!
Abstract
The effect of ‘busy’ directors serving on multiple boards on CEO compensation
The impact of the corporate governance of a firm on executive compensation has gained an
increasing attention from researchers mainly because of the recent scandals and the financial crisis. Prior studies do not find conclusive relationships between most of the different factors. The focus of this thesis is the relationship between ‘busy’ directors and CEO pay. Besides that, other control variables are used that present a clear outline of how CEO compensation is composed. The sample is drawn from 139 S&P500 manufacturing firms in the United States. The main findings of this study show that a higher percentage of ‘busy’ directors on the board of directors have a significant effect on CEO compensation and that firm size is also a big factor in the composition of executive compensation.
Paul van Dijk
10805222
Economics and Business
Finance and Organisation
Ross Gardner
Statement of Originality
This document is written by Student Paul van Dijk who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of contents Abstract 1 Table of contents 3 1. Introduction 4 1.1 Background 4 1.2. Research question 6 2. Literature review 6 2.1 Agency theory 6
2.1.1 Optimal contracting view 7
2.1.2 Managerial power view 9
2.2 Board structure 10
2.2.1 Board size 11
2.2.2 Outside directors 11
2.2.3 Outside directors who serve on multiple boards 13
2.2.4 CEO duality 14 2.2.5 CEO Tenure 14 2.3 Hypothesis 15 3. Research methodology 15 3.1 Sample Selection 16 3.2 Variable descriptions 16 4. Results 19 4.1 Descriptive statistics 19 4.2 Assumptions 21 4.3 Correlations 21 4.4 Findings 22 5. Discussion 26 6. Conclusion 27 7. References 29 8. Appendix 34
1. Introduction 1.1 Background
CEO compensation has been subject to much public criticism over the last decade, especially after the financial crisis of 2008. Coverage of executive compensation of bank managers flood the news and Shareholders, politicians, and also ordinary people felt involved in the discussion. Many people believe that CEOs in comparison to the average workers earn too much money. This phenomenon is not a recent issue, people have been talking about growing income inequality for several decades now. The rich get richer, and the poor get poorer. It is not strange that scandals involving companies giving their management extraordinarily high bonuses like Enron and HP were reported in the news in detail. It does not stop at these scandals, every year situations like these pop up, which upsets the average citizens every time.
All this attention to the topic attracted researchers in the academic field of corporate governance and firm performance to study what drives these compensation levels to rise. Different factors will influence the level and composition of the compensation package of executive board members and especially the CEO, which this research paper will focus on. Many researchers tried to find a relationship between firm performance and CEO compensation or the relationship of firm size to CEO pay, but most of these studies failed to explain a significant part of the CEO compensation. Researchers looked into the way that different corporate governance structures influence the
executive compensation (Core et al., 1999; Adams et al., 2008; Fernandes, 2008). Core et al. (1999) state that executive pay is higher when executives are powerful and the corporate governance of the firm is weak. This thesis will take the most important governance factor as control variables to test the effect of what is relatively new and, therefore, understudied in the academic research. The research done will focus on the influence of independent board directors that sit on three or more
boards, and therefore is the main variable. Throughout the thesis, this is also referred to as ‘busy’ directors.
As was stated, it is important to take all the important variables that make up the board of directors and control for the different characteristics. Also included are the variables of firm performance and firm size; especially firm performance should have a significant impact on CEO compensation. Tosi et al. (2001) found that more than 50 percent of the variance in CEO compensation was explained by the firm size, so it is important to take that variable into consideration as well. Jensen (1993) argues that the compensation of the CEO will be higher when the board has more members in it because the directors will pay less attention to the day-to-day activities and therefore lead to a weaker corporate governance. Aside from the influence of the busy directors, the addition of the CEO gender also creates new insights and can lead to more research being done on the subject. The dataset used in this thesis includes firms from the manufacturing industry that are listed in the S&P 500. The reason for this choice is that S&P 500 companies tend to have very transparent filings of the remuneration policy, and that the average compensation level is high. The
manufacturing industry is an understudied industry even though it accounted for 17.8 percent of GDP in the US economy in 2016 (Bureau of Economic Analysis, 2018) and that is why it is chosen for this sample.
The thesis is structured as follows. In the next section, an overview of existing theories on executive compensation and what corporate governance factors influence it is presented. This is followed by the discussion of the empirical research by further reviewing the dataset, the different variables and how they are operationalized for using them in the regression model. After that, the descriptive statistics, the correlations between the variables, and the OLS regression are presented. Finally, the
results are discussed and a conclusion is drawn from the output of the tests and suggestions for further research are given.
1.2. Research question
Is there an effect on executive compensation when directors serve on three or more boards in the manufacturing industry?
2. Literature review
In this section, the literature about executive compensation and corporate governance will be discussed. First, the executive compensation literature is introduced with the multiple views on the agency theory, as the structure of executive compensation is largely built on that. It begins with the description of the agency theory itself, followed by the optimal contracting view and finally the managerial power view is defined and how the compensation contracting is influenced by these visions. Next, the potential importance of board structure and size will be presented. Finally, the potential importance of both having directors that serve on multiple boards and that of the characteristics of the CEO will be included in this analysis.
