The existence of the gender pay-gap
on the corporate executive level in
the U.S. of the last decade
Bachelor Thesis
Name: Mats Rooijers
Student Number: 11255579
Name Supervisor: Alejandro Hirmas
Date: 30 June 2020
Programme: Economie en Bedrijfskunde
Abstract The focus of this paper is the existence of the gender pay-gap for corporate executives in the US during the last decade. Reasons for the existence of the pay-gap are given in the first part of this paper. Thereafter, the effect of gender is measured using a Pooled OLS regression on annual base salary, annual bonus and annual total compensation. The sample is constructed of executives of companies from the S&P 500 from 2010 to 2019. The results of the used model, which are all significant for at least a 5% level, conclude that being a female could have a negative influence on the compensation package. However, this result is not totally clear as there are some missing control variables that could explain the model better.Table of Contents
Introduction ... 3 Literature review ... 4 Hypothesis ... 6 Research method ... 7 Sample ... 7 Model ... 7 Descriptive statistics ... 8 Results ... 10 Regular model ... 10 Alternative dependent variables ... 12 Conclusion ... 14 Discussion ... 15 Reference list ... 17 Appendix ... 20 Appendix A – Table 2 ... 20 Appendix B – Table 3 ... 21 Appendix C – table 4 ... 22Introduction In this day and age almost every country has laws regarding the diversity within the workplace. There are many quota’s and laws that see to the fact that both women and men are represented within companies. This wasn’t always the case as traditionally men worked and made a living while women took care of their families and the house. This traditional view changed over the last halve century and nowadays there are many women working for big companies (Law, 2020). The traditional problem however was that women didn’t had the same opportunities as men, which resulted in a pay-gap between men and women. Due to this many countries implemented legislation to ensure an equality of opportunity for both men and women. The discussion about the pay-gap however still remained, with many different viewpoints in the current political landscape (Abrams, 2019). Within the US there are also a lot of anti-discrimination laws and there are a lot of women attending good universities (“Anti-Discrimination Laws in USA,”2017). This results in the fact that women are becoming more represented in boards of companies all over the country. However, in the US the majority of the people still think a gender pay-gap still exists (Renzulli, 2019). Bugeja (2012) gives in his study plenty examples of other studies that came to the conclusion that the pay-gap still exists, like the study of Bell (2005). There it is stated that there is a gender pay-gap within the board composition in the United States. Although the study of Bugeja gives all these examples, his study instead states that for CEOs there is no pay-gap. Due to the different results of studies there is not a clear answer yet on the phenomenon that is called the gender pay-gap. This paper tries to formulate an answer on these topics. The research question is the following: Is there a pay-gap between men and women anno 2019 within the composition of a board in the US? and if there is a gap what could be the reasons for the gap? This paper only contains data on the US market and specifically the corporate executives in the S&P 500. The reason for this was that it was the best public accessible data. The first part of this paper contains the eventual reasons for the gender pay-gap. It deals with the question what the reasons could be the pay-gap exists. It also reviews the results of previous research and their reasoning for the existence of the gap. Following this literature review, I conducted my own small research to asses if there still exists a pay-gap between men and women in the executive positions of companies using linear regression. This contains the used research method and the results found. I conclude with an
explanation of the results and a discussion on the process and possible improvements on the analysis. The main conclusion of this paper is that gender has an influence on the compensation of executives. All the regressions result in a negative influence of being a woman and all these results are significant for at least a 5% level. Literature review Historically, some reasons have been given why the gender pay-gap exists. One of the oldest is that there is a difference in interest between men and women. Men were more likely to be educated in for example business which regularly results in a high wage, while women chose different educations of jobs were they would earn less (Polacheck, 1981). It was stated that women chose to work more par-time and they choose different fields of work that paid less. This argument doesn’t apply in this case as the research will be done on subjects in the same kind of job. The focus is on corporate executives of S&P 500 companies, which means that there should not be a difference in education as most corporate executives attended the best business schools all over the world. Croson & Gneezy (2009) give three reasons why man and women differ. These reasons could lead to a difference in payoff between men and women. The first reason is that women are more risk averse than men which would mean they would rather choose a higher fixed salary then bonuses. The second reason is the different social preferences between men and women. The last reason is that women are less competitive than men which results in worse negotiating skills. The first reason focusses on the difference in men and women in risk taking. Following Croson & Gneezy (2009) many studies show that a man is more willing to take risk than a woman. A reason for this could be that women are considered more emotional than men. Women can experience emotions more strongly which means they are more sensitive for the emotion that causes to think about the dangers of risk taking (Harshman et al, 1987). Due to this, women are more likely to select in jobs that generate more fixed income which is less risky than variable income. Fixed income tends to generate a lower outcome than variable income like bonuses or stock options (Corgnet et al, 2019). This could mean that because women are getting paid more in fixed wage which is lower than variable wage they earn less overall.
