25.06.2020 University of Amsterdam
Graduate School of Communication
Master’s Programme Communication Science Corporate Communication
Master’s Thesis
“Moms, wives, smart interns”:
How media coverage frames female CEO appointments and the consequence for the firm’s stock price
Jana Alisa Sailer
Supervisor Dr. Sandra Jacobs
Deadline: 26.06.2020
Student ID: 12490423
Abstract
Gender-biased news coverage of female CEOs contributes to a detrimental public perception of women in leadership positions, creating additional challenges women have to face when attempting to break the glass ceiling. To examine gender stereotypes and biased framing of female CEOs in
comparison to male CEOs this study analyzed 435 news articles about CEO appointments between 2006 and 2020 from 13 economic newspapers in the US using a quantitative content analysis. Additionally, the stock prices during the time of the CEO appointment were collected to assess a potential correlation between gender-biased reporting and the stock markets reaction to the
appointment. The results show that gender stereotypes are present in articles about female CEOs in the categories parenthood and intelligence, and mention the CEO’s gender disproportionately often. Furthermore, in comparison to male CEOs female CEOs were framed frequently with frames that support gender stereotypical assumptions, e.g. women being more emotional and caring. An
association between framed media coverage of CEOs and a change of the company’s stock price was not supported. The findings indicate that traditional gender role stereotypes are still a crucial part of media reporting about female CEOs, and continue to question the appropriateness of female
leadership.
“Moms, wives, smart interns”: How media coverage frames female CEO appointments and the consequence for the firm’s stock price
While women in management positions are still a rare phenomenon in general, they are even less present at upper management levels (Catalyst, 2019). Despite being equally well- educated and qualified (US department of Education), female managers stay far behind their male competitors in terms of numbers. From the Standard & Poor’s 500 Index only 27 out of the 500 companies currently have a female CEO, making up only 5 % of all public companies at the US stock market (Catalyze, 2020). Though with slight differences per country, those numbers are representative for the status quo of female CEOs all over the world. Despite a number of positive organizational outcomes of female leadership found by various
researchers (Chadwick, Dawson, 2018; Amore et al., 2014), there seem to be distinct challenges women face when attempting to ‘climb the ladder’.
One particular phenomenon women that achieved a position of visible power in society have to face is gender-biased media coverage. Regardless of the profession and industry, women that hold influential positions and great responsibility have been presented through the media in a different way than men. A prominent example are female politicians who oftentimes are described in terms of gender, physical appearance, and their
newsworthiness (Adcock, 2010), next to their role as mothers, wives, and mothers-to-be (Fernéndez, 2017; Garcia-Blanco & Wahl-Jorgensen, 2012). Aligned with those findings studies on media portrayal of female managers found reinforcing traditional gender values and roles (Tijani-Adenle, 2015).
While female politicians are more common now in the political landscape, little to no research has been done on how media treats female CEOs. The study on hand therefore focuses on media coverage of newly appointed female CEOs and analyzes which stereotypes (e.g. parenthood, appearance, intelligence), tone, and media frames are used to describe women and men in executive positions.
Media frames that portray women CEOs according to certain gender norms and stereotypes (e.g. collaborative, diplomatic, compassionate) could contribute to a biased perception of female CEOs in society and among the company’s shareholders, thus creating additional barriers women have to overcome when striving for a leadership position.
To inspect if differences in media coverage between female and male CEOs are related to actual changes of organizational outcomes, the companies’ stock price after appointing the CEO will be examined. Various studies analyzed how CEOs characteristics relate to the companies financial performance or stock market performance (e.g. Smith et al., 2010, Waldman et al., 2004). While some found a positive effect between CEO charisma and stock price, or CEOs celebrity level and firm performance (e.g. Waldman et al., 2010; Flynn & Staw, 2010), other studies found no or even a negative effects between CEO and financial measures (e.g. Tosi et al., 2004; Fannelli, 2009). Not only are those findings inconsistent, they also only include male CEOs and have been published more than ten years ago. Female CEOs remain an unusual research topic due to the fact that they are a rather new phenomenon. Past studies show that they experience a higher amount of media attention than male CEOs, and that this coverage is connected to stock market performances (Lee & James, 2007). This suggests that a ‘novelty’ effect of women CEOs communicated through the media could lead to negative consequences for the company’s financial performance.
Against this background the paper on hand aims to analyze how media coverage differs between male and female CEOs and how those differences relate to changes on the companies financial performance on the stock market. The guiding research question is “To
what extent does media coverage about newly appointed CEOs differ between sexes
appointment will be retrieved from online finance portals to calculate the degree of change for female and male CEOs associated with gender-biased frames.
Theoretical Background
Stereotypes in media coverage
Several studies have found the recurring presence of gender stereotypes in media coverage about women and men in power positions (Adcock, 2010; Garcia Blano & Wahl Jorgensen, 2012; Tijani-Adenle, 2015). Stereotypes highlight certain aspects of a social group and create expectations about the group based on those aspects (Ellemers, 2019). Stereotypes therefore contribute to social classification regarding certain traits and behavior norms, oftentimes increasing the perceived differences between social groups. Gender stereotypes tend to
describe women as characterized by a concern for communality, warmth, and care, while men are defined through greater agency, assertiveness, and performance (Ellemers, 2019).
Female and male stereotypes and behavior norms have been widely found in media depictions of politicians and managers (Kotzaivazoglou et al., 2018; Bligh et al., 2012).
Based on the stereotypical assumption that the female role in society is primarily the one of a mother (Odenweller et al., 2017, Koropeckyi-Cox et al., 2007), and given the high frequency of references to motherhood in news articles about female politicians (Fernéndez, 2017), I expect that media coverage about female CEO appointments will mention parenthood more often than in articles about male CEO appointments.
H1a: Media coverage about female CEO appointments will mention more often parenthood of the CEO than for male appointments.
Furthermore, gender stereotypes associate women with the private realm instead of the public realm, which connect women to being at home and taking care of household matters, not visible to the public eye (Van Der Pas & Aaldering, 2020; O’Neill et al., 2016). I expect
private life (childhood, relationship status, hobbies, etc.) than it will for male CEOs, since male stereotypes perceive men as more naturally belonging to a public sphere.
H1b: Media coverage about female CEO appointments will mention more often private life of the CEO than for male appointments.
A related phenomenon is the presentation of woman in appearance-based terms: Media coverage about women in public, like politicians, includes to a high extent comments about their physical appearance and perceived attractiveness, decreasing their chances of being voted (Women’s Media Center, 2014).
In addition to mentioning parenthood and private life, I therefore expect that articles about female CEO appointments will equally show a higher level of comments regarding physical appearance than they will for male appointments.
H1c: Media coverage about female CEO appointments will mention more often physical appearance of the CEO than for male appointments.
Finally, I assume articles about female CEO will emphasize more often explicitly the intelligence of the woman to stress her qualification for the job. Estimate of intelligence has been found in several global and intercultural studies to be impacted by gender (Storek & Furnham, 2013; Furnham & Syzmanowics, 2011; Stumm et al., 2009): According to the domain-masculine-intelligence theory DMIQ (Storek & Furnham, 2013),
mathematical/logical and spatial intelligence are associated with male-stereotypical traits, resulting in men rating themselves and other men higher on those types of intelligence. Experimental studies asking participants to estimate different types of intelligence of themselves and others, even for family members, repeatedly rated IQ of women lower than for men, while only emotional intelligence was rated higher in women (Furnham, & Storek,
more times explicitly that she is intelligent, to compensate for her gender and to justify the choice.