2.1 Agency theory
The agency theory was initially developed by Berle and Means (1932). Later, the theory was revisited and further contributions were made by Jensen and Meckling (1976). The agency theory tries to portray the situation that occurs when two people or entities that work together pursue different goals and therefore have different interests. The theory assumes perfect rationality, contracting and informational conditions. It tries to interpret the problems that arise from the principal-agent relationship like the ‘hidden characteristics’ as well as the ‘hidden action’ problem. To help resolve this apparent conflict, instruments like structuring, monitoring, and alignment of the different parties’ contracts are applied. This, however, comes at a cost. These costs are referred to as
agency costs. Fama and Jensen (1983) argue that the value of output not exercised in a contract is also included in the agency costs if these are above the value of the higher performance of the agent. For decades researchers have tried to find an ideal combination of incentive instruments and monitoring practices to minimize the welfare loss that exists in contracts (Fama & Jensen, 1983). In the case of executive compensation, the CEO is the agent and the shareholders serve as the principal. In listed firms, the day-to-day activities are delegated to the executive board to run the company. It is of the shareholders’ concern that the CEO maximizes the firm value, but the CEO has to be motivated to do so. The salary is used to motivate the CEO in acting out of the interest of the shareholders (Bertrand & Mullainathan, 2001). Managers prefer their salary composed the way that they bear less risk of the salary decreasing, therefore they are risk-averse. In suppressing the risk, managers may pursue in activities that reduce the firm risk so the compensation will also contain less risk but also affect the stockholders’ wealth (Jensen & Murphy, 1990). However, shareholders are argued to be risk-neutral because they can diversify their portfolio to remove firm-specific risk. A problem is that the managerial actions are not fully observable by the shareholders.
2.1.1 Optimal contracting view
As firm value increases, the shareholders will benefit because they own the stocks of the company, but the CEO too has to benefit from the value increase. Therefore, granting executives with
sufficing incentives is essential. In the perspective of the optimal contracting view, the board of directors tries to put in these incentives in the contracts as cost-effect as possible through the executive compensation package (Bebchuk & Fried, 2003). An optimal contract could either be the result of efficient compensation negotiation between the managers and the board of directors or benchmarking from the other CEO compensation packages in the industry. It was Murphy (1986) who formed the optimal contracting view initially. Murphy states that because the CEO actions are hard to observe, the optimal way to measure the CEOs performance is to look at firm value.
Because the firm performance is not fully controlled by the CEO, a fragment like ‘luck’ has to be subtracted. Luck, in this case, is defined by exogenous variables which are beyond the control of the CEO, and rewarding him for this would, therefore, be useless (Murphy, 1986). In fact, this will only raise the costs because the CEO would have to be compensated for the extra risk he bears because of the salary being not only influenced by himself (Holmstrom, 1979). Murphy said that executives should be financially rewarded the same way as the shareholders would gain from the equity rise they got. This would remove the misalignment between the interest of both parties. Removing the proxy for luck, however, cannot be effectively done. Bertrand and Mullainthan (2001) found that in practice CEO compensation did have a significant relation to the luck variables.
So is there a relationship between the executive pay and the firm performance? Jensen and Murphy (1990) state that there is a relation between the two, all be it a small one. From their research, they saw a wealth increase of 3,25 dollars for the CEO for every 1000 dollars the firm value increases. They argued that the pay-for-performance method is the optimal one. Hall and Liebman studied the same relationship in their 1998 research paper, but whereas Jensen and Murphy did not found a strong relation, Hall and Liebman found that the pay-per-performance was notably higher. Both the executive compensation as well as its sensitivity to firm performance have risen considerably in the period from 1980 to 1994. The difference is explained by a few reasons.
To begin with, Jensen and Murphy used data from 1969 to 1983, but the issuance of stock to executives only really started in the 1980s, which had a positive effect on the relationship. Secondly, Hall and Liebman used a percentage change in the wealth of the CEO whereas Jensen and Murphy used dollars to express the number in. The absolute sensitivity may have seemed low for firms which their value is transcending one billion dollars, but in percentage, the variable part was a significant chunk of the total compensation. Therefore, they argued that firm performance is an important driver of the CEO’s salary.
2.1.2 Managerial power view
Besides these optimal contracting view on the agency theory, another perspective on executive compensation is important to look at; the managerial power view. This approach comes down to the idea that the CEO has quite a bit of control over the board, which includes the ability to set a large part of his own pay.
The managerial power view was first introduced by Bertrand and Mullainathan (2001), as a reaction to the optimal contracting view. The optimal contracting view and the managerial power view are the same if the compensation is not dependent on luck but only on the CEO’s actions to increase firm value. The problem with the optimal contracting view is that the executive compensation is only weakly linked to performance, so in this case, it is not a good compensation theory. Bertrand and Mullainathan say that the executive salary reacts on as they call it ‘lucky shocks’, in their research they used oil price, average industry performance and exchange rates as a proxy for luck. They concluded that the executive salary depends almost as much on CEO controlled variables as on the lucky variables.
Bebchuk and Fried continue to define the managerial power approach in their research. This approach differs from the principal-agent theory in that the principal-agent approach the CEO is paid just enough salary that he will not leave the firm for another, so the salary is relatively low. In the managerial power theory, the compensation is set as high as possible, with its limit determined by public perception. The public perception is defined as the market level of pay, but also an additional variable which Bebchuk and Fried call ‘Outrage costs’. Outrage costs arise when in this case, the CEO pay causes a public reaction for the reason that people think the salary is
Just as agency theory assumes that executives might not intrinsically seek to maximize the firm value, there is doubt to as the board directors will do the same. So, in that case, the board members are subject to an agency problem as well. Directors typically want to be re-elected to the board. Next to an appealing salary, a directorship can also yield beneficial business and social connections. And as the CEO plays a large role in the election of directors, directors have the tendency to favor the CEO (Bebchuk & Fried, 2003). It is therefore important to look at the different characteristics of the board directors.
