The influence of CEO Compensation on investing in Corporate Social
Responsibility, a manifestation of agency problems?
Master’s thesis Finance
1
Abstract
This paper examines how CEO compensation structure affects a company’s level of corporate social responsibility (CSR) investment, looking at the S&P500 companies in 2008-2018. Traditional agency theory suggests that CEOs engage in CSR for personal interests at the expense of shareholders, while the good governance view sees CSR as a business strategy that maximizes firm value. The results of this paper support the good governance view as they show that the level of CSR is negatively associated with a higher share of nonequity-based compensation, while it is positively associated with the share of equity-based compensation. Furthermore, breaking down equity-based compensation shows that CSR is positively associated with stock-based compensation since stock-based compensation reduces agency problem through linking the real wealth of CEOs to firm performance. CSR is negatively associated with option-based compensation, providing further evidence for the good governance view as with option-based compensation personal benefits can be pursued at the cost of firm value as CEOs can opt to not exercise the option, not affecting their own wealth levels much. The results are robust to various CEO-specific and company-specific variables that explain the CSR investment levels as well, including the presence of a CSR committee, CEO age and gender.
2
1. Introduction
Over the last years, corporate social responsibility (CSR) has become a widely known concept for companies and investors. New technologies and new uses of media enables consumer to collect information about bad business practices and enables them to express themselves about it, forcing companies to see CSR practices as a necessity for successful business. A study by Cone Communications (2017) reveals that over 76% of American consumers will refuse to buy products or services of a company supporting an issue contrary to their beliefs. Furthermore, the study shows that 78% feels that companies should take responsibility in addressing important social justice issues. Despite the growing attention to CSR practices and the investment considerations in companies, research investigating whether being socially responsible improves the financial performance of the company gives mixed results. Van Beurden and Gössling (2008) find that the majority, 68%, of the included studies in their research has a positive relationship between CSR and corporate financial performance (CFP), while only six percent finds a negative relationship. A major reason for the mixed results is the inconsistency in the way research measures the relationship between CSR and CFP as there is no standard measure of both variables. Mixed evidence from prior studies raises the question why CEOs intend to engage in CSR. Classic agency theory introduced by Jensen and Meckling (1976) argues that CEOs investing in CSR tend to pursue their own interests rather than to maximize shareholder value. The good governance view argues that CSR is a value-maximizing activity: engaging in CSR practices could strengthen the relationship with various stakeholders which improves firm performance in turn (Cheng, Hong, and Shue, 2013).
3
compensation is seen with a higher level of CSR investments. With equity-based compensation, firm performance becomes a focus area of CEOs as their compensation is linked to it. Seeing a higher CSR investment when equity-based compensation is relatively higher gives evidence that CSR is a manifestation of good governance. Vice versa, if a CEO has mainly compensation that is nonequity-based, it allows him to pursue their own interests at the possible expense of shareholders as their compensation will not be affected significantly by unprofitable investments despite changes in firm performance, creating agency problems. If agency problems are relatively higher, through a higher proportion of nonequity-based compensation, CEOs are more inclined to invest in investments that benefit themselves personally. Thus, if CSR investments are found together with higher agency problems CSR investment can be seen as a manifestation of agency problems that is not a benefit to shareholder value per se, but a benefit to CEOs predominantly. This study differs from previous work on CSR in two ways. First, previous literature has mainly been focused on the relationship between CSR and financial performance. When looking at accounting numbers for financial performance (e.g. profit, return on assets), a ‘perfect’ relationship with CSR investment does not exist since these accounting numbers are influenced by many variables that cannot be fully accounted for, e.g. accounting practices. Similarly, when looking at market financials to measure financial performance (e.g. market capitalization, stock price), many external factors influence these numbers. To my knowledge, this is the first paper to look at the link between CSR performance and the structure of CEO compensation. By looking at the relationship between CSR and CEO compensation structures, I try to avoid the imperfect relationship between CSR and financial performance. By looking at CEO compensation, I derive the extent of agency problems through the alignment between shareholder interests and CEO interests. Alignments are, generally, higher when the percentage of CEO compensation that is linked to shareholder value is higher. The relationship between the level of alignment and the level of CSR investments will be used to determine whether CSR is a manifestation of agency problems or of good governance.