H1d: Media coverage about female CEO appointments will mention more often explicitly intelligence of the CEO than for male appointment
Frames
Another way to include gender stereotypes in media coverage is through the usage of different frames for men and woman. According to Goffman (1975) frames serve to create relevance, meaning and suggest behavior norms. For media texts this translates to depicting a certain aspect of perceived reality above others, to influence or promote a particular interpretation of a discourse, situation, or problem (Entman, 1993). By highlighting one narrative or
interpretation of a topic media impacts how the topic will be perceived and understood by society (De Vreese, 2005). Regarding CEO descriptions, prominent frames include titles like ‘the father’, ‘the preacher’, or ‘gardener’ and are oftentimes highly metaphorical (Hassain Shari, 2013). No frame research could be found that includes female CEOs, raising the question if women are portrayed with the same male archetypes.
On an individual level media frames have been found to affect the perceived
credibility and competence of professionals described in news articles (Martins et al., 2018), demonstrating that media frames can have a substantial effect on how the public receives and interprets information with immediate or even long-term consequences for individuals or population groups like women.
Hassain Shari (2013) created an overview of all existing CEO media frames and was able to divide the majority into the four prominent cluster from the Competing Values Framework (CVF): collaborate, create, compete, and control (Quinn & Cameron, 2006). The CVF was created by Quinn and Rohrbaugh originally in 1984 but redefined by Quinn and Cameron in 2006. It describes leadership orientations and management choices leaders face
Frank, 2008). Media frames tend to highlight one leadership aspect above others when reporting about CEOs.
Based on the four CVF categories, eight frames were identified (Hassain Shari, 2013) that summarize most of the existing CEO frames: coach, diplomat, visionary, innovator, commander, hero, constructor and expert.
Coaches and diplomats are described as being more collaborative, compassionate, and sensitive to employees needs, and match the assumptions of female gender stereotypes more than any other found CEO frame.
Commanders and heroes follow an aggressive, direct and power-demonstrating leadership style, oftentimes described through war and battle metaphors. Those two frames match closely male gender stereotypes of being assertive and in power.
Innovator and visionaries are focused on inventions and creations, and can be seen as the classical inventor or researcher CEO. Gender stereotypes are less prominent in this frame and do not match explicit male or female behavior stereotypes, therefore I define them as neutral frames and do not expect an association with gender of CEO.
Lastly, constructors and experts are described as more practical and knowledgeable, bringing a high level of experiences and expertise to the position.
I as well categorize this sector of the CVF cluster as a neutral one, which cannot clearly be prescribed to gender stereotypical assumptions.
Based on the categorization of coaches and diplomats as female gender frame, I expect both frames will be more present in articles describing the appointments of female CEOs than in articles describing the appointments of male CEOs.
Since male gender stereotypes are most prominent in the frames commander and hero, I expect, that male CEOs appointments will be described more often with those two frames, than female CEO appointments.
H2b: Media coverage about male CEO appointments will include more commander and hero frames, than it will for female CEO appointments.
Finally, the four neutral frames expert, constructor, visionary, and innovator should be equally distributed in the articles, regardless the gender of the appointed CEO.
H2c: The neutral CEO frames (constructor, expert, visionary, innovator) will be equally divided among both genders.
Gender emphasis
Media coverage about women in politics has been found to highlight their rarity and exceptionality, emphasizing indirectly how it is not common for the female gender to be an active part of the political realm (O’Neill et al., 2016). Since business equally is characterized with male stereotypes, I argue that in media coverage about female CEOs a higher level of gender emphasis will be present than in media coverage of male CEOs. I define gender emphasis as highlighting explicitly the gender of the CEO, which stresses the rarity of female leadership („She is the first woman to take this office“, „A working mum is the new leader of GM“).
H3: Media coverage about female CEO appointments will include more often gender emphasis than for male appointments
Tone
Since both, gender stereotypes and the low number of women as CEOs, conclude that leadership positions are not natural or obvious for women, I expect that the evaluation of female CEOs will be more negative than for male CEOs. Past research about female
male politicians (Miller et al., 2010), however other studies found contradicting results (Van Der Pas & Aaldering, 2020) and no gender bias in news tone.
A negative tone about female CEOs during the time of their appointment does not only consist of clearly negative wording; it is also present in remarks about how unusual or surprising the choice is. These expressions are not explicitly negative, but do contribute to create an atmosphere of insecurity and doubt. Given the fact that economic newspapers and especially CEO appointment will be most likely read by the company’s shareholders, investors, competitors, and employees this can translate to negative outcomes for the CEO and the company as well. In this sense I expect that a negative tone in news articles about female CEOs will be present more often than for male CEOs, given the rarity of the situation and the likelihood of the media to highlight this factor.
H4: Media coverage about female CEO appointments will have more often a negative tone regarding the CEO than media coverage about male CEO appointments.
Media effects on financial performance of the organization
Media coverage of CEOs influences public perception of leadership, stabilizes problematic gender stereotypes and establishes the appropriateness of female behavior.
Above these societal consequences, media coverage of CEOs also directly impacts not only the individual life of the CEO, but also the entire organization: Liua et al. (2017) found a clear economic link between the amount and type of media coverage and the CEOs level of human capital and opportunities after retirement; Bai et al. (2019) demonstrated that CEOs
engagement with social media directly relates to the companies performance and reputation; and Qiaoa et al. (2018) found evidence on CEOs media exposure having the potential to
outsider appointments caused a negative stock market reaction. Cook and Glass (2011) however showed that the gender-composition of the industry of the company affects the degree of change in the stock price. Women appointed as CEOs in female dominant industries (e.g. fashion, cosmetic) did not experience a decrease of stock price of their company.
Jadiappa et al. (2019) found that new female CEOs negatively predicted return of assets and equity caused by an increase in agency costs, and Metha et al. (2014) found that the stock price return in the first year of a new female CEO was significantly lower than for male CEOs, a difference that did not exist anymore the following year. Those findings indicate that women CEOs suffer a setback in their firm’s stock market value unrelated to their ability as a leader, but connected to how society and shareholders perceive women in leadership.
Public perception of female CEOs and the expectations of shareholders are
immediately related to how the media covers them. I expect that female gender stereotypical frames for female CEOs (coach and diplomat) will be associated with a greater decrease in the company’s stock price, in comparison to neutral (expert, constructor) or male stereotypical frames (commander, hero, visionary, innovator), because they emphasize the incompatibility of the female role stereotypes and the male-biased image of business leadership.
H5: Female-stereotypical frames for female CEOs will be associated with a greater decrease of the stock price than male stereotypical or neutral frames for female CEOs.
Method
To explore how female and male CEO appointments differ in their media coverage regarding stereotypes, tone, and frames a quantitative content analysis of economic newspaper articles from 13 different US newspapers was conducted. Unit of analysis were the individual articles per appointment. The sampling procedure consisted of three steps: 1. sampling companies with female and male CEO appointments, 2. sampling newspapers and articles, and 3. retrieving the stock market prices.
Sampling procedure: companies with CEO appointments
Multiple requirements applied for including CEO appointments in the sample: The appointment date had to be between 2006 and 2020, since former research did not extend the year 2006 (Lee & James, 2007; Cook & Glass, 2011).
The company that appoints the CEO had to be part of the Standard & Poors 500 Index at the time of the appointment. This guaranteed that, a) the companies were public companies listed on the stock market; b) the companies were big enterprises with a high net value, therefore receiving more media attention; c) the companies were well known and their leadership changes were most likely discussed in major economic newspapers; and d) the companies were all part of the US market, thus enabling an US-based analysis without cultural differences.
Female appointments: Companies that appointed female CEOs were extracted using the website Catalyst.org, through the list „Women CEOs of the S&P 500“, which is regularly updated (last update May 2020). Since in total during the years 2006 until 2020 only 47 female CEOs were appointed in S&P500 companies, all 47 companies were included in the sample. The sample was therefore a non-probability, census sample including all elements of the specific population.