2.2 Board structure
Now the theory around how executive compensation is made up and where it comes from has been discussed, and now the earlier mentioned board of directors is discussed. The board of directors has an important place in the corporate governance of a firm. The board has the power to fire, hire and determine the compensation of the management (Baysinger & Butler, 1985). The role of the board is to resolve the agency problems, which were discussed earlier, by firstly replacing executives that are not creating firm value and secondly, by making and confirming the remuneration policy of the firm so that the interests of the executive board and shareholders are aligned (Hazarika et al., 2012). Another role of the board is to grant the executives access to resources for running the firm
(Hillman et al., 2008). Commonly in North-America, firms operate using a staggered board. This means that the elections for the board are done in different timeframes. The main reason for this is to protect the board against the control of a hostile bidder and helps the managers to hold off hostile bids that may be favorable for shareholders (Bebchuk & Fried, 2003). Core et al. (1999) stated that if the board is relatively weak, CEO compensation is higher. This thesis will further cover the factors that influence the quality of the board of directors. In 2002 a law called the Sarbanes-Oxley Act was introduced in America, which pushed firms to be more transparent in their corporate governance and increase the quality of the financial reporting. By demanding this, the government
wanted to restore the confidence in the market after several cases of bankruptcy caused by fraud, like WorldCom and Enron. As a result of the Sarbanes-Oxley Act, boards have become bigger, are more independent and have more responsibilities (Adams et al., 2010). Now that the activities of the directors on the board and the characteristics are discussed, the influence of the different
components of the board composition on executive compensation is discussed.
2.2.1 Board size
What is meant by board size is the number of overall directors on the board of directors, so both inside and outside directors. There is much debate in the academic field about what the optimal board size of a company is. There are several studies that argue that that board size is negatively related to firm performance (e.g., Lipton & Lorch, 1992; Jensen, 1993; Hermalin & Weisbach, 2003). The reasoning behind this is that small boards might be better at discussing problems and making quick decisions. Hermalin and Weisbach (2003) state that agency problems will not happen that much in small boards as it does in larger boards. They also state that the negative relation between board size and firm performance is one of the more prominent findings across the scientific studies. Jensen (1993) sketched the idea that larger boards are less effective than smaller boards because they may be less able to discuss efficiently and monitoring may become less effective. Jensen states that therefore the bargaining power of the CEO will increase. This is in line with what Core et al. (1999) found. Their study results show that the CEO gets more salary if the board size is larger. Core et al. argue that the reason for this is that corporate governance structure tends to be weaker for larger boards.
2.2.2 Outside directors
Corporate boards are composed of inside and outside directors, also known as respectively
dependent and independent directors. In North-America, the board structure that is most commonly used is called a one-tier board. Dependent directors are the executives of the firm, also called the
management. The independent directors form the supervisory board. The (in)dependence of the director refers to the relationship of the director with the management of the firm. The independent director is selected by the shareholders to oversee the activities of the management board (Brudney, 1982) and has no business ties to the firm aside from the board membership. In this context, the outside directors are the principal and the insiders are the agent. However, as the outsiders are not the owner of the firm, their incentives are not clear (Hermalin & Weisman, 2003). Fama and Jensen (1983) argue that their incentives are to build a reputation of being a good monitoring director, on the other hand, state Hermalin and Weisman that the reputation of not being a troublemaker for the CEO can also be an incentive. Outside directors may be beneficial for the firm for the reason that they can evaluate the executive decisions with another perspective and fire inefficient insiders (Hermalin & Weisbach, 1988). Independent directors can also be harmful to firm performance because, in contrast to dependent directors, they lack expertise and easy access of independent information to objectively challenge executive decisions and the compensation packages (Bebchuk & Fried, 2003). Adams et al. (2010) state that the Sarbanes-Oxley Act included several
requirements that increased the monitoring work, and therefore the demand for independent directors. This drove firms to have a board with the majority of the directors being independent (Adams et al., 2010).
Some prior studies show that a more independent board has a positive correlation with firm value, like that of Barnhart et al., (1994). However, some studies also state that if there is a majority of outside directors this does not significantly affect firm performance, like the research of Bhagat and Black (1999). So the impact of outside directors on firm performance is vague, but what about executive compensation? Core et al. (1999) find that executive compensation and the percentage of independent directors are positively related. Outside directors may not be better at monitoring than insiders (Fernandes, 2008).
2.2.3 Outside directors who serve on multiple boards
In prominent academic research, an independent director who also is a director who serves on multiple other is called a ‘busy’ director (Core et al., 1999; Ferris et al., 2003; Fich & Shivdasani, 2006). In these papers, the definition of a busy director is an outside director sitting on three or more boards in total, which will be used in the regression as well. The definition of a ‘busy’ director. Shareholders condemn firms which have multiple busy outside directors in the board because it is in their opinion that those directors cannot effectively monitor the executive board of multiple firms (Ferris et al., 2003). Too many board appointments might cause the director to oversee things in their monitoring job or that they might be unavailable at times they are needed the most (Korn, 1998).
However, they could bring additional value to the board. Their external corporate ties can contribute to a fresh perspective in regard to what the optimal compensation package is for the CEO, so
Larcker and Tayan (2011) suggest. Furthermore, busy directors could have the reputation of having high integrity which can commit to an increased demand for their service on boards. (Tian et al., 2011).
