, D. (2013). Women Political Leaders and the Media. New York, NY: Palgrave Macmillan.
Catalyst. Women CEOs of the S&P 500. (2016, February 4). Retrieved from:
http://www.catalyst.org/knowledge/women-ceos-sp-500
20 Coombs, W. (1998). An analytic framework for crisis situations: Better responses from a
better understanding of the situation. Journal of Public Relations Research, 10(3), 177-191.
Costa, G., Oliveira, L., & Rodrigues, L. (2013). Factors associated with the publication of a CEO letter. Corporate Communications: An Internal Journal, 18(4), 432-450.
Dawson, J., Kersley, R., & Natella, S. (2014). The CS Gender 3000: Women in Senior Management. Zurich: Credit Suisse.
Diehl, M., Owen, S., & Youngblade, L. (2004). Agency and communion attributes in adults’
spontaneous self-representations. International Journal of Behavioral Development, 28(1), 1-15.
Drucker, P. (2011). Managing the non-profit organization. New York, NY: Routledge Taylor & Francis Group.
Eagly, A. (1987). Sex differences in social behavior: A social-role interpretation. Hillsdale, NJ: Erlbaum.
Eagly, A. (2007). Female leadership advantage and disadvantage: resolving the contradictions. Psychology of Women Quarterly, 31, 1-12.
Eagly, A., & Johannesen-Schmidt, M. (2001). The leadership styles of women and men.
Journal of Social Issues, 57(4), 781-797.
Eagly, A., & Johnson, B. (1990). Gender and Leadership Style: A Meta-Analysis.
Psychological Bulletin, 108(2), 233-256.
Eagly, A., & Karau, S. (1991). Gender and the emergence of leaders: a meta-analysis. Journal of Personality and Social Psychology, 60(5), 685-710.
Eagly, A., & Wood, W. (1985). Gender and influenceability: Stereotype vs. behavior. In O'Leary, V., Unger, R., & Wallston, B. (Eds.), Women, gender, and social psychology (pp. 225-256). Hillsdale, NJ: Erlbaum.
Eagly, A., & Wood, W. (1991). Explaining sex differences in social behavior: A meta analytics perspective. Personality and social psychology bulletin, 17(3), 306-315.
21 Eagly, A., Wood, W., & Diekman, A. (2000). Social role theory of sex differences and
similarities: A current appraisal. In T. Eckes & H. M. Trautner (Eds.), The developmental social psychology of gender (pp. 123–174). Mahwah, NJ: Erlbaum.
Elsaid, E., & Ursel, N. (2011). CEO succession, gender and risk taking. Gender in Management: An International Journal, 26(7), 499-512.
Fanelli, A., & Misangyi, V.F. (2006). Bringing out Charisma: CEO Charisma and External Stakeholders. The Academy of Management Review, 31(4), 1049-1061.
Gaucher, D., Friesen J., & Kay, A. (2011). Evidence That Gendered Wording in Job Advertisements Exists and Sustains Gender Inequality. Journal of Personality and Social Psychology, 101(1), 109-128.
Greenglass, E., Antonides, G., Christandl, F., Foster, G., Katter, J., Kaufman, B., & Lea, S., (2014). The financial crisis and its effects: perspectives from economics and psychology. Journal of Behavioral and Experimental Economics, 50, 10-12.
Ha, H., & Riffe, D. (2015). Crisis-related research in communication and business journals:
An interdisciplinary review from 1992 – 2011. Public Relations Review, 41, 569-578.
Haddon, A., Loughlin, C., & McNally, C. (2015). Leadership in a time of financial crisis:
what do we want from our leaders?. Leadership & Organization Development Journal, 36(5), 612-627.
Haslam, S., Platow, M., Turner, J., Reynolds, K., McGarty, C., & Oakes, P. (2001). Social identity and the romance of leadership: the importance of being seen to be ‘doing it for us’. Group Processes and Intergroup Relations, 4(3), 191-205.
Heilman, M., Block, C., Martell, R., & Simon, M. (1989). Has anything changed? Current characterizations of men, women, and managers. Journal of Applied Psychology, 74, 935-942.
Hupfer, M. (2002). Communicating with the Agentic Woman and the Communal Man: Are Stereotypic Advertising Appeals Still Relevant. Academy of Marketing Science Review, 3, 1-15.
Hunt, J., Boal, K., & Dodge, G. (1999). The effects of visionary and crisis-responsive leadership. Leadership Quarterly, 10(3), 423-448.