Male appointments: To extract companies with male CEO appointments I used a non-probability, purposive sampling method to create a representative sample with the same characteristics as the female sample. I manually researched 47 companies that were part of the S&P 500 Index since 2006, using the Wharton Research Data Services, followed by web searches to assess if they had a male CEO appointment between 2006 until 2020. The exact CEO appointment day in all companies was manually researched, as well as the companies
companies and companies with missing data, the final sample consisted of 94companies. See appendix A for a table of all included companies, CEOs names, and industries
Sampling procedure: Newspapers
I used a purposive, non-probability sample and only included articles from major economic newspapers, guaranteeing a high circulation and popularity, therefore increasing the probability to impact public perception of the CEO choice. The following newspaper were included based on these criteria: The New York Times, The Economist, The Fortune
Magazine, The Wall Street Journal, The Forbes Magazine, The Associated Press (online and international), The Business Wire, The Financial Wire, and The Financial Times. The articles were downloaded through the research data base Nexis Uni or academic subscription
accounts. A table showing the number of articles included per newspaper can be found in appendix B.
Sampling procedure: Articles
Newspaper articles that were included in the content analysis were selected based on three criteria: 1. The article had to be published in a time frame of maximum three days after the CEO appointment, and not earlier than the exact appointment day. 2. All articles had to indicate in their title or subtitle that a CEO change was being discussed. 3. The article had to be published in one of the selected newspapers.
Search terms included the company name, or the CEO name. Using this approach in total 502 articles were collected for the manual content analysis, after excluding duplicates 460 articles remained. After the coding process the data was inspected and irrelevant cases were filtered out of the data set, resulting in 435 cases. When necessary, variables were recoded into numerical variables. Around 2% of cases were missing depending on the variable, caused by interrupted or inconsistent coding. Of the 435 included articles in the sample 204 articles discussed male CEO appointments and 231 female ones.
Sampling procedure: Stock market data
To analyze a potential correlation between media coverage of CEO appointments and stock market changes, the stock price values of the companies three days before the appointment day and three days after the appointment day were manually extracted to compare an
averaged value of before and after, which is more precise and allows to some extent to control for other influencing factors. The historical stock price values for the in total 6 days per appointment (3 days before, 3 days after) were obtained using the history section of the website Yahoo! Finance. It was controlled that no company announced the CEO change after closing hours of the stock exchange, which would alter the time frame. See appendix C for a table of all individual stock prices.
Operationalization of Variables
Stereotypes
Parenthood: Parenthood was measured as article includes comments about the parent status of the CEO (e.g. “mom of two”, “busy dad”).
Private Life: The category private life in the codebook is defined as article mentions the aspects of the CEO’s life unrelated to work, such as relationship status (past and present), family matters (excluding parenthood), hobbies, place of birth, personal aspirations, and others. Comments about the CEO’s age were explicitly not included under private life. Appearance: To measure the stereotype of women being presented majorly through their physical appearance, the coder had to indicate if the article mentions the appearance of the CEO, which was defined as comments about the face, body, clothes, accessories, or hair of the CEOs.
Gender emphasis: Gender emphasis was defined as the article mentioning the gender of the CEO in explicit words or close synonyms, e.g. first female CEOs, lady in charge, female boss, a though guy, a real man. To assess the level of gender emphasis per article the amount of gender references had to be counted manually. Gender emphasis could only be counted once per sentence and did not occur for titles such as Mr, Ms, or Mrs. Titles and subtitles were included, as well as quotes of the CEO or others.
Tone: To assess the tone of the article regarding the CEO, clearly positive or negative (including surprised) words had to be present in the article regarding the CEO or the CEO choice, but not the company. If negative consequences following the CEO choice were described the tone had to be coded as negative. If the tone was unclear or missing, no evaluation had to be coded.
Frames
The eight frames originate from the research of Hassain Shari (2013) and were further
operationalized in clearly distinguishable categories. All frames include a detailed description with examples in the codebook, that followed a brief first assessment based on the most prominent feature of each frame group: Relationships (coach, diplomat),
innovations/strategies (visionary, innovator), power/challenges (commander, hero), and knowledge/experiences (constructor, expert). A second level for internal differences in each group was included (e.g. Group Relationships: motivating= coach, compassionate= diplomat). The detailed descriptions of all variables can be found in the codebook under appendix F.
Degree of stock price change
The average stock price for the three days prior to the CEO appointment were calculated by adding the three values of these days together and dividing the result by 3. The same
calculation was used for the average of the stock prices the three days after the appointment (including the announcement day). For better comparison the degree of change between both
means was calculated as percentages: The average stock price before the appointment was subtracted from the average stock price after the appointment, and the result divided by the average stock price before the appointment (e.g. [24.89-22.91]/22.91=8.6%). Appendix B includes a table with all individual stock prices and means.
Intercoder Reliability
To test intercoder reliability a random sample of 10 % of all articles (43 articles in total) was coded by an independent coder, who received 45 minutes of coding training before the coding process to become familiar with the different categories. Especially the decision criteria for tone, gender emphasis, and frames were practiced given their level of subjectivity. Since percentage of agreement cannot control for agreement by chance, observed disagreement, and expected disagreement the two additional indices Cohen’s kappa and Krippendorff’s alpha were included due to their more conservative nature (Lombard et al., 2002).
Krippendorff’s alpha has a threshold of .67 to consider a variable reliable; if percentage of agreement was higher than 90% variables below .67 were still included in the analysis. After calculating all reliability values, no variables were excluded from the analysis (a table including all values can be found in appendix D). Table 1 shows the Krippendorff’s alpha values for all content variables:
Results
The purpose of the following statistical analyses was to examine gender stereotypes and CEO frames in news articles and if their frequency differs between genders.
Table 2 includes an overview of the frequencies of all categorical variables in the sample, while Table 3 includes all numerical variables.
Stereotypes: Parenthood
There was a significant association between gender of CEO and article mentions parenthood of CEO, X²(1, N=428)=14.73, p>.001, Tau=.034.
If the CEO was a woman the probability that the article mentioned her status as a mother was statistically higher: The strength of association was moderate, the prediction of article
mentions parenthood of CEO improved 34 % for female CEOs. Hypothesis 1a „Media coverage about female CEO appointments will mention more often parenthood than for male appointments“ was supported. Table 4 summarizes the statistical results of the Chi square tests that were conducted for appearance, parenthood, private life, and intelligence.
Stereotypes: Private life
As can be seen in Table 4 there was no statistically significant difference between genders regarding the frequency that articles mentioned their private life. Hypothesis 1b „Media coverage about female CEO appointments will mention more often private life than for male appointments“ was not supported by the findings.
Stereotypes: Physical appearance
The articles did not mention physical appearance statistically significant more often for female CEOs, X²(1, N=426)=2.34, p=.180, Tau=.005. Hypothesis 1c „Media coverage about female CEO appointments will mention more often physical appearance than for male appointments“ was not supported.
Stereotypes: Intelligence
Articles mentioned intelligence statistically significant more when describing female CEOs,
X²(1, N=428)=4.61, p=.036, Tau=.011. The association was weak, prediction of article
mentions intelligence improved 11% when taking gender of the CEO into account.
Hypothesis 1d: “Media coverage about female CEO appointments will mentions more often intelligence than for male appointments” was supported.
Frames
A set of chi square tests was conducted to compare gender-biased and neutral CEO frames between the genders. A summary of the statistical results can be seen in Table 5.
Coach and Diplomat
The association between the presence of frame coach and gender of CEO was significant but weak, X²(1, N=426)=6.10, p=.015, Tau=0.14. Gender of CEO improved the presence of the frame coach by 14 %. Women were more likely to be framed with the frame coach than men. The association between the presence of the diplomat frame and gender of CEO was as well significant, X²(1, N=428)=4.55, p=.044, Tau=.011. The association was weak, prediction of presence of diplomat frame improves by 11 % when the gender of the CEO was taken into account. Hypothesis 2a „Media coverage about female CEO appointments will include more coach and diplomat frames than it will for male CEO appointments“ was supported.