The results of the studies on this subject are indecisive. On the one hand, do researchers like Fich and Shivdasani (2006) and Core et al. (1999) find that busy directors are correlated with less
monitoring, higher CEO pay and worse firm performance. On the other hand, does research done by Ferris et al. (2003) find no relationship between busy directors and their monitoring capabilities. In this analysis, the potential relationship of busy directors on executive compensation will be
2.2.4 CEO duality
In North-America, it is usual that the CEO also is the chairman of the board, and therefore has both the responsibility of being CEO but also that of being the head of the board (Rechner & Dalton, 1991). The difference of the two functions is that the CEO is responsible for managerial work, whereas the chairman of the board is accountable for picking executive directors, reaching to decisions with the board and for putting things on the agenda). It serves a prominent role in the communication both internal and external and has the control over the composition of the board of directors (Larcker & Tayan, 2011). Many studies found that if a CEO is also chairman of the board, he will acquire more power over the board, which weakens the objectivity and consistency of the board (Patton & Baker, 1987; Hambrick & Finkelstein, 1987; Harrison et al., 1988; Morck et al., 1989).
CEO duality would compromise the ability of the board to monitor the executive directors in an independent manner (Fama & Jensen, 1983). Larcker and Tayan (2011) came to the conclusion that when the CEO and chairman position were occupied by different people that the probability of opportunistic behavior by the CEO decreases. Sauerwald et al. (2014) found that CEO hires more dependent directors if the CEO is also chairman, which magnifies his power. A merger of the two jobs of the CEO and chairman will also make it harder for the board of directors to replace an underperforming CEO, which also would increase power (Goyal & Park, 2002).
2.2.5 CEO Tenure
CEO tenure is the duration of the period between when a particular CEO was appointed as the CEO and when he stopped as being the CEO. Hill et al. (1991) argued that the tenure of a CEO was an important factor in determining how much power a CEO has over his own compensation. A reason why might be that a CEO who has a longer tenure can elect new executives throughout his
longer tenure can be that he might be able to learn how the company informational system works, which he can manipulate to let some information be withheld (Hill et al., 1991). Hermalin and Weisbach (1998) state that CEOs with a longer tenure have more power over the board of directors in general because the CEO can indirectly set his own targets in the long-run. This is why it is important to test this characteristic against CEO compensation as well. Johnston (2002) finds that CEO compensation is linked to his tenure, but that effect is from the relatively good performance the CEO has performed over the years with his firm. What Johnston (2005) found in his newer paper was that tenure is negatively correlated with the total compensation package, contradicting the theory by Finkelstein and Hambrick and Hill et al. (1991).
2.3 Hypothesis
Directors sitting on multiple boards have a positive effect on executive compensation in the manufacturing industry.
3. Research methodology
In this section, the methodology will be explained. It contains data descriptions, a definition of the variables, the hypothesis and the regression model that is used for the empirical study. There will be an explanation of why the specific variables are being used and how they are calculated. The
hypothesis will be formulated based on the literature and empirical findings of others. For the specific research done, the research methodology is based on a quantitative approach for the purpose of applying economic reasoning to the statistical generalizations made. Quantitative research has a few advantages like the universality of the output in numbers, and they can, for example, be interpreted globally. However, quantitative research has to make sense. The context of the results has to fit into realistic boundaries to really be useable (Curwin et al., 2013, p. 66).
3.1 Sample Selection
In this thesis, several corporate governance variables are tested on CEO compensation, as well as some control variables. The research sample will consist of 139 firms from the S&P500 which are drawn from the manufacturing sector using Standard Industrial Classification (SIC codes 2000 till 3999). Because most information is available for North-American firms, the tests contain firms from that country as there are large institutions that register the corporate governance data, and the data will not completely be available for other countries. The corporate governance data used for the analysis of the characteristics of the board is drawn from ISS, and executive compensation data and financial data about the firm is drawn from Standard & Poor’s ExecuComp and Compustat Capital IQ, which contains information about public companies in the US. The three databases will be merged together using the unique firm Committee on Uniform Securities Identification
Procedures (CUSIP) codes so each firm is linked correctly with its characteristics. Firms with missing values on executive compensation or governance data will be removed. Data covered is from the year 2015 so that changing variables like the economic growth and switching CEO
positions in firms through the years can be ruled out. The year 2015 has been chosen due to the fact that the financial crisis has passed, and the prediction is that firms will be on the output level that they were on before the crisis. In this study, a regression model is used to analyze the relationship of the independent variables on the dependent variable.
3.2 Variable descriptions
Formula 1
In this section, all the variables used in formula 1 will be discussed. The influence of the
independent variables on CEO compensation will be tested, plus the control variables that are added to correct for unwanted effects and to ensure the validity of the regression.
First of all, the dependent variable of executive compensation is going to be defined. In formula 1 it is defined as y(COMP). For the research, the total annual compensation package reported by the SEC is used, which contains the base salary, bonuses received, the total value of restricted stock granted, other annual compensation, restricted stock grants, long-term incentive payouts and the value of option grants. Other compensation, for instance, can be things like tax reimbursements or other personal benefits. The data from the SEC is very useful because it contains the important things to make up the compensation package of the CEO, and it is very reliable because it comes from the SEC. This data was available in the ExecuComp database.