22 Hyland, K. (1998). Exploring Corporate Rhetoric: Metadiscourse in the CEO's Letter.
Journal of Business Communication, 35, 224-245.
Kakabadse, A., Kakabadse, N., & Barratt, R. (2006). Chairman and Chief Executive Officer (CEO): That Sacred and Secret Relationship. Journal of Management Development, 25(2), 1-28.
Kendall, J. (1993), Good and Evil in the Chairman's "Boiler Plate": An Analysis of Corporate Visions of the 1970s. Organization Studies, 14, 571-592.
Kirtley, M., & Weaver III, J. (1999). Exploring the Impact of Gender Role Self-Perception on Communication Style. Women’s Studies in Communication, 22(2), 190-209.
Lord, R., Foti, R., & De Vader, C. (1984). A test of leadership categorization theory:
internal structure, information processing, and leadership perceptions. Organizational Behavior and Human Performance, 34(3), 343-378.
Lord, R., Brown, D., Harvey, J., & Hall, R. (2001). Contextual constraints on prototype generation and their multilevel consequences for leadership perceptions. Leadership Quarterly, 12(3), 311-338.
Luhmann, N. (1992). What is Communication?. Communication Theory, 2(3), 251-259.
Madera, J., Hebl, M., & Martin, R. (2009). Gender and Letters of Recommendation for Academia: Agentic and Communal Differences. Journal of Applied Psychology, 94(6), 1591-1599.
Markus, H., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 20 568-579.
Mastekaasa, A., & Smeby, J. (2008). Educational choice and persistence in male- and female-dominated fields. Higher Education, 55, 189-202.
McDonald, M., Navarrete, C., & Vugt, van, M. (2012). Evolution and the psychology of intergroup conflict: the male warrior hypothesis. Philosophical Transactions of the Royal B Society, 367, 670-679.
Meindl, J. (1993). Reinventing leadership: a radical social psychological approach, in Murninghan, J.K. (Ed.). Social Pyschology in Organizations (pp. 89-118). Englewood Cliffs, NJ: Prentice Hall.
23 Mio, C., Fasan, M., & Ros, A. (2016) Owners’ preferences for CEOs characteristics: did the
world change after the global financial crisis?, Corporate Governance, 16(1), 116-134.
Mohan, N., & Ruggiero, J. (2003). Differences between Male and Female CEOs for Publicly Traded Firms: A Nonparametric Analysis. The Journal of the Operational Research Society, 54(1), 1242-1248.
Mullins, W., & Schoar, A. (2015). How do CEOs see their roles? Management philosophies and styles in family and non-family firms. Journal of Financial Economics, 119(1), 24-43.
Oakley, J. (2000). Gender-Based Barriers to Senior Management Positions: Understanding the Scarcity of Female CEOs. Journal of Business Ethics, 27(4), 321-334.
Osborn, R., Hunt, J., & Jauch, L. (2002). Toward a contextual theory of leadership, The Leadership Quarterly, 13(6), 797-837.
Palmer, I., King, A., & Kelleher, D. (2004). Listening to Jack: GE's Change Conversations with Shareholders. Journal of Organizational Change Management, 17, 593-614.
Pearson, C., & Clair, J. (1998). Reframing crisis management. Academy of Management Review, 23(1), 59-76.
Pennebaker, J., Booth, R., Boyd, R., & Francis, M. (2015). Linguistic Inquiry and Word Count: LIWC 2015 [Computer software]. Austin, TX: LIWC.
Pennebaker, J., Mehl, M., & Niederhoffer, K. (2003). Psychological Aspects of Natural Language Use: Our Words, Our Selves. Annual Review Psychology, 54, 547-577.
Powell, G. (1990). One more time: do female and male managers differ? The Executive, 4(3), 68-75.
Powell, G., & Butterfield, D. (1979). The “Good Manager”: Masculine or Androgynous?
Academy of Management Journal, 22(2), 395-403.
Powell, G., & Butterfield, D. (1989). The “Good Manager”. Did androgyny fare better in the 1980s? Group & Organization Studies, 14(2), 216-233.
Powell, G., Butterfield, D., & Parent, J. (2002) Gender and Managerial Stereotypes: Have the Times Changed? Journal of Management, 28(2), 177-193.
24 Prasad, A., & Mir, R. (2002). Digging Deep for Meaning: A Critical Hermeneutic Analysis of
Shareholder Letters to Shareholders in the Oil Industry. Journal of Business Communication, 39, 92-116.