Commander and Hero
Women CEOs had a statistically higher amount of commander frames in the sample compared to men, X²(1, N=428)=4.70, p=.032, Tau=.011. Prediction of commander frame improved 11% when taking female gender into account. For the hero frame the association was not significant, X²(1, N=428)=2,37, p=.187, Tau=.006. Neither men nor women were
coverage about female CEO appointments included more male gender stereotypical frames than it did for men.
Constructor, expert, visionary, and innovator
There was no difference among the genders regarding the frequency of being framed as
constructors, X²(1, N=427)=2.32, p=0.144, Tau=.005, or experts, X²(1, N=427)=1.28, p=.278, Tau=.003. Both genders were framed as constructors or experts equally.
No gender was more likely to be framed as a visionary: X²(1, N=426)= 2.10, p=0.157,
Tau=.005 or innovator, X²(1, N=427)=2.63, p=.122, Tau=.006. Both genders were equally
likely to be framed as visionaries or innovators. Hypothesis 2c: “The neutral CEO frames (constructor, expert, visionary, innovator) will be equally divided among both genders” was therefore supported.
Gender emphasis
All requirements to conduct an independent samples t-test using the t-distribution were met. The independent samples t-test for male and female gender of CEO and count of gender
emphasis was significant, t(225.62)=-8.474, p >.001, 95% CI [-1.97, -1.23].
Male CEO gender had an average count of gender emphasis per article of M=.01, SD=.01, while female CEO gender had an average count of gender emphasis per article of M=1.61,
SD=2.83. The articles on average mention 1.61 times explicitly the gender of the CEO for
women, and only .01 for men. The effect was weak to moderate, d=.43. Hypothesis 3 „Media coverage about female CEO appointments will include a higher number of gender emphasis than for male appointments“ was supported by the findings.
Tone
The association between gender of the CEO and tone was not significant, X²(2, N =
428)=3.17, p=.205, Tau=.004. The articles did not differ regarding the tone to describing the CEO for the different genders. Women were not portrayed more negatively in terms of tone
than men. Hypothesis 4 „Media coverage about female CEO appointments will have more often a negative tone than media coverage about male CEO appointments“ was not supported.
Stock price changes
To analyze the association between media frames and degree of change in stock price after the appointment I followed a four step process: 1. Manually calculating the degree of change as percentage for each individual company, 2. Manually calculating the mean degree of change in stock price per gender, 3. Conducting a simple, linear regression with gender as independent variable and stock price change as dependent variable to see if the difference between genders and stock price change is statistically significant. 4. Finally, conducting a multiple linear regression with gender and the media frames as independent variables and
stock price change as dependent variables to assess if the degree of change in stock price is
associated with media frames. Descriptions and results of each stage can be found below. It has to be noted that the mere stock price of a company is not an appropriate indicator of the firm’s financial value: Since companies can choose freely how many shares they want to make available for trading (capitalization) they can influence how high or low their stock price is. Therefore it was essential to choose degree of change of stock price as the dependent variable.
Step 1: Calculating degree of change in percentage for each company
To account for the degree of change between the different moments of measuring the stock price, I manually calculated the difference as percentages using the formula (T2-T1)/T1, e.g. for Juniper Networks: (24.89-22.91)/22.91=8.6% increase after appointment. I repeated this calculation for all companies in the sample (see a table for all stock price values and means in
Step 2: Calculating degree of change in percentage per gender
To see how high for each gender the degree of change in stock price in general was, I added all percentages and divided the results by the number of items. I repeated this for each gender, resulting in a value of -0.32% for companies that appointed a male CEO and -0.43% for companies that appointed a female CEO. The result showed that on average all companies in the sample experienced a decrease in stock price after the leadership change, but the loss was higher when the newly appointed CEO was a woman.
Step 3: Linear regression for stock price change per gender
To further analyze if the difference in loss in stock price between men and women after the appointment is statistically significant and a consequence of stereotypical framing of media coverage a linear regression analysis was conducted.
The assumptions for conducting a linear regression analysis were sufficiently met: The
residuals were sufficiently normally distributed and a lack of homoscedasticity was not found. The first linear regression model, including female CEO gender as independent variable and
degree of stock price change as dependent variable, was not significant F(1, 411)=2.39, p=.123.The difference between the decrease in stock price for both genders was not high enough to be statistically significant.
Step 4: Multiple linear regression with media frames, gender and stock price change A second multiple regression model including female CEO gender, and the eight frames as independent variables, and degree of stock price change after appointment (in percentage) as dependent variable was equally not significant F(9, 411)=1.52, p=.138. Neither gender nor media frames did predict a change in stock price after the appointment of a new CEO. Hypothesis 5 „Female-stereotypical frames will be associated with a greater decrease of the stock price than male stereotypical or neutral frames“ was not supported (see appendix E for a table summarizing the Regression analysis outcome).
Conclusion
The presented findings answer the research question “To what extent does media coverage
about newly appointed CEOs differ between sexes regarding the presence of stereotypical assumptions, and how are these differences related to changes in the companies’ stock price?” as follows: Articles about female CEO appointments mention disproportionately
often their children, their intelligence, and their gender in comparison to their male
colleagues. Women were also more likely to be portrayed with frames that have a high level of female stereotypes (caring, emotional, focused on relationships). The study provides evidence that gender-bias does exist in media coverage but only for female CEOs:
No association could be found between male CEOs and a higher frequency of stereotypes or frames (commander, hero). On the contrary, women were more framed as commanders than men, which accentuats combative, aggressive leadership behavior of women. Tone of article and CEO gender was not associated, neither was appearance or private life (excluding parenthood). Regarding the second part of the research question, no correlation between gender, frames, and stock price change of the company after the appointment could be found. Neither gender nor gender-biased frames did predict degree of stock price change.
Discussion
Several of the findings of the current paper are aligned with those of previous studies discussing leadership and gender-bias.
The high frequency of comments about the female CEOs children, gender, and intelligence is consistent with existing research about female politicians (Adcock, 2010; Bligh et al., 2012) and managers (Tijani-Adenle, 2015; Braun et al., 2017), and questions the appropriateness of
notion, gender-biased frames (diplomat, coach) create the image of a female CEO as focused on relationships and emotions before mentioning competency, experience, or qualification. This indirectly benefits men as CEOs, since male-stereotypical traits are still associated by many with how a leader should behave or look like (Stoker et al., 2012; Fischbach et al., 2015).
Braun et al. (2017) found that female trait stereotypes not only hinder women from achieving leadership positions, but also push women towards followership positions, which are associated with communality. The study on hand found that media coverage of CEOs is confirming the association between women and communality by framing women as focused on relationships and compassion. Emphasizing feminine trait stereotypes when reporting about female CEOs is therefore most likely to have a detrimental effect on how the public perceives female leaders in business, due to both a “think manager think male” and a “think follower think female” effect. Therefore women have to face two types of gender-biased assumptions when aspiring to become corporate leaders: The prototypical male perception of leadership and the ascribed stereotypes of the female gender role as incompatible with this image of leadership.
Furthermore, women were more likely to be portrayed with the commander frame, a leadership style defined through competition, battle, power, and toughness. This finding is aligned with past research on female politicians who were portrayed with an overemphasis on combative, aggressive behavior (Van Der Pas & Aaldering, 2020). The findings suggest there are only two possible types of frames for female CEOs: either as caring mothers-in-chief (coach, diplomat) or as aggressive commanders. Interestingly, from the three frames that were associated with women, the commander frame is the only one not including followership traits such as communality, therefore not indirectly questioning the legitimacy of female leadership. Indeed, some research found that the emotion anger is associated with managers
(Fischbach 2015), and a more aggressive, uncooperative behavior is associated with leadership, instead of followership (Braun 2017).