The second reason for the existence of the gender pay-gap is that men and women have different social preferences. The main observation Croson & Gneezy (2009) take out of this statement is that women tend to be more sensitive to social cues than men. This could mean that women have a lack of consistency or fixed pattern. They are more liable to vary or change their minds about things. A study of Andreoni & Vesterlund (2001) Confirms this theory by stating that women are more concerned about the fact that different parties have an equal pay scheme. Men however are more interested in maximizing their efficiency. The last reason brought up is that men are more competitive than women, which was tested in the study by Gneezy, Niederle and Rustichini (2003). The study confirmed that men tend to perform better in competitive atmosphere than women. Another study showed that women also shy away from competition as well, even if the outcome of the competition is more profitable and they have the same or better skills (Niederle et al, 2007). A consequence of these findings is that women tend to do worse in negotiations than men. They are more likely to accept the first offer they get and often don’t opt for a renegotiation of the offered wage (Cronson et al, 2009). This all could lead to women getting paid less than men. The study of Eckel & Grossman (2001) confirms this theory concluding that women are more likely to accept lower offers than men. All the previous reasons for the existence of the gender pay-gap have to do with the psychological differences between men and women. But these are not the only reasons that there could exist a pay-gap. For instance, in the study of Elkinawy and Stater (2011) they make the argument that firms discriminate against women because of the possibility of pregnancy. Even if a woman stated that they are not interested in having a family they are still seen as potential mothers which would cause them to lay down their work for months. While companies should not discriminate between men and women, it could be very disrupting if a person in a management position leaves for maternity leave. This could cost the company a lot of money and could be a reason that women are paid less. Another potential reason is the “taste for discrimination” model developed by Becker (1971). In the model it is suggested that if an employer has a distaste for a certain ethnic group or gender the employer will pay this group less than his preferred group, while they are equally skilled. In this case, if an owner of a corporation does not want a female executive in his board, the woman would get paid less.
However, it is to simple to suggest that most of the male corporate executives discriminate against women and that the pay-gap exists solely because of discrimination against women. For example, the human capital theory offers different reasons why the gender pay-gap could exist. These reasons include self-selection into lower-paying jobs and a lower job mobility. Therefore, it is very hard to determine what parts of the gender pay-gap are existing because of discrimination or if it is related to productivity or just the psychological differences between men and women (Oaxaca, 1973). Because of the difficulty to determine if a gender pay-gap exists research is divided and come to a lot of different conclusions. Examples of different research that came to the conclusion that gender has a significant influence on the compensation of executives are Bartlett and miller (1988), Truman and Baroudi (1994), Gray and Benson (2003). Roth (2003), Mohan and Ruggiero (2007) and Bell (2005). Research was conducted on different industries and all these papers suggest that gender has