Secondly, the independent variables will be discussed. The emphasis of the independent variables lies in the board structure of the companies, and especially that of the percentage of busy directors, in formula 1 defined as DBUSY, and from the result section onwards, it is also called “board busyness”. As mentioned in the literature review, a busy director is defined as an outside director sitting on three or more other boards. This variable will be a percentage, namely busy directors divided by outside directors.
Board independence is calculated by adding up the outside directors on the board and that outcome divided by the total board members. Board independence in formula 1 is defined as INDEP. An outside director is defined as someone who has not been employed by the firm in the five previous years, does not own more than ten percent stock of the firm and is not a relative of any director. Board size is calculated by adding up all the board members and defined as BSIZE in formula 1. A dummy variable is used for CEO duality: when the CEO also is the chairman of the board, the variable equals 1, and 0 otherwise. The variable is defined as DUAL in formula 1 and is a binary variable. Another CEO characteristic to account for is the tenure period the CEO has been employed by the firm, defined in formula 1 as TENURE. This variable is computed by taking the years
between 2015 and the year the CEO started being the CEO. This starting date was taken from the ISS database.
To control for the outcomes of these variables, the age of the CEO and the gender of the CEO are added to the regression model, respectively defined as AGE and FEMALE. Age of the CEO is related to the executive compensation, but the effect decreases as the manager gets older (McKnight et al., 2000). Gender is a popular topic regarding salary these days, so that variable is included in the regression as well. Age is a discrete number, and if the gender of the CEO is 1, then she is female. There is some research done on the effect of gender, but results are mixed (Adams et al., 2007; Mohan et al., 2003). There is no proven theory to back up these results.
To ensure that there is validity in the regression model, certain firm characteristics are added as control variables. The first control variable is firm performance, for which Tobin’s Q is used to calculate. This variable is defined as FPERFORM in formula 1. Tobin’s Q ratio stands for the ratio between the market value and the replacement costs of its assets. This ratio is used because it has the strongest linkage with corporate governance as a performance measure (Love et al., 2011). A value under 1 is not favorable for investors to invest in the firm, while a value higher than 1 does. Besides firm performance, also included as a control variable is the size of the company, defined as FSIZE in formula 1. Bigger firms tend to compensate executives more because they run a bigger firm, with more responsibilities. Jensen and Murphy (1990) however say that salary is negatively correlated because bigger firms tend to have stricter governance; therefore, the CEO has a harder time influencing his contract terms. Tosi et al. (2000) found that firm size counted for more than 40 percent of the variance in CEO compensation levels. Therefore, the firm size in this regression is the total assets of a firm in the year of 2015, following Tosi et al. (2001).
To make sure the model would be a good predictor of CEO pay, more variables have been tested on CEO compensation. For the factor firm size these include; total employees and total sales of the year 2015. Alternatives used for firm performance include return on assets (ROA) and return on equity (ROE). The alternative tests and results of these tests can be found in the appendix in table 7. This table shows the variables used and results in the highest R-squared value. In the table also the reliability of the data is being shown because every variable mix gives more or less the same t-value with it.
4. Results
In this section, the characteristics of the variables as well as the results of the OLS regression are being analyzed.
4.1 Descriptive statistics
In this paragraph, the results of the OLS regression are presented and evaluated. First, the continuous variables are presented in Table 1. After the continuous variables, the categorical variables are being shown in Table 2.
Table 1: descriptive statistics of the continuous variables
The descriptive statistics presented appear to be normally distributed, with the exception of total assets that have a skewness and kurtosis that is relatively high. However, before removing outliers also CEO compensation, CEO tenure and Tobin’s Q were leptokurtic, and the variable of assets was
more leptokurtic than it is now. The sample began at 165 firms, but due to lack of information and other criteria these have been removed, and 139 unique values of firms remained. Criteria were the following; CEO pay and sales expressed in USD, financial data has to be published in 2015, and outliers are to be removed. Regeneron Pharmaceuticals’ CEO had a total compensation of
$47,462,526 which is removed from the database. Secondly, Gilead Sciences’ ROA of 34.93% has been removed. Thirdly, a few Tobin’s Q numbers have been removed due to their extraordinary high values; Vertex Pharmaceuticals Inc with a Tobin’s Q of 12.40, Illumina Inc with a value of 7.29, Monster Beverage Corp with a value of 5.35 and lastly Colgate-Palmolive Co with a value of 4.97.
The independent variable CEO compensation has a dispersion of $28,215,338 between the lowest and highest earning CEO; this, however, does not make the standard error relatively large to the mean. Where the standard deviation can be called problematic is the CEO tenure and even more so for the total assets. The standard deviation of the CEO tenure is almost as high as its mean, which implies that CEO tenure fluctuates a lot in the data, especially for the total assets. To remove the kurtosis and the skewness of the total sales and make the distribution of the variable more normal, a natural logarithm is taken. Also remarkable is the negative skewness of board independence and CEO age. Nevertheless, this does that not require manipulation of the variables to qualify for the regression.
In Table 2 the descriptive statistics of the dummy variables are shown. What is remarkable is the frequency of gender; a mere 8 of the 139 firms had a CEO that is a woman. Also, the high number of firms that had someone who was both the CEO and seated on the board as chairman was surprising, looking at that the Sarbanes-Oxley Act did not really take the CEO duality numbers to below 50 percent.