While those findings suggest women being framed by the media as commanders might be positive because it narrows the gap between the stereotypical depictions of leaders and women, other research found problematic outcomes for ‘angry women’ at the workplace: A study of Salerno and Peter-Hagene (2015) found that showing anger as a woman decreases their social influence and makes others less likely to consider their opinion. And Brescoll and Uhlmann (2008) found that regardless of their position (trainee versus CEO) women who expressed anger at the workplace were conferred lower status and their anger was attributed to internal characteristics (personality, loss of control) rather than external circumstances.
Therefore all three frames associated with female CEOs are more likely to be perceived as negative by the public since they either emphasize female stereotypes that are associated with followership, or show women as angry and aggressive which can lower their social influence and status.
The current research was not able to establish support for a connection between media frames, gender, and stock price reaction. This contradicts the original findings of Lee and James (2007) that found a decrease of stock price after the appointment of a female CEO, as well as research that established a connection between media variables, and the firm’s financial performance (Liua et al., 2017; Bai et al., 2019). It was not possible to isolate a specific media frame that accounts for the change of the company’s stock price after
appointing a CEO. This might be partly due to the inconsistency among the articles covering the same CEO regarding frames. Articles varied widely in the usage of frames for the same CEO, even when they were published in the same newspaper. This resulted in the same CEO
they read. Regardless of the inconsistnecy of frames covering the same CEO, the association between female leaders and gender-stereotypical assumptions throughout the articles
undermines female leadership and slows down the rise of female executives towards equal career opportunities and social status.
Limitations
The study on hand has several limitations that will be discussed in the following section. First, the choice of articles in the sample could affect the validity of the research. Even though only articles of economic newspaper where included, there is variety in the style of articles. Some are brief announcements with standardized wording, while other include the biography of the CEO, and a more colorful description of the person. Therefore framing, tone, and manifest variables at times varied to great extent between the different publications. Nevertheless I consider it a valid choice to include all types of articles and not focus on one specific genre. The public does not only choose standard announcements but also more subjective articles to form an opinion of the new CEO.
Second, the category ‘tone’ in the codebook only included three options (positive, negative, no evaluation), which did not leave any space for indicating a mixed tone. It could have been interesting to see if one gender gets described more often in a mixed tone than the other, since a not-clearly-positive evaluation could create a negative perception of the CEO as well.
Third, regarding the assessment of the stock price change, there are several limiting factors. I used a time period of each three days before and three days after the appointment to calculate both averages and degree of change to see a potential difference. Three days were chosen because the amount of articles published about an appointment decreases rapidly after three days. However, a longer period of time to see the stock price change could have been of value, because more influencing factors could have been controlled for. It is also questionable
if the news of a CEO change continues to impact a company’s stock price beyond the first day, which would explain the absence of significant correlations as well.
Fourth, as mentioned before in the discussion, one major limitation that explains the insignificance of the regression model is the lack of consistency of frames per appointment. Most appointments were described with different frames in each article, so no single frame could be isolated regarding its association with the company’s stock price.
Equally a number organizational factors is known to affect the firm’s financial performance on the stock market that were not considered due to practical limitations: The size of the company, the financial performance the year before, and the gender composition of the industry.
Finally, it is important to note, that the sample is a very specific one: I intentionally focused on S&P500 firms because their CEO appointments receive a high public attention and the number of female CEOs stayed remarkably low throughout the years. Thus the findings might not be generalizable for low to mid market companies, non-US-based companies, private or family companies.
Future Research
Future research could further analyze some of the categories in the codebook by including more subtle hints of gender bias, e.g. including a category to assess if the article features opinions of stock market analyst about the appointment. In multiple articles stock market analysts were asked if the CEO choice is good or bad, which can lead to a higher level of insecurity in shareholders who read professional opinions about the new CEO.
CEOs are rare, the CEO position oftentimes is the first executive position for the woman, making her more prone for negative comments about her lack of experience.
Additionally, future research could analyze how often media covers the professional ambitions of CEOs. Several articles discussed how the CEOs were outspoken about their career ambitions or switched companies when they were denied the leadership position in favor of someone else. It would be interesting to examine if the frequency regarding
comments about calculated professional decisions of CEOs is the same for men and women. While for men career ambition is considered to be an essential part of work life, (Harman & Sealy, 2017), it is contradicting the societal gender role of women as mothers and wives, therefore it could be perceived negative if women show strategic thinking regarding their future career.
Finally it is essential that future studies focus on the underlying mechanisms of gender-bias in media coverage about women. This can be researched in several ways: Examining gender-biased assumptions about leadership of journalists on an individual level, or analyzing the economic news industry as a whole to assess gender bias on organizational level (e.g. numbers of female writers, editors, amount of articles featuring men or women, gender-composition of readers). As long as gender stereotypes continue to dominate the media coverage of female leadership and either idealize motherly traits or exaggerate female aggression, the 5% threshold of female CEOs in the S&P500 Index will remain stable for many years to come.
References
Adcock, C. (2010). The Politician, The Wife, The Citizen, and her Newspaper: Rethinking women, democracy, and media(ted) representation. Feminist Media Studies, 10(2), 135–159. https://doi.org/10.1080/14680771003672254
Amore, M., Garofalo, O., & Minichilli, A. (2014). Gender Interactions Within the Family Firm. Management Science, 60(5), 1083–1097.
https://doi.org/10.1287/mnsc.2013.1824
Atwater, L., Tringale, A., Sturm, R., Taylor, S., & Braddy, P. (2018). Looking Ahead: How What We Know About Sexual Harassment Now Informs Us of the Future.
Organizational Dynamics. https://doi.org/10.1016/j.orgdyn.2018.08.008 Bai, L., Yan, X., & Yu, G. (2019). Impact of CEO media appearance on corporate
performance in social media. North American Journal of Economics and Finance, 50. https://doi.org/10.1016/j.najef.2019.100996
Belasen, A., & Frank, N. (2008). Competing values leadership: quadrant roles and personality traits. Leadership & Organization Development Journal, 29(2), 127–143.
https://doi.org/10.1108/01437730810852489
Bligh, M., Schlehofer, M., Casad, B., & Gaffney, A. (2012). Competent Enough, But Would You for Her? Gender Stereotypes and Media Influences on Perceptions of Women Politicians. Journal of Applied Social Psychology, 42(3), 560–597.
https://doi.org/10.1111/j.1559-1816.2011.00781.x
Brescoll, V., & Uhlmann, E. (2008). Can an Angry Woman Get Ahead?: Status Conferral, Gender, and Expression of Emotion in the Workplace. Psychological Science, 19(3),
manager-think male, think follower-think female: Gender bias in implicit followership theories.(Report). Journal of Applied Social Psychology, 47(7), 377–388.