Raithel, S., Wilczynski, P., Schloderer, M., & Schwaiger, M. (2010). The value-relevance of corporate reputation during the financial crisis. Journal of Product & Brand Management, 19(6), 389-400.
Ranft, A., Zinko, R., Ferris, G., & Buckley, M. (2006). Marketing the image of management:
The costs and benefits of CEO reputation. Organizational Dynamics, 35(3), 279-290.
Rigg, C., & Sparrow, J. (1994). Gender, Diversity and Working Styles. Women in Mangement Review, 9(1), 9-16.
Rosette, A., & Tost, L. (2010). Agentic Women and Communal Leadership: How Role Prescriptions Confer Advantage to Top Women Leaders. Journal of Applied Psychology, 95(2), 221-235.
Ryan, M., & Haslam, S. (2005). The glass cliff: Evidence that women are over-represented in precarious leadership positions. British Journal of Management, 16, 81-90.
Ryan, M., Haslam, S., Hersby, M., & Bongiorna, R. (2011). Think Crisis – Think Female: the Glass Cliff and Contextual Variation in the Think Manager – Think Male Stereotype.
Journal of Applied Psychology, 96(3), 470-484.
Schein, V. (1973). The relationship between sex role stereotypes and requisite management characteristics. Journal of Applied Psychology, 57, 95-100.
Schein, V. (1975). The relationship between sex role stereotypes and requisite management characteristics among female managers. Journal of Applied Psychology, 60, 340-344.
Schein, V. (1978). Sex role stereotyping, ability and performance: Prior research and new directions. Personnel Psychology, 31, 259-268.
Schein, V., Mueller, R., & Jacobson, C. (1989). The relationship between sex role stereotypes and requisite management characteristics among college students. Sex Roles, 20, 103- 110.
25 Spence, J. (1985). Gender identity and its implications for concepts of masculinity and
femininity. In T. Sondregger (Ed.), Nebraska Symposium on Motivation (pp. 59-95).
Lincoln: University of Nebraska Press.
Spence, J., Helmreich, R., & Stapp, J. (1974). The personal attributes questionnaire: A measure of sex role stereotypes and masculinity femininity. JSAS Catalog of Selected Documents in Psychology, 4, 43-44.
Sterkel, K. (1986). The relationship between gender and writing style in business communications. The Journal of Business Communication, 25(4), 17-38.
Thompson, J. (1967). Organizations in action. New York, NY: McGraw-Hill.
Vugt, Van, M. (2011). The male warrior hypothesis. In J.P. Forgas, A.W. Kruglanski & K.D.
Williams (Eds.), The Psychology of Social Conflict and Aggression (pp. 233-244).
New York, NY: Psychology Press.
Vugt, Van, M., Cremer, De, D., & Janssen, D. (2007). Gender Differences in Cooperation and Competition. Association for Psychological Science, 18(1), 19-23.
Westphal, J., & Zajac, E. (1998). The Symbolic Management of Stockholders: Corporate Governance Reforms and Shareholder Reactions. Administrative Science Quarterly, 43, 127-153.
Williams, J., & Best, D. (1990). Measuring sex stereotypes: A multinational study (revised).
Newbury Park, CA: Sage.
Zuckerman, E.W. (2000). Focusing the corporate product: Securities analysis and de diversification. Administrative Science Quarterly, 45, 591-619.
26 APPENDIX A
Communal and Agentic Words in Dictionary
Agentic (1/2) Communal (1/2)
Achieve Active Adorable Affectionate
Advantage Adventurous Agreeable Aversion
Affiliation Against Caring Charming
Agress* Allow Cheer* Child*
Always Ambitious Colleagues Commit*
Analy* Assert* Communica* Community
Becoming Autonom* Company Compassion
Better Best Connect* Considerate
Business Boast* Cooperat* Council
Cause Cash Counsel Dear
Challeng* Certai* Debate Depend*
Compet* Chief Divine Disclos*
Continent Confiden* Discussion Do
Courag* Convinc* Emotiona* Empath*
Daring Crisis Engag* Everyone
Decision* Decide Family Feminine
Determin* Decisive Flatterable Frien*
Ear* Domina* Genius Gentle
Escape Efficienc* Group Help*
Executive Ever Her Honest
Force* Expec* Him Human
Great Gain Husband Interacti*
Hardworking Greedy Interdependen* Interpersona*
Hierarch* Headstrong Involv* Kids
Hostil* Hostess Kind Kindship
Imagin* I Lovely Loyal*
Improve Impulsive Meeting Member
Increased Income Message Mind
Individual* Independen* Modesty Nag