influence. However, there is also a lot of research that suggests otherwise. While some see a difference in compensation between men and women they argue that is not due to there gender but because of other reasons. Examples of these kinds of research are Lausten (2001), Bertrand and Hillock (2001), Bowlin, Renner and Rives (2003), Khan and Vieito (2008). The conclusion is that it is hard to measure if gender has an influence on the compensation of executives and that there is no definite conclusion found yet. Hypothesis In the literature review I gave examples of different studies that come to the conclusion that a pay gap exists. Due to the fact that it is still a highly debated subject with still no clear answer. However, most previous research came to the conclusion that a small pay-gap exists. Because of this, I suspect there to be a small difference ass well. H0: 𝛽𝑔𝑒𝑛𝑑𝑒𝑟 = 0 H1: 𝛽𝑔𝑒𝑛𝑑𝑒𝑟 < 0 However, as stated in the study by Cronson & Gneezy (2009), women tend to be more risk averse. This could mean that they are probably more likely to accept a contract with a somewhat higher fixed wage but with not much of variable wage options. There fore I suspect that the gender will play a smaller role in the fixed part of the wage then in the variable part of the wage. H0: 𝛽𝑔𝑒𝑛𝑑𝑒𝑟 𝑟𝑒𝑔𝑟𝑒𝑠𝑠𝑖𝑜𝑛 1 = 𝛽𝑔𝑒𝑛𝑑𝑒𝑟 𝑟𝑒𝑔𝑟𝑒𝑠𝑠𝑖𝑜𝑛 2
H1: 𝛽𝑔𝑒𝑛𝑑𝑒𝑟 𝑟𝑒𝑔𝑟𝑒𝑠𝑠𝑖𝑜𝑛 1 < 𝛽𝑔𝑒𝑛𝑑𝑒𝑟 𝑟𝑒𝑔𝑟𝑒𝑠𝑠𝑖𝑜𝑛 2 Research method Sample The data that is used in this thesis is collected out of WRDS using the Compustat extension. This dataset focuses mainly on the North American market. Initially my idea was to get data on the executives in the Netherlands, but this data was not accessible to me. That is the reason I went with the North American market instead. From Compustat the databases for Financial Fundamentals Annual and Execucomp Annual Compensation where used. The timeframe that is used is from the last decade so 2010 to 2019. I focused only on the companies from the S&P 500 to exclude small market companies. The data out of Execucomp was transformed into panel data to adjust for time. The panel was created on a yearly basis of every individual executive. It could not be done for every company as then there were to many observations that had the same company and the same year and many executives had to be dropped if this was the panel ID variable. After the panel was established the right variables were calculated out of the Financial Fundamentals Annual database and these were matched to the existing panel. Some values could not be matched which meant that I had to drop 678 observations which brought the total up to 25547 observations. In the sample it is possible for firms to have multiple observations, due to the fact that they have multiple people in the same fiscal year. That is why for all the following regressions the robust errors are used (Bugeja et al, 2012). Model The model that was used to estimate if gender plays a role in the salary of corporate executives is: 𝐶𝑜𝑚𝑝 = 𝛽0 + 𝛽1𝑓𝑒𝑚𝑎𝑙𝑒 + 𝛽2𝑎𝑔𝑒 + 𝛽3𝐿𝑜𝑔 𝑆𝑎𝑙𝑒 + 𝛽4𝑀𝐵𝑉 + 𝛽5𝑅𝑂𝐸 + 𝛽6𝑅𝑂𝐴 + ∑𝛽𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑒 𝑑𝑢𝑚𝑚𝑖𝑒𝑠 + Σ𝛽𝑆𝑡𝑜𝑐𝑘 𝑒𝑥𝑐ℎ𝑎𝑛𝑔𝑒 + 𝜀 The dependent variable in this model is the compensation of executives. The compensation is described in three different ways: total annual salary, annual base salary and annual bonus. This results in three models being looked at. To measure the total annual salary I use the natural logarithm of the TDC1 value in Execucomp. This value measures the entire annual earning including base salary, bonuses, stock options and all the other forms of compensation (lntotal). For the annual base salary (salary) and the annual bonus (bonus) I used the regular dollar values given in the database (Bugeja et al, 2012).