https://doi.org/10.1111/jasp.12445
Cameron, K. S. (2014). Competing values leadership. Cheltenham, UK: Edward Elgar. Catalyst (2019): Women in Management. Quick take. Retrieved from:
https://www.catalyst.org/research/women-in-management/ Catalyst (2020): List: Women CEOs of the S&P 500. Retrieved from
https://www.catalyst.org/research/women-ceos-of-the-sp-500/
Chadwick, I., & Dawson, A. (2018). Women leaders and firm performance in family
businesses: An examination of financial and nonfinancial outcomes. Journal of Family Business Strategy, 9(4), 238–249. https://doi.org/10.1016/j.jfbs.2018.10.002 Cook, A., & Glass, C. (2011). Leadership change and shareholder value: How markets react
to the appointments of women. Human Resource Management, 50(4), 501–519. https://doi.org/10.1002/hrm.2043
De Vreese, C. H. (2005). News framing: Theory and typology. Information Design Journal & Document Design, 13(1), 51–62. https://doi.org/10.1075/idjdd.13.1.06vre
Entman, R.M. (1993), Framing: Toward Clarification of a Fractured Paradigm. Journal of Communication, 43: 51-58. doi:10.1111/j.1460-2466.1993.tb01304.x
Ellemers, N. (2019). Gender Stereotypes. Annual Review of Psychology, 69(1), 275–298. https://doi.org/10.1146/annurev-psych-122216-011719
Fanelli, A., Misangyi, V., Tosi, H. (2009). In Charisma We Trust: The Effects of CEO Charismatic Visions on Securities Analysts. (2009). Organization Science, 20(6), 1011–1033. https://doi.org/10.1287/orsc.1080.0407
comparative analysis of the media representation of the first female Defense minister in Spanish history. Communication Papers, 4(07).
https://doi.org/10.33115/udg_bib/cp.v4i07.22047
Fischbach, A., Lichtenthaler, P., & Horstmann, N. (2015). Leadership and Gender Stereotyping of Emotions. Journal of Personnel Psychology, 14(3), 153–162. https://doi.org/10.1027/1866-5888/a000136
Flynn, F., & Staw, B. (2004). Lend me your wallets: the effect of charismatic leadership on external support for an organization. Strategic Management Journal, 25(4), 309–330. https://doi.org/10.1002/smj.377
Gaughan, K., & Smith, E. (2016). Better in the Shadows? Media Coverage and Market Reactions to Female CEO Appointments. Academy of Management Proceedings, 2 016(1), 1069–1074. https://doi.org/10.5465/AMBPP.2016.301
Garcia-Blanco, I., & Wahl-Jorgensen, K. (2012). The Discursive Construction of Women Politicians in the European Press. Feminist Media Studies, 12(3), 422–441.
https://doi.org/10.1080/14680777.2011.615636
Goffman, E. (1975). Frame analysis: an essay on the organization of experience. Northeastern University Press. Retrieved from:
https://is.muni.cz/el/1423/podzim2013/SOC571E/um/E.Goffman-FrameAnalysis.pdf Hayes, D., Lawless, J., & Baitinger, G. (2014). Who Cares What They Wear? Media, Gender,
and the Influence of Candidate Appearance.(Report)(Author abstract). Social Science
Quarterly, 95(5), 1194–1212. https://doi.org/10.1111/ssqu.12113
Harman, C., & Sealy, R. (2017). Opt-in or opt-out: exploring how women construe their ambition at early career stages. Career Development International, 22(4), 372–398.
Jadiyappa, N., Jyothi, P., Sireesha, B., & Hickman, L. (2019). CEO gender, firm performance and agency costs: evidence from India. Journal of Economic Studies, 46(2), 482–495. https://doi.org/10.1108/JES-08-2017-0238 Koropeckyj-Cox, T., Romano, V., & Moras, A. (2007). Through the Lenses of Gender, Race, and Class: Students’ Perceptions of Childless/Childfree Individuals and Couples. Sex Roles, 56(7-8), 415–428. https://doi.org/10.1007/s11199-006-9172-2 Kotzaivazoglou, I., Hatzithomas, L., & Tsichla, E. (2018). Gender stereotypes in advertisements for male politicians: longitudinal evidence from Greece. International Review on Public and Nonprofit Marketing, 15(3), 333–352. https://doi.org/10.1007/s12208-018-0202-x
Lee, P., & James, E. (2007). She’‐e‐os: gender effects and investor reactions to the
announcements of top executive appointments. Strategic Management Journal, 28(3), 227– 241. https://doi.org/10.1002/smj.575
Liua, B., Mcconnell, J., & Xu, W. (2017). The power of the pen reconsidered: The media, CEO human capital, and corporate governance. Journal of Banking and Finance, 76, 175–188. https://doi.org/10.1016/j.jbankfin.2016.12.004
Lombard, M., Snyder‐Duch, J., & Bracken, C. (2002). Content Analysis in Mass Communication: Assessment and Reporting of Intercoder Reliability. Human
Communication Research, 28(4), 587–604. https://doi.org/10.1111/j.1468-
2958.2002.tb00826.x
Martins, N., Weaver, A., & Lynch, T. (2018). What the Public “Knows” About Media Effects Research: The Influence of News Frames on Perceived Credibility and Belief
Change. Journal of Communication, 68(1), 98–119. https://doi.org/10.1093/joc/jqx004 Mehta, M., Neeraja, B., & Chandani, A. (2014). Women CEOs and financial performance of
Miller, M., Peake, J., & Boulton, B. (2010). Testing the Saturday Night Live Hypothesis: Fairness and Bias in Newspaper Coverage of Hillary Clinton’s Presidential Campaign. Politics & Gender, 6(2), 169–198.
https://doi.org/10.1017/S1743923X10000036
Women’s Media Centre (2013): Name it. Change it. An Examinanation of the Impact of Media Coverage of Women Candidates' Appearance.
http://www.nameitchangeit.org/page/-/Name-It-Change-It-Appearance-Research.pdf Neto, F., & Furnham, A. (2011). Sex differences in parents’ estimations of their own and their
children’s multiple intelligences: a Portuguese replication.(Report). Spanish Journal of
Psychology, 14(1), 99–110. https://doi.org/10.5209/rev_SJOP.2011.v14.n1.8
Odenweller, K., & Rittenour, C. (2017). Stereotypes of Stay-at-Home and Working Mothers. Southern Communication Journal, 82(2), 57–72.
https://doi.org/10.1080/1041794X.2017.1287214
O’Neill, D., Savigny, H., & Cann, V. (2016). Women politicians in the UK press: not seen and not heard? Feminist Media Studies, 16(2), 293–307.
https://doi.org/10.1080/14680777.2015.1092458
Petrides, K., Furnham, A., & Martin, G. (2004). Estimates of Emotional and Psychometric Intelligence: Evidence for Gender-Based Stereotypes. The Journal of Social
Psychology, 144(2), 149–162. https://doi.org/10.3200/SOCP.144.2.149-162
Qiao, P., Fung, H., & Wang, W. (2018). Chinese firms’ crash risk and CEO media exposure. Applied Economics Letters, 25(14), 1034–1037.
https://doi.org/10.1080/13504851.2017.1420871
Relationship Between Sexualized Appearance and Perceptions of Women’s Competence and Electability. Sex Roles, 79(11-12), 671–682.
https://doi.org/10.1007/s11199-018-0898-4
Smith, K., Smith, M., & Wang, K. (2010). Does brand management of corporate reputation translate into higher market value? Journal of Strategic Marketing, 18(3), 201–221. https://doi.org/10.1080/09652540903537030
Stoker, J., Velde, M., & Lammers, J. (2012). Factors Relating to Managerial Stereotypes: The Role of Gender of the Employee and the Manager and Management Gender Ratio. Journal of Business and Psychology, 27(1), 31–42.
https://doi.org/10.1007/s10869-011-9210-0
Tijani-Adenle, G. (2016). She’s homely, beautiful and then, hardworking. Gender in Management: An International Journal, 31(5/6), 396–410.
https://doi.org/10.1108/GM-06-2015-0053
Tosi, H., Misangyi, V., Fanelli, A., Waldman, D., & Yammarino, F. (2004). CEO charisma, compensation, and firm performance. The Leadership Quarterly, 15(3), 405–420. https://doi.org/10.1016/j.leaqua.2004.02.010
US department of Education (2016): Total undergraduate fall enrollment in degree-granting postsecondary institutions, by attendance status, sex of student, and control and level of institution: Selected years, 1970 through 2026. Retrieved from:
https://nces.ed.gov/programs/digest/d16/tables/dt16_303.70.asp
Van Der Pas, D., Aaldering, L., & Political Science and Public Administration. (2020). Gender Differences in Political Media Coverage: a Meta-Analysis. Journal of
Communication, 70(1), urn:issn:0021–9916.