4.2 Assumptions
Several assumptions have to be met in order to draw conclusions from an OLS regression (Curwin et al., 2013, p. 488). One of the assumptions is that independent variables do not have a perfect linear relationship. To check the data for this, the Variance Inflation Factor (VIF) of the variables have to be analyzed (Curwin et al., 2013, pp. 503). If this value rises above 10, there is a chance of multicollinearity. As shown in Table 3 in the appendix, the VIF values lie between 1.07 and 1.71 with a mean of 1.40. This means that that assumption is valid for this data. The data of the regression is also normally distributed because the natural logarithm of assets has been taken.
4.3 Correlations
A correlation value equals 0 if there is no linear relationship between two variables. As shown in Table 4, the largest correlation between two independent variables is between CEO duality and CEO tenure (r(129)=0.52) which is quite high. A possible explanation for this might be that CEOs who also serve as chairmen of the board of directors have more power to lay off plans to choose another CEO in his or her place if the board is not satisfied with the performance of the CEO. Another explanation for the large correlation value may that more experienced CEOs are considered to also become chairman. Looking at the second highest correlation value, the value between board independence and board size (r(129)=0.50) also is relatively high. Because the way the board independence is calculated, namely dividing the outside directors by the total directors on the board (which equals board size), the variables will correlate. The importance of both the variables in this
study exceeds the need to drop one of the two variables. Finally, the most remarkable correlation figure is discussed, the negative correlation between the natural logarithm of total assets and Tobin’s Q (r(129)=-.041). This is simply due to the fact that Tobin’s Q is calculated by dividing the market value of the assets by total assets on the balance sheet. So, the Tobin’s Q decreases when total assets increase. Despite some high values in Table 4, the coefficients are small, and, therefore, measure other aspects of the composition of the CEO pay. This means that the variables will not cause any trouble for the regression.
4.4 Findings
The results of the multiple regression are shown in Table 5. What can be seen is an R-squared of 0.545 which is high, because the findings in former literature state that a large part of the CEO pay variance is explained by luck(Bertrand & Mullainathan, 2001). The F-test gets a value of F(9,129) = 20.54 which is significant (p<.001).
Table 4: Correlation coefficients
Kolom1 CEO compen sa0on Board busynes s Board indepen dence Board size CEO duality CEO
tenure CEO age Female
Tobin's Q ln of total assets CEO compen sa-on 1.00 Board busynes s 0.29 1.00 Board indepen dence 0.30 0.02 1.00 Board size 0.27 0.05 0.50 1.00 CEO duality 0.28 0.03 0.17 -0.08 1.00 CEO tenure -0.05 0.02 -0.01 -0.04 0.52 1.00 CEO age 0.24 0.03 0.17 0.01 0.34 0.37 1.00 Female 0.31 -0.02 0.13 0.08 0.08 -0.03 0.07 1.00 Tobin's Q -0.18 -0.06 -0.13 -0.11 -0.04 0.20 0.14 -0.14 1.00 ln of total assets 0.63 0.22 0.32 0.32 0.08 -0.24 0.03 0.22 -0.41 1.00
As shown in Table 5, board busyness has a significant positive effect on the CEO compensation with an α <.01 significance level. With a beta of 54228, this means that a 1 percent increase in the board busyness means a $54,228 increase in the CEO compensation. The hypothesis was; directors sitting on multiple boards have a positive effect on executive compensation in the manufacturing industry. It can be concluded that the hypothesis is right.
Board independence has an insignificant positive effect on CEO pay. It was expected that the higher the ratio of outside directors to the total directors on the board, the compensation would be lower because monitoring was argued to be tighter, and so the CEO would not have as much power over his or her salary. On the contrary, a positive beta of $3250 is shown. This result is in line with the findings of Fernandes (2008), who found that firms with more outside board members pay higher wages to their executives, and thus also the CEO.
Following board independence, the result of the effect of board size is insignificant as well. All be it that it has a positive effect on the compensation. This is somewhat unsurprising because of how former literature was clear on that boards composed of more directors are less effective than those composed of fewer people. This is in line with what Core et al. (1999) found; however, due to the insignificance conclusions cannot be made.
CEO duality is an important factor in the composition of CEO pay, as it is significant at the 1 percent confidence level. If the CEO also is chairman, he sees his compensation climb up by approximately 2.85 million dollars. This is backed by former research findings such as that of Finkelstein and Hambrick (1996) and Core et al. (1999).
An unforeseen effect is shown of CEO tenure. From the literature available, one would think that a longer tenure would increase the pay of a CEO, but what this research found is that compensation actually decreases with an increase in CEO tenure with a non-statistical significance of p= .10,
which is officially not significant. This might be explained by that CEOs with a longer tenure instead of salary want and get long-term stock options (Hill et al., 1991) where this research did not focus on. Another reason might be that nowadays, more than ever, there is a big competition
between companies wanting the most suitable person for the position of CEO and that can be why beginning pay is higher now than it was twenty years ago. In the correlations section, the correlation between CEO tenure and CEO duality was reasoned by the more experienced a CEO is, the more chance he has to become chairman, what is only included in Table 6 in the appendix.