The main explanatory independent variable is a dummy variable for gender (female). All the other independent variables are control variables that are associated with the compensation of executives. The data is transformed into panel data so the model is adjusted to year effects. The control variables are: age, log(sale), market-to-book value, return on equity, return on assets, dummies for all industries and dummies for the exchanges the companies trade on. The variable age is included as research has shown that the age of an executive has influence on the compensation he receives (Ryan and Wiggins III, 2001). The control variables log(sale), market-to-book value, return on equity and return on assets are all economic control variables that control for the performance of the firm and thereby the executive board. The natural logarithm for sales (LogSale) is used as a control variable for size as executives of bigger companies tend to be paid more (Smith and Watts, 1992). The variable market-to-book value (MBV) is used to measure a companies’ investment opportunities as executives get paid more if their company has better investment opportunities (Murphy, 1985). The variables return on equity (ROE)and return on assets (ROA) are used to measure firm performance. Return on equity is used to measure the return on a firm’s stock and return on assets for the return on their assets (Core et al, 1999). The last control variables are dummies for different industries and dummies for the stock exchange the companies trade. The dummies for different industries are included because a different industry could pay lower or higher than the other. In total there are 126 dummies for different industries, as there were 127 industries but for the model I have to exclude one due to multicollinearity. This reasoning also counts for the dummies of the different exchange markets the companies change on. A different exchange market could mean that the compensation differs. There are three different stock exchange dummies on which the companies trade: The Nasdaq, The New York State Exchange and other exchanges. Two of the dummies are included in the model to account again for multicollinearity. Descriptive statistics On first glance the numbers suggest that there is a difference in the compensation among executives in the US. The averages of the three compensation categories used here are all higher for men during the last decade. This can be seen in table 1 where the descriptive
statistics are given for the entire sample. The difference between the averages are quite high, which suggest that the gender pay-gap could exist. Table 1- descriptive statistics Gender Male Female Mean Std. Mean Std. Salary 715.939 490,965 627.402 300.214 Bonus 194.898 807,173 169.910 709.858 Total 5.690.769 7.330,539 4.677.220 5.617.075 #observations 23.125 2.541 These high differences could be explained by the fact that the compensations of the CEOs are included in this sample. This is done because the CEOs are corporate executives as well, but they do drive up the averages compensation packages. Most of the CEOs in this sample are male so this could lead to higher averages for male executives in comparison to the female executives. In table 2, given in the appendix, the same descriptive statistics are shown as in table 1 except the CEOs are excluded to account for the higher compensation of CEOs. Nevertheless, while the averages are down they are still quite far apart in favor of the male executives. Both table 1 and 2 focus on the descriptive statistics of the entire sample over the whole period. Firstly, it is interesting to note that during the last decade the means of all three compensation measures have grown. Both male and female executives earn more on average in 2019 then they did in 2010. However, even if you look at every year individual, in table 3 in the appendix, you see that the means are favorable for men in every year in the last decade for all three compensation measures. All these descriptive statistics suggest that gender could have an influence in the compensation package of corporate executives as for every category a man earns more on average than a woman. A positive note however is that it seems more women are becoming part of executive boards. The last decade the percentage of women constantly grew from approximately 8% in 2010 to 12,5% in 2019. This means that boards are becoming more diverse which is good for a company’s performance (Carter et al, 2013). This is also beneficial for the executives as a good performance of their companies generates a higher
compensation package for them. The growth of women in the executive boards of companies is shown in figure 1, while the specific numbers for each year can be found in the appendix in table 4. figure 1 Results Regular model The results of the Pooled OLS Regression on the model are presented in table 5 for all three the compensation measures. All the regressions are conducted in the same way but the dependent variable in every regression is different as the dependent variable compensation is expressed in three ways. That means that in every regression all the control variables were used and that all the regressions are adjusted for time. For all the three compensation measures the explanatory variable gender seems to have a negative influence. In the first regression the variable gender has a negative influence on the base salary of 77,99 dollars at a significance level of 0,1%. Also the variables age and LogSale are significant both a 0,1% significance level and seem to both have a positive influence. As expected as higher sales of the company and a higher age of an executive both should have a positive influence on