Leadership Matter? CEO Leadership Attributes and Profitability under the Conditions of Perceived Environmental Uncertainty. Academy of Management Journal, 14(1), 134-144. https://doi.org/10.2307/3069341
Waldman, D., Javidan, M., & Varella, P. (2004). Charismatic leadership at the strategic level: A new application of upper echelons theory. The Leadership Quarterly, 15(3), 355– 380. https://doi.org/10.1016/j.leaqua.2004.02.013
Appendix A Table
Sample characteristics: Companies, CEOs, Industries
Female CEOs
Company CEO Industry
Wendys Kerrii Anderson Restaurants
ADM Patricia Woertz Data processing & outsourced services Mondelez Irene Rosenfeld Packaged Foods & Meats
Pepsico Indra K. Nooyi Soft Drinks
Wellpoint Angela Braly Managed Health Care
Xerox Ursula Burns Tech. hardware, storage & peripherals
Keybank Beth Mooney Regional banks
Campbells Denise Morrison Packaged Foods & Meats
Tegna Gracia Martore Broadcasting
IBM Ginni Rometty IT Consulting & other services Mylan Heather Bresch Pharmaceuticals
Alliant Energy Patricia Kampling Electric Utilities Avon Products Sherilyn McCoy Personal Products
Yahoo Marissa Mayer Internet Content and Information General Dynamics Phebe Novakovic Aerospace & Defense
Lockheed Martin Marillyn Hewson Aerospace & Defense Duke Energy Lynn Good Electric Utilities Ulta Beauty Mary Dillon Specialty Stores HCP Lauralee Martin Healthcare Facilities General Motors Mary Barra Automobile Manufacturers American Water Susan Story Water Utilities
Oracle Safra Catz Application Software
AMD Lisa Su Semiconductors
Occ. Petroleum Vicki Hollub Oil & Gas Exploration & Prod. Staples Shira Goodman Specialty Retail Discretionary Progressive Patricia Griffith Property & Casualty Insurance
Company CEO Industry Nasdaq Adena Friedman Financial Exchanges & Data Mattel Margo Georgiadis Leisure
Hersheys Michele Buck Packaged Foods & Meats PG&E Geisha Williams Utilities Regulated Electric Signet Jewelers Virgina Coleman Luxury Goods
Kohls Michelle Gass General Merchandise Stores Northrop Grum. Kathy Warden Aerospace & Defense Celanese Lorri Ryerkerk Specialty Chemicals
BestBuy Corrie Barry Computer & Electronics Retail Accenture Julie Sweet IT Consulting & other services
Gap Sonia Syngal Apparel Retail
Zoetis Kristin Peck Pharmaceuticals
DuPont Ellen Kullman Specialty Chemicals TJX Companies Carol Meyrowitz Apparel Retail
Sunoco Lynn Elsenhans Oil & Gas Refining & Marketing Sempra Energy Debra Reed-Klages Multi-Utilities
Xylem Gretchen McCain Industrial Machinery Ross Stores Barbara Rentler Apparel Retail
Arista Jayshree Ullal Comm. Equipment
Progressive Susan Griffith Property & Casualty Insurance Avon Prod. Angela Cretu Personal Products
Male CEOs
Company CEO Industry
Adobe Shantanu Narayen Application Software
Company CEO Industry
Mastercard Ajaypal Banga Data processing & outsourced services Juniper Networks Kevin Johnson Communications Equipment
HP Enrique Lores Tech. Hardware, Storage & Peipherals Hormel Foods Jim Snee Packaged Foods & Meats
Home Depot Frank Blake Home Improvment Retail Hasbro Brian Goldner Leisure Products
Gap Arthur Peck Apparel Retail
Vulcan Materials Thomas Hill Construction Materials Walt Disney Bob Chapek Movies & Entertainment Walmart Doug McMillon Hypermarkets & Super Centers Estée Lauder Fabrizio Freda Personal Products
Flowserve Mark Blinn Industrial Chemicals Starbucks Kevin Johnson Restaurants
Textron Scott Donnelly Aerospace & Defense Walgreens Boots Stefano Pessina Drug Retail
Williams Alan Armstrong Oil & Gas Storage & Transport. Xylem Steven Loranger Industrial Machinery
Yum Brands Geg Creed Restaurants
Alphabet Sundar Pichai Interactive Media & Services CarMax Thomas Folliard Specialty Stores
H. & P. John Lindsay Oil & Gas Drilling Kellogg Steven Cahiliane Packaged Foods & Meats Lilly Eli David Ricks Pharmaceuticals
Lowes Marvin Ellison Home Improvment Retail Microsoft Satya Nadella Systems Software
Kimberly Clark Michael Hsu Household Products Quest Diagnostics Steve Rusckowski Health Care Services Chipotle Brian Niccol Restaurants
A.O. Smith Ajita Rajendra Building Products Hologic Stephen MacMillan Health Care Equipment
Ansys Ajei Gopal Application Software
Citrix Systems David Henshall Application Software Ford Motor James Hackett Automobile Manufacturers
Company CEO Industry General Mills Jeffrey Harmening Packaged Foods & Meats Kinder Morgan Steve Kean Oil & Gas Storage & Transport. Sherwin Williams John Morikis Specialty Chemicals
Parker Hannifin Thomas Williams Industrial Machinery
Fiserv Jeffrey Yabuki Data processing & outsourced services Sysco Corp Tom Bené Food Distributord
Genuine Part Paul Conahue Specialty Stores
Lamb Weston Thomas Werner Packaged Foods & Meats
Appendix B
Sample characteristics: articles per newspapers and genders Variable N Gender Female Male Newspaper New York Times The Economist Financial Times Wall Street Journal Business Wire Financial Wire Fortune Magazine Forbes Magazine CNN News Associated Press
Associated Press Online Associated Press Int. Others Sectors 231 204 43 3 61 132 40 25 22 8 19 57 8 8 8
Appendix C