The second distinction is that I subdivide equity-based compensation instead of treating them as one type of compensation. Previous literature (e.g. Mehran, 1995) does not make any distinction between different types of equity-based compensation but acknowledges that the types may have different impacts on decision-making by CEOs (see section 2.1.). Hence, in this paper I make a distinction between option-based compensation and stock-based compensation.
The main research question is: Are CSR investments a manifestation of agency problems or of
4
First, I look at the proportion of CEO compensation that is nonbased as well as equity-based and examine the effects on CSR investments. If investing in CSR is seen with a high proportion of nonequity-based compensation, CSR investments can be seen as a manifestation of agency problems since agency problems are higher when CEO compensation is not linked to shareholder value. Similarly, if CSR investments are higher when equity-based compensation is higher, CSR investments can be seen as a manifestation of good governance instead, as agency problems are lower with equity-based compensation. Second, I will incorporate different CEO-specific and company-CEO-specific variables that may influence the level of CSR investments as well, such as board size, CEO age, and CEO gender. Third, I investigate whether the presence of CEO duality will strengthen the relationship between compensation schemes and CSR investments. With CEO duality, CEOs experience more power and may therefore be able to increase CSR investments further than in the situation where they do not possess the role of chairman as well. Finally, to avoid cultural differences between countries that may explain structural differences in compensation schemes, I will only look at companies from one country in my research, the USA. Because widely dispersed is a common feature in many U.S. firms, the subsequent relatively low managerial ownership makes equity-based compensation more important and effective than in many other countries. I will incorporate data from the S&P500 in the period 2008-2018.
5
2. Literature review
2.1. Agency theory and executive compensation
In 1976, Jensen and Meckling published an article that is seen as the basis of the agency theory, defining the agency relationship as “a contract under which one or more persons (the principal(s)) engage another person (the agent) to perform some service on their behalf which involves delegating control and decision-making authority to the agent.” The most famous corporate example is the relationship between shareholders and executives where shareholders hire executives to run the company on their behalf on a day-to-day basis. Agents can show self-serving behavior where they try to maximize private returns (e.g. empire-building, shirking) instead of maximizing the return on investment of the principal (e.g. profits, share price).
Executive compensation is not solely a means for attracting and retaining executives, but to motivate them as well. The positivist agency theory proposes that outcome-based compensation is an effective way to constrain opportunism by agents. These compensations forms will co-align the preferences of the agent with those of the principal since the rewards for both depend on the same actions (Eisenhardt, 1989).
6
executives with those of shareholders, resulting in an increase in firm performance. However, equity-based compensation is clustered in this study as it makes no distinction between different types of equity-based compensation. An important area of future research mentioned in the paper is to make a distinction in types of equity-based compensation as not all types are suitable for reducing agency problems and increasing firm performance.
Stock options could encourage riskier behavior by eliminating down-side risks for the executive since it fails to penalize executives when investments fail and firm value declines. Sanders (1999) finds that executives will choose to not exercise the option when the share price falls below the option price and, consequently, their current wealth would not change. So, unlike stock compensation, executives will not experience any immediate losses with option-based compensation even though shareholders could lose considerably. When executive compensation consists of stocks, failed investments and a subsequent decrease in stock price will result in a real reduction of executive wealth. With future research it thus is of importance to make a distinction between these two types of equity-based compensation
In short, existing literature discussed is in line with the agency theory by suggesting that incentive compensation is a better way to incentive executives to increase firm value compared to salaries and bonusses. However, differences in types of equity-based compensation exist, thus requiring the subject to be investigated further and more thoroughly than has been done so far.