the compensation. It seems that except for LogSale all other economic control variables are not significant in this regression. This means that it looks like these variables are not accurate in predicting changes in the dependent variable base salary. The R2 value however for the entire model is the highest of all three the regressions which means that it best predicts the entire model. The second regression the variables are regressed on the annual bonus. The result is slightly less significant then the results of regression one with a significance level of 5%. The explanatory variable gender is again negative however, by 47,65 dollars for the bonus. Both age and LogSale are again significant with a 0,1% level. From the other economic control 0% 2% 4% 6% 8% 10% 12% 14% 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Percantage women in Executive Boards
variables only the return on equity (ROE) is significant with a 5% level. The control variables explain the model slightly better with more significant variables but the R2 of the entire model is lower. This model is also the only one in which the constant is not significant for at least a 5% significance level. The last regression is on the natural logarithm of the total annual salary. Again, the explanatory independent variable female is negative, this time at a 0,1% significance level as in regression 1. If the executive is female the dependent variable lowers by 12,5%. The third regression is the regression with the most significant control variables out of the three. Again age and LogSale are significant at a 0,1% level while ROE joins them. The market-to-book value is in this regression also significant at a 1% level. The only variable that is not significant is return on assets (ROA). This variable is not significant in any of the regressions and an argument could be made that it has to be excluded in following research. Although this regression has the most significant control variables, the R2 measure is lower than in regression 1 which means that the regression fits the data less good. Table 5 – Pooled OLS regression with dependent variables: base salary, bonus, natural logarithm total salary Independent variabels (1) (2) (3)
salary bonus lntotal
female -77.99*** -47.65* -0.129*** (10.46) (23.27) (0.0266) age 16.27*** -5.750*** 0.0278*** (1.029) (1.059) (0.00213) LogSale 109.5*** 91.41*** 0.268*** (10.71) (12.33) (0.0177) MBV 0.0000391 -0.0000595 0.000000320** (0.0000953) (0.0000853) (0.000000103) ROE 2.890 9.094* 0.0427*** (2.354) (3.905) (0.0129) ROA 0.712 -85.78 -0.0146 (41.52) (44.27) (0.0797)
_cons -1471.3*** -237.6 3.642***
(205.3) (133.1) (0.263)
N 21960 21960 21264
R2 0.3940 0.2184 0.2208
industry dummy Yes Yes Yes
Exchange dummy Yes Yes Yes
Adjusted for time Yes Yes Yes
Standard errors in parentheses Values of regression 1 & 2 x1000 * significant at a 5% level, ** significant at a 1% level, ***significant at a 0,1% level Alternative dependent variables The explanatory independent variable female had a lower effect on the dependent variable annual bonus in the first three regressions. This is surprising as in previous research it was stated that women were more averse to risk what could result in higher base salary and lower variable compensation like bonuses. It was also stated in that variable compensation is often higher than fixed compensation (Corgnet et al, 2019). Following this reasoning the variable female should have more of a negative influence on the annual bonus than on the annual base salary. The main reason that is not the case in this sample is because many executives in the dataset received a bonus of 0. Nevertheless, the executives often had way higher total compensations than the base salary which means that there is a variable component to their compensation scheme. All these 0 values in the sample resulted in less of an impact on the explanatory variable. Because of this I conducted two more regressions with different dependent variables for the annual received variable compensation, or bonus. These two new regressions are shown in table 6. The first regression is the same as the second regression in the first model. It is there to compare it to the two new dependent variables. The second regression in this table has the dependent variable variable wage. This variable is conducted by subtracting the annual base salary of the total compensation to create a measure for variable compensation. This variable is not flawless as there could be some annual amounts that are paid every year to the executive but most of this measure is considered to be variable compensation. The result is that the variable female has a way bigger negative influence than for the previous regression concerning the bonus. Being a woman has a negative influence of 1021,30 dollars on the variable compensation and the result is significant with a 0,1% level, which is better than the first regression. The control
variables age and LogSale are again significant at a 0,1% level. Both MBV and ROE are significant at a 5% level. This is slightly better than the first regression in which only ROE is significant. Unfortunately, the estimates don’t fit the data as well as in the first regression with a lower R2 value. The reason for this could be that the dependent variable’s value is not totally accurate with it containing some form of fixed compensation. The third regression has again just the bonus value as it’s dependent variable, however the 0 values are dropped so that you get a better estimate of the value of variable compensation. The result is that the female again has a negative effect on the bonus of 177,60 dollars at a significance level of 1%. This is more than in the influence on the base salary which the literature would suggest. The control variables however are not as good as in the previous regressions. Nonetheless, the R2 value is way higher which suggests that the estimates do fit the data better. So an argument can be made this regression better explains the influence of the explanatory variable. Table 6 – Pooled OLS regression with dependent variables: Annual Bonus, Variable Wage (Total salary – Base salary), Bonus without 0 values (1) (2) (3)