The first CEO characteristic control variable, CEO age, is also significant and in line with what former literature findings say. With a significance level of p = .05, it can be concluded that age has a significant positive effect on CEO compensation. A possible explanation would be that older CEOs have a longer career; therefore are compensated for their extra experience. The second CEO characteristic is the gender of the CEO. This variable also is significant at the p = .05 level and is positively related to CEO compensation. There is no real conclusion over why this is the case. A possible driver for the higher compensation can be that now with the discussion of women earning less money for the same job as men that companies decide to have countered this. This, however, does not explain the positive relation. A possibility is that because 75 percent of the females in this sample is also chairman of the board, that this CEO duality influences the gender variable as well. The controlling variable for firm performance, Tobin’s Q, was positive but was not significant. This was surprising because one would argue that this has to be a substantial part of the composition of the CEO pay. But in line with Tosi et al. (2000), firm performance does not matter much in the composition of CEO pay. This finding makes a review of compensation composition seem needed. An explanation for this could be that luck variables are far more important, like Bertrand and Mullainthan (2001) found. Finally, controlling for firm size, the effect of the ln of total assets is a part of the regression. And also in line with what Tosi et al. (2000) found, the variable total assets is
responsible for a large part of the CEO compensation. It is positive and significant at a significance level of p = .01 with a t-value of over 5. The firm size is the best predictor in this regression.
5. Discussion
As shown in the results section, a high percentage of the variables have a significant effect on CEO compensation. The starting point of this thesis was to show that a higher percentage of busy
directors in the board of directors has an effect on the compensation of the CEO, but what was clear from the literature study which was done, is that a lot of corporate governance factors influence the CEO pay as well. The most important variables according to former research were used in the OLS regression to reduce the omitted variable bias.
Looking at the influence of the variables of the board busyness, CEO duality, CEO age and the total assets of the firm, the findings of this research are backed up by former research (Tosi et al., 2000; Core et al., 1999; McKnight et al., 2000). Other variables such as CEO tenure and CEO gender gave the opposite effect of what was predicted, which was remarkable. The negative relationship of the CEO tenure to CEO pay and the positive relationship of CEO age to CEO pay was strange initially, however it is justifiable if more thought is put into the matter. Another surprising result of the OLS regression is the positive significant effect of the CEO being female on her compensation level. In the news and on other media it is argued that women are getting paid less than their male colleagues that are doing the same work, on the contrary, female CEOs make more money than male CEOs in this sample group. In the literature, there is no real economical reasoning as to why this is, but what has to be said is that in this sample the amount of female compared to male CEOs is very low and that 75 percent of the female CEOs is also chairman of the board. This might exaggerate the positive effect of being female as a CEO is to their compensation. If this relationship is to be researched in the future, a sample distribution of 50 percent female and 50 percent male CEOs should be drawn to really say something about the differences in gender to CEO pay. Furthermore, looking at the percentage of women on board of directors can also contribute to academic research. For the manufacturing industry as of 2015 the average amount of women on board in the sample
used, were lying around 10 percent of the board, which theoretically cannot have a significant effect. In other industries and countries the percentage of women on the board of directors differs though.
Furthermore, a few variables that had a predicted relationship with CEO compensation did not have a significant outcome. The board independence, board size, CEO tenure, and Tobin’s Q fall under this category. Tobin’s Q should have been significant looking at the theoretical composition of the compensation that says that firm performance is a big motivator for a higher compensation. Therefore, it is strange that this variable is not significant at all. The board independence is
predicted to have a negative relationship with executive compensation from the literature, but in the regression outcomes, it is positive and insignificant. The outcome is the same as what Fernandes (2008) found, that independent directors may not be better at monitoring the executives than
dependent directors. In other words, a higher percentage of board independence does not change the CEO pay.
6. Conclusion
This section presents the conclusion of the research done and provides an overview of the
contribution to former research. This paper adds to prior research because it uses a different set of variables together to focus only on the manufacturing industry in the United States of America, and with it, some surprising findings.
Executive compensation was and still is a popular topic in the news, and to this day researchers are not sure what the different building blocks are of executive pay. In this research, the effect of busy outside directors on executive compensation is was investigated and concluded that more busy directors on the board does influence the CEO compensation significantly in a positive way. Also, the variables that make up the highest R-squared are being tested besides that. Besides the main
variable, the variables of CEO duality, CEO age and CEO gender, and the logarithm of total assets have proven to be significant. First of all, if the CEO also is chairman of the board, the CEO pay becomes larger. Secondly, if the CEO is female, she will earn more then her male colleagues. After that, the age of the CEO has a positive correlation with executive compensation. The older the CEO is, the higher his salary is. The variable total assets has a significant positive effect on the level of executive pay. To summarize, enough evidence is found to validate to relationship between the percentage of busy directors and CEO compensation, and therefore, the hypothesis is accepted.
7. References
Adams, S. M., Gupta, A., Haughton, D. M., & Leeth, J. D. (2007). Gender differences in CEO compensation: Evidence from the USA. Women in Management Review, 22(3): 208-224. Adams, R., Hermalin, B., & Weisbach, M. (2010). The Role of Boards of Directors in Corporate
Governance: A Conceptual Framework and Survey. Journal of Economic Literature, 48(1), 58-107.
Baysinger, Barry D., & Henry N. Butler (1985). Corporate governance and the board of directors: Performance effects of changes in board composition. Journal of Law, Economics, and Organization 1.1: 101-124.
Barnhart, S., Marr, M. & Rosenstein, S. (1994). Firm performance and board composition: some new evidence. Managerial and Decision Economics 15, 329-340.
Bebchuk, L., & Fried, J. (2003). Executive Compensation as an Agency Problem. Journal of Economic Perspectives, 17(3), 71-92.
Berle, A., & Means, G. (1932). The modern corporate and private property. McMillian, New York, NY.
Bertrand, M., & Mullainathan, S. (2001). Are CEOS Rewarded for Luck? The Ones without Principals Are. The Quarterly Journal of Economics, 116(3), 901-932.