2.2. CSR
7
managerial agency problems inside the firm. Previous literature suggests that managers engaging in CSR benefit themselves at the expense of the shareholders (Krüger, 2015). They, for example, earn a good reputation with key stakeholder such as local politicians or non-governmental organizations. Overall, CSR is not in the benefit of shareholders as it is not a value-maximizing investment under the agency view.
Results of research testing the two views remains mixed. Cheng et al (2013) find evidence consistent with the agency view. Looking at managerial ownership only, they find that measures of the company’s goodness decreases when managerial ownership increases. With a high level of managerial ownership, agency problems decline as managers have more incentive to increase firm performance than managers with low or zero ownership. The found decline in firm’s goodness shows that CSR is not value-maximizing but a manifestation of agency problems. Krüger (2015) finds evidence consistent with the agency view as well: investors react slightly negatively when positive news about a company’s CSR policies is revealed. Shareholders thus see CSR policies not as a benefit to themselves but as a benefit primarily to managers only, who, for example, earn a good reputation with their key stakeholders. A study by Hillman and Keim (2001) looks at the relationship between social issue participation and shareholder value creation, with social issue participation being defined as “using corporate resources to pursue social issues that are not directly related to the relationship with primary stakeholders”. They find a negative correlation between social issue participation and shareholder value creation. Finally, Jensen (2001) conclude that it is impossible for firms to maximize in more than one dimension; corporate social performance (CSP) and corporate financial performance (CFP) do not go hand in hand. Other papers find that CSR has positive effects on various company aspects. Lee and Faff (2009) find that firms that focus on CSP have significantly lower idiosyncratic risk. Goss and Roberts (2011) look at social responsibility concerns and find that banks with more concerns have a higher cost of debt capital. In other words, focusing on CSP and reducing the environmental concerns of your company declines the cost of debt capital. Furthermore, a study by El Ghoul, Guedhami, Kim, and Park (2018) provides evidence that the cost of equity capital is lower for firms with higher corporate environmental responsibility. Finally, Deng, Kang, and Low (2013) look at mergers to determine whether CSR creates value for shareholders. Their results are consistent with the good governance view as shareholders experienced a positive long-term stock return and mergers were less likely to fail than when acquirers had low CSR levels.
8
view, CSR is a business activity that improves firm value. When CEO compensation is equity-based, CEOs will be compensated for such a value-maximizing investment and therefore motivated to invest in such value-maximizing activities. With this good governance view, I hypothesize to find a positive relationship between the percentage of compensation that is equity-based and CSR investment levels, as they will be inclined to invest in value-maximizing activities with equity-based compensation.
H1a. If CSR is an investment that improves firm value, CSR performance is positively associated with the proportion of CEOs' equity-based compensation, while it is negatively associated with the proportion of CEOs’ nonequity-based compensation.