Bonus Variablewage Bonusnonzero
female -47.65* -1021.3*** -177.6** (23.27) (189.4) (61.72) age -5.750*** 114.0*** 20.18*** (1.059) (12.59) (3.533) LogSale 91.41*** 1694.0*** 352.0*** (12.33) (108.8) (32.21) MBV -0.0000595 0.00208* 0.000405 (0.0000853) (0.000824) (0.000717) ROE 9.094* 138.4* 39.87 (3.905) (58.20) (224.8) ROA -85.78 661.9 -61.40 (44.27) (787.1) (288.4) _cons -237.6 -18034.6*** -3000.0***
(133.1) (1182.7) (348.4)
N 21960 21268 3620
R2 0.2184 0.1542 0.5278
industry dummy yes yes Yes
Exchange dummy Yes Yes Yes
Adjusted for time Yes Yes Yes
Standard errors in parentheses Values x1000 * significant at a 5% level, ** significant at a 1% level, ***significant at a 0,1% level Conclusion This paper is about the existence of a gender pay-gap between corporate executives in the US during the last decade. In the beginning of the paper I try to explain the reasons why there could exist such a pay-gap and what previous research had concluded. Following this I conducted my own small research using a Pooled OLS regression model, with the variable female as it’s explanatory variable and executive’s compensation as the dependent variable. My hypothesis was that being female had a small negative influence on the compensation of executives and that the influence would be more negative for variable compensation like annual bonus. The results show that in this model the female gender plays a negative role for compensation. For the annual base salary and natural logarithm of the total compensation it was negative with a 0,1% significance level. For the bonus it was also negative but at a 5% level. However, for the annual bonus the explanatory variable had a less negative impact than for annual base salary, which was not expected. After running two more regressions with different measures for variable compensation the results changed and the negative impact on the bonus was higher than that on the base salary, which was inline with the theory, with a regression that explained the data the best. The main conclusion from this paper is that there is evidence that there could exist a pay-gap between men and women for corporate executives in the US. For every regression made the gender female was negatively related to the dependent variable and all these results were significant. The second conclusion is that the effect on variable compensation is more negative than that on fixed compensation. This second conclusion is not as strong as the first conclusion however as the data had to be altered quite a bit to come to this conclusion.
The reason the female gender has a negative influence could be that women don’t like competition as much which could lead to poor negotiation skills of contracts. Or that women tend to be more risk averse which could lead to lower variable compensation and thereby a lower total compensation package. All these reasons are not hard evidence however, so there is not a clear reason why the pay-gap exists. Discussion The conclusions and results of this paper do have some limitations. The first limitation has to do with missing information on the part of the executives. Originally I wanted to include more information about the executives personally like their experience and their educational background. This could be included as these factors are assumed to influence the compensation scheme (Ryan and Wiggins III, 2001). I could not find this information in the datasets that were accessible to me. There is also a limitation in the conclusions that I can make from the test results. There is a chance that there is a form of multicollinearity in the sample. In this case it is imperfect multicollinearity which means that there is a close linear relation within the regressors of the model. The reason for this imperfect multicollinearity could be because the standard deviations of the OLS-estimators are very high. Because the multicollinearity is imperfect I could still calculate the OLS estimators but it could lead to an imprecise estimation. Another limitation is excluding the other factors of the compensation package. I only focused on the base salary, annual bonus and annual total compensation. There are a lot of other components that influence the compensation package. Most of them could influence the statements made about risk aversion of women. Out of the variable pay scheme only the annual bonus is used in this paper. While I tried to calculate a measure for total variable compensation by taking the difference between base salary and total salary, is that not a totally accurate representation. The results could differ if you take every single component into account. For further research, all these other variables that were not used in this paper could be included to make a model that better estimates the influence of gender. The results would better explain the sample and would give more insight on the influence of gender on the compensation of executives in the US. Another interesting insight for further research is that it would be interesting to create a same kind of database as Execucomp for the
Netherlands or other European countries. Data on executive compensation for these countries is harder to find which makes it more difficult to do research on these countries.