Bhagat, S. & Black, B. (1999). The uncertain relationship between board composition and firm performance. The Business Lawyer 54, 921-963.
Brudney, V. (1982). The independent director: Heavenly city or Potemkin village?. Harvard Law Review: 597-659.
U.S. Bureau of Economic Analysis, “Table 1 Gross output by industry,”
reqid=51&step=51&isuri=1&5114=a&5102=1#reqid=51&step=51&isuri=1&5114=a&5102 =15 (accessed January 10th, 2018).
Core, J.E., Holthausen, R.W. & Larcker, D.F. (1999). Corporate governance, chief executive officer compensation, and firm performance. Journal of Financial Economics 51, 371-406. Curwin, J., Slater, R., & Eadson, D. (2013). Quantitative methods for business decisions (7th ed.).
London: Chapman and Hall.
Fama, E. F. & Jensen, M. N. (1983). Separation of ownership and control, Journal of Law and Economics, 26, 301-325.
Fernandes, N. (2008). EC: Board compensation and firm performance: The role of “independent” board members. Journal of Multinational Financial Management 18, 30-40.
Ferris, S., Jagannathan, M., & Pritchard, A. (2003). Too Busy to Mind the Business? Monitoring by Directors with Multiple Board Appointments. Journal of Finance, 58(3), 1087-1111. Fich, E., & Shivdasani, A. (2006). Are Busy Boards Effective Monitors? Journal of
Finance, 61(2), 689-724.
Goyal, V., & Park, C. (2002). Board leadership structure and CEO turnover. Journal of Corporate Finance, 8(1), 49-66.
Hall, B., & Liebman, J. (1998). Are CEOs Really Paid Like Bureaucrats? The Quarterly Journal of Economics, 113(3), 653-691.
Hambrick, D. C., & Finkelstein, S. (1987). Managerial discretion: A bridge between polar views of organizational outcomes. Research in Organizational Behavior, 9, 369-406.
Harrison, J., Torres, D., & Kukalis, S. (1988). The Changing of the Guard: Turnover and Structural Change in the Top-Management Positions. Administrative Science Quarterly, 33(2), 211-232.
Hazarika, S., Karpoff, J. M., & Nahata, R. (2012). Internal corporate governance, CEO turnover, and earnings management. Journal of Financial Economics, 104(1), 44-69.
Hermalin, B. E., & Weisbach, M. S. (1998). Endogenously chosen boards of directors and their monitoring of the CEO. American Economic Review, 96-118.
Hermalin, B.E., Weisbach, M.S., (2003). Boards of directors as an endogenously determined institution: a survey of the economic literature. Economic Policy Review 9, 7-26.
Hill, C. W., & Phan, P. (1991). CEO tenure as a determinant of CEO pay. Academy of Management Journal, 34(3): 707-717.
Hillman, A. J., Cannella, A. A., & Paetzold, R. L. (2000). The resource dependence role of
corporate directors: Strategic adaptation of board composition in response to environmental change. Journal of Management studies, 37(2): 235-256.
Holmstrom, B. (1979), Moral Hazard and Observability, Bell Journal of Economics, 10: 74-91.
Jensen, M. C. (1993). The modern industrial revolution, exit, and the failure of internal control systems. the Journal of Finance, 48(3), 831-880.
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of financial economics, 3(4), 305-360.
Johnston, J. (2002). Tenure, promotion and executive remuneration. Applied Economics, 34(8), 993-997.
Johnston, J. (2005). Reward design and CEO succession in the UK. Applied Economics, 37(13), 1535-1541.
Larcker, D. & Tayan, B. (2011), Corporate Governance Matters: A Closer Look at Organizational Choices and Their Consequences. Pearson Education, Inc. Publishing as FT Press
Lipton, M., & Lorsch, J. W. (1992). A modest proposal for improved corporate governance. Business Lawyer, 48(1), 59.
McKnight, P., Tomkins, J., Weir, C., & Hobson, C. (2000). CEO Age and Top Executive Pay: A UK Empirical Study. Journal of Management and Governance, 4(3), 173-187.
Mohan, N., & Ruggiero, J. (2003). Compensation differences between male and female CEOs for publicly traded firms: A nonparametric analysis. Journal of the Operational
Research Society, 54(12): 1242-1248.
Morck, R., Shleifer, A., & Vishny, R. W. (1989). Alternative mechanisms for corporate control. The American Economic Review, 79(4), 842-852.
Murphy, K. (1986). Incentives, Learning, and Compensation: A Theoretical and Empirical
Investigation of Managerial Labor Contracts. The RAND Journal of Economics, 17(1), 59. Patton, A., & Baker, J. C. (1987). Why won't directors rock the boat? Harvard Business Review
, 65(6), 10-18.
Rechner, P. L., & Dalton D. R. (1989). The impact of CEO as board chairperson on corporate performance: Evidence vs. rhetoric. Academy of Management Executive, 3(2): 141–143. Sauerwald, S., Lin, Z., & Peng, M. (2016). Board social capital and excess CEO returns. Strategic
Tian, J., Haleblian, J. J., & Rajagopalan, N. (2011). The effects of board human and social capital on investor reactions to new CEO selection. Strategic Management Journal, 32(7), 731-747.
Tosi, H., Werner, S., Katz, J., & Gomez-Mejia, L. (2000). How Much Does Performance Matter? A Meta-Analysis of CEO Pay Studies. Journal of Management, 26(2), 301-339.
8. Appendix
Table 3: Variance Influence Factors of the independent variables