The second view, the agency view, suggests that CEOs invest in CSR to benefit themselves personally the cost of the firm. Barnea and Rubin (2010) find evidence supporting this view, their results show that investment numbers in CSR are higher when CEO have lower insider ownership. The fact that CSR investments are low when insider ownership is high supports the agency view by suggesting that, on average, each incremental dollar invested in CSR declines the firm value. The higher levels of CSR investments with lower CEO ownership suggests that CEOs tend to invest in CSR when they bear little costs for such activity due to low levels of ownership. With such a “CEO-friendly” compensation scheme, with low levels of CEO ownership, CEOs bear little personal risk since they have a high proportion of fixed, nonequity-based compensation and a low proportion of equity-based compensation that depends on performance. CEOs then are less inclined to maximize firm value and have a bigger possibility to pursue their own interests since their actions bears little consequences for their compensation levels. Under the agency view, CEOs engage in CSR is to strengthen their reputation as a “good” CEO and corporate citizen in the community and on the managerial labor market, to leave a personal legacy at the cost of the shareholders, or to pursue their personal values. (see, e.g., Barnea et al, 2010; Cespa and Cestone, 2007; Hemingway and Maclagan, 2004; Fabrizi, Mallin, and Michelon, 2014). Thus, to hypothesize:
9
Finally, a distinction will be made in this paper between two different equity-based compensations: option-based compensation and stock-based compensation. As mentioned in section 2.1., it is important to make a distinction between different types of equity compensation as it has different effects on the real wealth of CEOs (Sanders, 1999). Particularly, with option-based compensation, CEOs will not experience any immediate real wealth losses but opt to leave their options unexercised if investments turn out to be unprofitable. With stock-based compensation, CEOs do experience immediate real wealth losses. So, with stock-based compensation CEOs are more motivated to improve firm value as it impacts their wealth more. If CSR is an investment that improves firm value, CSR investment levels will be higher if stock-based compensation is higher. If CSR is an investment that does not improve firm value but improves CEO personal benefits, CSR investment levels will be lower if stock-based compensation is higher. With option-based compensation personal benefits can be pursued at the cost of firm value as CEOs can opt to not exercise the option, keeping their wealth level unaffected even though firm value is affected. If CSR is an investment that does not improve firm value but improves CEO personal benefits, CSR investment levels will be higher if option-based compensation is higher. To hypothesize:
H2a. If CSR is an investment that improves firm value, CSR performance is positively associated with the share of CEOs' stock-based compensation, while it is negatively associated with the share of CEOs’ option-based compensation.
H2b. If CSR is an investment that does not improve firm value but improves CEO personal benefits, CSR performance is positively associated with the share of CEOs' option-based compensation, while it is negatively associated with the share of CEOs' stock-based compensation.
2.3. Board structure
10
person that is both the CEO and chairman of the board of directors, as the contracts between the CEO and the other executives in the board are unlikely to maximize firm value as expected by shareholders. If hypothesis H1a holds and CEOs invest in CSR as it is an investment that maximizes firm value, having CEO duality will enable CEOs to engage in such as value-maximizing investment even further. Meanwhile, if CSR investments are in line with the agency view, if hypothesis H1b holds, having CEO duality and the subsequent increased CEO power will enable CEOs to engage in CSR for purely their own interests even further.
2.4. Company-specific and CEO-specific variables
Several characteristics regarding the company and the CEO may have an impact on the level of CSR investments. The first characteristic discussed is board size. Previous literature has found both positive and negative relationships between board size and firm performance. Some papers find that smaller boards are more effective in monitoring and reach consensus quicker (Jensen, 1993; Cheng, 2008; Lakhal, 2005; Ahmed, Hossain, and Adams, 2006) while other papers find smaller boards to have disadvantages as they have lower quality of advice due to undiversified experience and backgrounds, less ability to create links for gaining access to critical resources (Dalton, Daily, Johnson, and Elstrand, 1999; Bonn, 2004), and less efficiency in workload allocation and responsibilities distribution (Beiner, Drobetz, Schmid, and Zimmermann, 2004).
Regarding the effects of board size on CSR, Rao, Tilt, and Lester (2012) find that environmental reporting increases as board size increases. They mention that decisions regarding the content and extent of environmental disclosure require “intensive involvement, more unanimity, effective communication, and coordination by members”. Previous literature finds that these requirements are more likely to be found in smaller boards. In other words, it is hypothesized that CSR investment is higher in companies with smaller boards. Kassinis and Vafeas (2002) provide further support for the effectiveness of smaller boards as they find that larger boards are less effective in preventing environmental lawsuits. Further, they find that larger boards are less effective in preventing behavior that leads to environment-related lawsuits. Therefore, it is expected that CSR investment is lower in companies that have a larger board size.
A company-specific variable that may have a positive effect on CSR investment is the presence of a CSR sustainability committee. Having such a committee or team in place suggests that the company has an interest in engaging in CSR. Hence, it is expected that having a CSR sustainability committee has a positive relationship with the level of CSR investment.