Reference list Abrams, A. (2019, April). Here's What All the 2020 Candidates Have Said and Done About Equal Pay. Retrieved from https://time.com/5562209/equal-pay-day-2020- candidates/ Andreoni, J., & Vesterlund, L. (2001). Which is the fair sex? Gender differences in altruism. The Quarterly Journal of Economics, 116(1), 293-312. Anti-Discrimination Laws in USA. (2017, Mar). Retrieved from https://knowledge.leglobal.org/anti-discrimination-laws-in- usa/#:~:text=It%20is%20illegal%20under%20U.S.,)%2C%20disability%20or%20geneti c%20information. Bartlett, Robin L. and Timothy I. Miller (1988). “Executive Earnings by Gender: A Case Study,” Social Science Quarterly, 69 (4): 892-909. Bell, L. A. (2005). Women-led firms and the gender gap in top executive jobs. Bertrand, Marianne and Kevin F. Hallock (2001). “The Gender Gap in Top Corporate Jobs,” Industrial and Labor Relations Review, 55 (1): 3-21. Blau, F. D., & Kahn, L. M. (2007). The gender pay gap: Have women gone as far as they can?. Academy of Management Perspectives, 21(1), 7-23. Bowlin, W.F., Renner, C.J., and J.M. Rives (2003). “A DEA Study of Gender Equity in Executive Compensation,” Journal of the Operational Research Society, 54: 751-57. Bugeja, M., Matolcsy, Z. P., & Spiropoulos, H. (2012). Is there a gender gap in CEO compensation?. Journal of Corporate Finance, 18(4), 849-859. Carter, D. A., Simkins, B. J., & Simpson, W. G. (2003). Corporate governance, board diversity, and firm value. Financial review, 38(1), 33-53. Core, J., Holthausen, R., Larcker, D., 1999. Corporate governance, chief executive officer compensation, and firm performance. J. Financ. Econ. 51, 371–406. Corgnet, B., & Hernan-Gonzalez, R. (2019). Revisiting the trade-off between risk and incentives: The shocking effect of random shocks?. Management Science, 65(3), 1096-1114. Croson, R., & Gneezy, U. (2009). Gender differences in preferences. Journal of Economic literature, 47(2), 448-74. Eckel, C. C., & Grossman, P. J. (2001). Chivalry and solidarity in ultimatum games. Economic Inquiry, 39(2), 171-188.