All stock prices and averages per company before and after appointment Female CEOs Company 3 days prior 2 days prior 1 day prior MEAN 1 appoin tment 1 day after 2 days after MEAN2 Wendys 18,86 19,20 19,16 19,07 19,12 19,46 19,60 19,39 ADM 36,67 36,93 36,19 36,60 36,34 38,00 41,90 38,75 Mondelez 31,15 31,17 31,00 31,11 30,85 30,77 30,97 30,86 Pepsico 62,84 63,20 63,33 63,12 63,95 63,59 63,29 63,61 Wellpoint 81,69 81,76 81,50 81,65 81,13 77,90 79,39 79,47 Xerox 17,23 17,81 18,18 17,74 17,97 17,87 18,26 18,03 Keybank 8,65 8,68 8,67 8,67 8,71 8,82 8,63 8,72 Campbells 34,23 34,22 33,98 34,14 33,11 33,90 34,02 33,68 Tegna 4,79 5,13 5,47 5,13 5,47 5,45 5,61 5,51 IBM 177,25 181,63 182,25 180,38 180,36 181,97 185,88 182,74 Mylan 18,04 18,53 18,21 18,26 19,56 20,35 20,28 20,06 Alliant Energy 21,56 21,61 21,31 21,49 21,54 21,38 21,13 21,35 Avon Products 22,19 22,75 23,42 22,79 22,69 22,22 22,71 22,54 Yahoo 15,80 15,69 15,74 15,74 15,65 15,60 15,70 15,65 General Dynamics 62,72 61,96 61,98 62,22 63,62 63,69 64,06 63,79 Lockheed Martin 94,87 91,15 89,92 91,98 89,98 89,81 90,18 89,99 Duke Energy 67,49 67,64 67,65 67,59 68,23 66,72 65,10 66,68 Ulta Beauty 95,04 93,40 94,15 94,20 97,18 97,89 98,71 97,93 HCP 37,30 38,17 38,04 37,84 36,27 35,56 36,30 36,04 General Motors 39,09 40,17 40,90 40,05 40,4 40,16 40,05 40,20 American Water 41,10 40,93 40,36 40,80 40,75 40,80 40,96 40,84 Oracle 40,66 41,19 41,14 41,00 41,55 39,80 39,58 40,31 AMD 3,40 3,36 3,28 3,35 3,28 2,95 2,72 2,98 Occidental Petroleum 69,20 67,19 67,83 68,07 68,3 65,92 66,54 66,92 Staples 8,46 8,55 8,54 8,52 8,31 8,36 8,37 8,35 Progressive 33,15 33,60 33,15 33,30 33,56 33,24 33,04 33,28 Nasdaq 66,93 67,10 66,44 66,82 64,73 64,48 64,71 64,64 Mattel 29,51 29,59 29,52 29,54 30,91 30,47 30,07 30,48 Hersheys 101,16 101,46 102,61 101,74 103,17 104,44 103,80 103,80 PG&E 59,41 58,50 58,48 58,80 58,15 59,10 58,71 58,65
Female CEOs Company 3 days prior 2 days prior 1 day prior MEAN 1 appoin tment 1 day after 2 days after MEAN2 Accenture 190,31 191,09 191,71 191,04 194,67 196,21 195,32 195,40 Gap 14,03 18,82 14,28 15,71 13,36 13,19 11,77 12,77 Zoetis 124,59 125,12 122,29 124,00 124,28 127,05 126,07 125,80 DuPont 50,98 53,46 51,11 51,85 48,76 49,23 50,00 49,33 Male CEOS Company 3 days prior 2 days prior 1 day prior MEAN appoin tment 1 day after 2 days after MEAN 2 Adobe 46,17 45,03 43,24 44,81 42,19 40,86 41,02 41,36 Stryker 55,59 55,81 55,66 55,69 54,90 54,67 54,42 54,66 Target 60,30 61,10 61,38 60,93 59,59 59,85 60,70 60,05 NetApp 33,02 33,26 33,40 33,23 33,16 33,34 33,24 33,25 National Oilwell Varco 68,48 67,48 67,03 67,66 67,36 69,47 69,45 68,76 Praxair 59,40 59,48 59,15 59,34 58,05 58,25 58,51 58,27 Mastercard 25,57 25,69 25,89 25,72 25,96 25,85 26,31 26,04 Juniper Networks 23,49 22,36 22,89 22,91 22,57 26,57 25,52 24,89 HP 38,96 39,27 39,04 39,09 37,06 37,08 36,65 36,93 Hormel Foods 38,26 38,21 38,46 38,31 38,65 38,06 37,51 38,07 Home Depot 39,56 39,73 40,16 39,82 41,07 40,57 39,79 40,48 Hasbro 24,75 25,63 25,87 25,42 26,41 26,94 27,11 26,82 Gap 41,20 41,20 41,90 41,43 36,67 36,34 36,05 36,35 Vulcan Materials 64,92 64,50 65,16 64,86 64,37 64,20 64,41 64,33 Walt Disney 140,37 138,97 133,01 137,45 128,19 123,36 118,04 123,20 Walmart 78,90 78,86 79,81 79,19 80,43 80,68 80,93 80,68 Estée Lauder 12,40 12,52 12,63 12,52 13,00 12,86 12,32 12,73 Flowserve 29,51 29,50 29,80 29,60 28,75 27,60 28,29 28,21 Starbucks 57,59 58,17 57,97 57,91 58,51 57,21 57,50 57,74 Textron 19,44 19,25 19,37 19,35 19,20 18,34 17,88 18,47 Walgreens Boots 85,39 87,25 85,91 86,18 89,55 93,11 94,48 92,38 Williams 15,75 16,08 16,07 15,97 15,99 17,55 17,36 16,97 Xylem 27,36 27,53 28,23 27,71 27,99 27,93 28,24 28,05 Yum Brands 55,85 55,34 55,35 55,51 54,82 54,78 54,85 54,82 Alphabet 1312,99 1304,96 1289,92 1302,62 1295,28 1320,54 1328,13 1314,65 CarMax 16,55 16,34 16,20 16,36 15,81 15,88 15,78 15,82 Helmerich & Payne 61,28 60,67 60,43 60,79 60,57 60,25 61,14 60,65 Kellogg 63,04 62,81 62,37 62,74 62,72 62,02 62,42 62,39 Lilly Eli 84,74 82,19 81,84 82,92 83,02 83,10 82,66 82,93 Lowes 85,47 86,34 87,39 86,40 85,75 94,69 96,49 92,31 Microsoft 36,86 37,84 36,48 37,06 36,35 35,82 36,18 36,12
Male CEOS Company 3 days prior 2 days prior 1 day prior MEAN appoin tment 1 day after 2 days after MEAN 2 Quest Diagnostics 61,17 59,63 58,80 59,87 59,57 58,83 57,98 58,79 Chipotle 266,01 255,46 254,45 258,64 251,33 289,91 286,65 275,96 A.O. Smith 14,95 14,90 14,85 14,90 14,85 14,90 14,66 14,80 Hologic 22,26 22,47 22,29 22,34 22,12 22,21 21,77 22,03 Ansys 98,51 98,50 95,35 97,45 95,09 94,58 95,66 95,11 Citrix Systems 78,67 79,27 79,93 79,29 78,56 80,43 80,37 79,79 Ford Motor 10,67 10,70 10,79 10,72 10,98 11,05 10,96 11,00 General Mills 57,51 56,60 55,92 56,68 56,05 57,05 57,30 56,80 Kinder Morgan 41,22 40,77 41,53 41,17 41,82 42,00 42,08 41,97 Sherwin Williams 242,51 236,08 243,17 240,59 242,87 239,60 238,36 240,28
Appendix D
Table Reliability
Variable name % agreement Kappa α
Relevance 100% 1.00 1.00 Newspaper 95% .948 .972 Date 93% .908 .901 Picture 97% .944 .945 Name of CEO 100% .954 .977 Gender of CEO 100% 1.00 1.00 Physical Appearance 100% 1.00 1.00 Parenthood 100% 1.00 1.00 Private Life 93% .857 .859 Professional Background 97% .848 .850 Intelligence 97% .880 .812 Tone 91% .869 .870 Frame Coach 100% 1.00 1.00 Frame Diplomat 100% 1.00 1.00 Frame Visionary 95% .827 .830 Frame Innovator 95% .854 .855 Frame Commander 100% .791 .793 Frame Hero 100% 1.00 1.00 Frame Constructor 100% .791 .793 Frame Expert 86% .708 .709 Gender Emphasis 97% .956 .999
Appendix E
Regression Models 1 and 2 Outcome
ANOVAa Model
Sum of
Squares df Mean Square F Sig.
1 Regression 36,155 1 36,155 2,393 ,123b Residual 6193,468 410 15,106 Total 6229,623 411 2 Regression 205,329 9 22,814 1,522 ,138c Residual 6024,294 402 14,986 Total 6229,623 411
a. Dependent Variable: % change in average b. Predictors: (Constant), female
c. Predictors: (Constant), female, Q30Constructor_num=yes, Q31Expert_num=yes, Q29Hero_num=yes, Q25Diplomat_num=yes, Q28Commander_num=yes,
Q26Visionary_num=yes, Q27Innovator_num=yes, Q24Coach_num=yes