11
12
3. Methodology
3.1. Variables
In order to test the hypotheses, several data inputs are collected. Detailed compensation data is needed, including data on salary, stock options, bonusses, and stocks. These data are needed to create several variables measuring the compensation schemes that will be used in the methodology of this research: (1) percentage of total compensation in grants of new stocks awarded, (2) percentage of total compensation in grants of new options awarded, (3) percentage of total compensation that is equity-based, (4) percentage of total compensation in salary plus bonusses (non-equity). Total compensation is the sum of salary, bonusses, other annual compensation, restricted stock grants, LTIP payouts, and value of options grants. Next, I obtain data regarding CEO-specific and company-specific variables: board size, the presence of a sustainability committee, the age and gender of the CEO, and the presence of CEO duality as these variables are hypothesized to affect the level of CSR investments. Finally, data regarding the control variables for size, the market capitalization of the company, time effects, the fiscal year, are collected, as well as the market-to-book-ratio and the return on assets (ROA).
3.2. Methodology
As described in the previous section, two hypotheses regarding the association between CSR investment and CEO compensation will be tested. If CEOs invest in CSR for their own benefits, this behavior will be reflected in the relationship between CSR investments and CEO compensation schemes. The compensation scheme will show constructions that are CEO-friendly to let them be able to pursue their own interests. This CEO-friendly package will consist of a high proportion of nonequity-based compensation and a low proportion of equity-based compensation. If CEOs were to invest in CSR for their own benefit, this compensation package will be seen with a high level of CSR investments. On the other hand, if CEOs invest in CSR as it is an investment that improves firm value, we will expect to see a positive relationship between CSR and a low proportion of nonequity-based compensation and a high proportion of equity-based compensation. The hypotheses are tested using ordinary least-squares (OLS) analysis, following previous research (see, e.g., Hillman et al, 2001; Mehran, 1995). For this analysis, I use the following regression of CSR investment on CEO compensation structures:
𝐶𝑆𝑅𝑖𝑡 = 𝛼 + 𝛽0𝑁𝑜𝑛𝑒𝑞 + 𝛽1𝐸𝑞 + 𝛽2𝑁𝑜𝑛𝑒𝑞 𝑥 𝐷𝑢𝑎𝑙 + 𝛽3𝐸𝑞 𝑥 𝐷𝑢𝑎𝑙 + 𝛽4𝑌 + 𝛽5𝑀𝑎𝑟𝑘𝐶𝑎𝑝
13
where the dependent variable, CSR, is the ESG score of company i in fiscal year t, which proxies for the CSR investment level. The ESG score measures how good a company scores on the environmental, social, and governance pillars, based on published information. The independent variables include: Noneq is the percentage of total compensation that is nonequity-based, such as salary and bonus, Eq is a vector of 2 that measures the percentage of total compensation that is equity-based, where a distinction could be made between stock-based compensation and option-based compensation.
Noneq x Dual and Eq x Dual are interaction variables that measure whether having duality, Dual,
strengthens the relationship between compensation and CSR, Y is a vector of 5 CEO-specific and company-specific variables discussed in section 2.4. that may influence the level of CSR investments. Finally, I will incorporate several control variables. The first variable, Size, measures market capitalization of the company at the end of the fiscal period and accounts for company size. Accounting for size has been done extensively in previous research because bigger firms have more possibilities to invest in CSR than smaller firms. Since the S&P500 includes big as well as relatively smaller companies, market capitalization will differ considerably and by taking the log of firm size to control for firm size cross-sectional comparisons between companies is enabled. The second control variable, Time, is a dummy variable for each year to account for time-fixed effects. As suggested by Schipper and Thompson (1981), time effects will need to be included in order to allow for changes in legal restrictions and market conditions during the sample period plus the recent growth in CSR investments that has been noticed, which will be discussed further in the next section. ROA is the return on assets, a profitability measure that is used as a control variable to see whether profitable firms invest more in CSR in a year than unprofitable firms, while
MTB, the market-to-book-ratio, predicts future profitability using market values.