Elkinawy, S., Stater, M., 2011. Gender differences in executive compensation: Variation with board gender compensation and time. J. Econ. Bus. 63, 23-45. Gneezy, U., Niederle, M., & Rustichini, A. (2003). Performance in competitive environments: Gender differences. The quarterly journal of economics, 118(3), 1049-1074. Gray, Samuel. R and Philip G. Benson (2003). “Determinants of Executive Compensation in Small Business Development Centers,” Nonprofit Management and Leadership, 13 (3): 213-27. Harshman, R. A., & Paivio, A. (1987). " Paradoxical" sex differences in self-reported imagery. Canadian Journal of Psychology/Revue canadienne de psychologie, 41(3), 287. Khan, Walayet A. and Joao Paulo Vieito (2008). “Gender and Executive Compensation in S&P Listed Firms,” Working Paper, 21st Australasian Finance and Banking Conference 2008 Paper, Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1136060 Lausten, Mette (2001). “Gender Differences in Managerial Compensation – Evidences from Denmark,” Working Paper, Institute of Local Government Studies, Copenhagen, Denmark. Retrieved From http://research.asb.dk/fbspretrieve/858/01-4_ml.pdf Law, T. (2020, January). Women Are Now the Majority of the U.S. Workforce — But Working Women Still Face Serious Challenges. Retrieved from https://time.com/5766787/women-workforce/ Mohan, Nancy and John Ruggiero (2007). “Influence of Firm Performance and Gender on CEO Compensation,” Applied Economics, 39 (7-9): 1107-13. Murphy, K., 1985. Corporate performance and managerial remuneration: An Empirical Analysis. J. Account. Econ. 7, 11-42. Niederle, M., & Vesterlund, L. (2007). Do women shy away from competition? Do men compete too much?. The quarterly journal of economics, 122(3), 1067-1101. Oaxaca, Ronald (1973). “Male-Female Wage Differentials in Urban Labor Markets,” International Economic Review, 14 (3): 693-709 Polachek, S. W. (1981). Occupational self-selection: A human capital approach to sex differences in occupational structure. The review of Economics and Statistics, 60-69. Renzulli, K.A. (2019, April). 46% of American men think the gender pay gap is ‘made up to serve a political purpose’. Retrieved from
https://www.cnbc.com/2019/04/04/46percent-of-american-men-think-the-gender- pay-gap-is-made-up.html Roth, Louise Marie (2003). “Selling Women Short: A Research Note on Gender Differences in Compensation on Wall Street,” Social Forces, 82 (2): 783-802. Ryan, Harley E. jr., Roy A. Wiggens III (2001), The Influence of Firm- and Manager Specific Charasteristics on the Structure of Executive Compensation, Journal of Corporate Finance, 7:101 - 123 Smith Jr, C., Watts, R., 1992. The investment opportunity set and corporate financing, dividend, and compensation policies. J. Financ. Econ. 32, 263-292. Truman, Gregory E. and Jack J. Baroudi (1994). “Gender Differences in the Information Systems Managerial Ranks: An Assessment of Potential Discriminatory Practices,” MIS Quarterly, 18 (2): 129-142.
Appendix
Appendix A – Table 2 Gender Male Female Mean Std. Mean Std. Salary 666.635 484.155 596.677 269.141 Bonus 184.245 756.167 171.329 715.369 Total 4.881.630 5.813.745 4.088.087 4.165.341 #observations 19.403 2.330Appendix B – Table 3
Fiscal Year Mean Salary Women
Mean Salary Men Mean Bonus Women Mean Bonus Men Mean Total
Women Mean Total Men
2010 569.464 649.678 207.544 242.720 3.859.800 4.899.884 2011 553.535 660.880 164.140 199.425 4.021.387 5.246.613 2012 561.261 674.113 152.236 193.570 4.140.593 5.258.527 2013 611.393 689.853 164.166 187.286 4.152.284 5.386.555 2014 629.758 704.206 177.276 186.617 5.185.432 5.864.594 2015 641.101 723.265 157.746 174.893 4.902.329 5.633.721 2016 646.093 735.203 155.855 170.885 4.976.729 6.017.178 2017 661.245 745.436 135.693 184.230 4.450.855 5.729.810 2018 669.842 783.961 184.830 195.839 5.176.337 6.191.608 2019 673.414 815.393 211.175 222.015 5.322.270 6.868.420 average all periods 627.402 715.939 169.910 194.898 4.677.220 5.690.769
Appendix C – table 4
Year FEMALE MALE Total percentage women
2010 189 2,245 2,434 7,765% 2011 209 2,378 2,587 8,079% 2012 224 2,392 2,616 8,563% 2013 226 2,406 2,632 8,587% 2014 243 2,412 2,655 9,153% 2015 272 2,367 2,639 10,307% 2016 280 2,37 2,65 10,566% 2017 304 2,351 2,655 11,450% 2018 309 2,249 2,558 12,080% 2019 285 1,955 2,24 12,723% Total 2,541 23,125 25,666