Testing hypothesis 1, the primary focus is on the regressions coefficient of the compensation measures, β0 and β1. If β0 is positive, the results will support Hypothesis H1b, CSR is an
investment that does not improve firm value but improves CEO personal benefits. With nonequity-based compensation, CEOs can invest in CSR for their personal benefits as their compensation scheme will not affect their real wealth significantly since their compensation is more nonequity-based. As their actions have little impact on their compensation levels, CEOs may thus be less inclined to maximize firm value and are more inclined to pursue their own interests. Therefore, the positive coefficient shows that CSR investments are high when the proportion nonequity-based compensation is higher. Vice versa, if β0 is negative, results will provide
14
compensation next, if β1 is positive, the results will provide evidence in favor of Hypothesis H1a.
When CEO compensation is linked to firm performance through a relatively high proportion of equity-based compensation, CEOs have a “shareholder-friendly” compensation scheme since interests are more aligned with shareholders and consequently, agency problems will decline. They will be more inclined to engage in activities that increase firm value as their compensation is linked to it. The found positive relationship between CSR investments and a relatively higher proportion of equity-based compensation, a positive β1, are thus in line with the good governance view. If the value of β1 is negative, it will provide evidence for Hypothesis H1b as it
shows that CEOs do not invest in CSR if their compensation is linked to firm performance. It provides evidence to the agency view that CSR does not improve firm performance and since CEO compensation is linked to performance, they shy away from the investment. Now subdividing β1
into a stock-based part and option-based part, if the coefficient of the proportion of option-based compensation were positive, support for hypothesis H2b is found. Then, higher levels of ESG scores are in line with the agency view since, with option-based compensation, CEOs real wealth will not change with a decline in firm value allowing CEOs to engage in CSR for their personal interests rather than shareholder interests. They can opt to not exercise the stock option, which thus encourages self-interested behavior instead of encouraging to only invest in investments because they increase shareholder value. Looking at stock-based compensation, if the coefficient were positive, support for hypothesis H2a is found. With stock-based compensation, changes in firm value are directly translated into real changes in CEO wealth which encourages CEOs to focus on value-maximizing investments. A positive relationship between stock-based compensation and CSR investment suggests that CEOs see it as a value-enhancing investment.
15
4. Data and descriptive statistics
4.1. Measurement of variables
I collect data regarding executive compensation structures for S&P500 companies using the Execucomp database which covers over 12,500 executives of organizations, including the S&P500 for the period 2008 to 2018. The relevant variables that will be analyzed to determine agency problems are equity-based and based compensation. First, I collect the nonequity-based variable: total current compensation, which includes salary and bonus. Second, I will use three relevant equity-based variables; value of stock awards, value of options awards, and value of equity awards, which is the sum of stock awards and option awards. These numbers are the pay for each CEO in a fiscal year.
I will create four measures of compensation that will be used in the methodology of this research: (1) percentage of total compensation in grants of new stocks awarded, (2) percentage of total compensation in grants of new options awarded, (3) percentage of total compensation that is equity-based, and (4) percentage of total compensation in salary plus bonusses. Equity-based compensation is the sum of the grants of stock awards and options awards and total compensation is the sum of salary, bonusses, other annual compensation, restricted stock grants, LTIP payouts, option grants and all other types of compensation. Both compensation variables are available in Execucomp.
16
Further data concerning CEO-specific variables is collected using Execucomp, as it provides data about the age of the CEO and gender. I consult the attached information of executives to determine whether there is CEO duality. Finally, looking at company-specific variables, data regarding the presence of a CSR sustainability committee is provided by Thomson Reuters Eikon database. The last company-specific variable to test the hypotheses, board size, is collected from the Execucomp database.
Finally, market capitalization to control for company is size is collected using the Thomas Reuters Eikon database, while the control variables ROA and market-to-book-ratio are derived from Compustat. to measure current and past profitability as well as future profitability. This controls for the possibility that firms mainly invest in CSR if they are profitable (in the future).
4.2. Sample
The sample consists of executive compensation schemes and CSR investments (ESG scores) of the S&P500 between 2008 and 2018 in the USA. An advantage of using one country in research is that no fundamental differences in executive compensation between countries caused by cultural differences or differences in the level of development and legal systems must be accounted for. Furthermore, the advantage of looking at the USA is that in many other countries widely dispersed ownership is not as optimal since investors are not as well-protected. Therefore, in these settings, managers often already retain a high ownership in the company and having equity-based compensation may therefore be ineffective in reducing agency problems (La Porta, Lopez-de-Silanes, Shleifer, and Vishny, 1996). The advantage of the USA is that it is a good example of a country where shareholders are well-protected, and each firm is required to be transparent in their disclosures. These features contribute to widely dispersed ownership in many U.S. firms. Consequently, the relatively low managerial ownership makes contracts that include equity ownership more important than in other less protected countries and is, therefore, a good country to investigate in this paper. After collecting data and accounting for missing values, the final sample consists of 4,218 observations.
4.3. Sample characteristics
17
dollars, the average options awards for CEOs are worth 2.133 million US dollars. The share of non-equity compensation, salary and bonus, accounts to just below 13% of total compensation. Column 2 shows that salary and bonusses have lower mean values compared to equity-based compensation as well: 1.106 million US dollars and 0.218 million US dollars, respectively. To conclude, panel A of table 1 shows that most compensation is equity-based in the sample. However, with standard deviations being high compared to their mean values for nearly every compensation component, see column 4, compensation structures differ considerably between companies in the sample. A column depicting the composition of CEO compensation containing the values of compensation types as well as percentages of total compensation is provided in figure 1.
18 Table 1
Summary statistics
Panel A: Values of components of CEOs total compensation for the 2008-2018 data of the S&P500 in
salary plus bonus (non-equity awards), stock options awards, stock awards, and equity awards as well as the percental distribution of CEO’s total compensation among these components. All compensation variables are in thousand dollars. Total compensation is the sum of salary, bonusses, other annual compensation, restricted stock grants, LTIP payouts and value of options grants. Nonequity-based compensation is the sum of the values of salary and bonusses. Equity-based compensation is the sum of values of stock awards and option awards. The definitions of all variables are described in Appendix A.
Variables N Mean Median St. Dev. Min. Max.
Salary ($000s) 4,218 1,106 1,038 463.5 0 5,600 Bonus ($000s) 4,218 218.4 0 864.1 0 12,000 Stocks ($000s) 4,218 4,879 3,853 5,092 0 98,323 Options ($000s) 4,218 2,133 936 4,694 0 102,107 Non-equity compensation ($000s) 4,218 1,325 1,100 1,043 0 13,068 Total compensation ($000s) 4,218 10,804 8,934 8,977 0 156,078 Nonequity-based share (%) 4,218 16.71 12.54 14.84 0 100 Stock-based share (%) 4,218 50.90 43.04 168.5 0 7,815 Option-based share (%) 4,218 20.80 12.49 117.1 0 6,496 Equity-based share (%) 4,218 71.70 63.84 254.0 0 12,115
Panel B: Distributions of the level of CSR investments, the presence of a sustainability committee, the
presence of CEO duality size of the board of directors, CEO age and CEO gender for the average of 2008-2018 data of the S&P500. CEO duality is a dummy variable with value 1 if the situation of CEO duality is present in the company. Sustainability committee is a dummy variable taking the value of 1 if a committee is present in the company. Gender is a dummy variable with value 1 if the CEO is male, and 0 if female. The definitions of all variables are described in Appendix A.
Variables N Mean Median St. Dev. Min. Max.
ESG pillar score 4,218 47.99 44.25 16.45 8.621 94.27
Environmental score 4,218 59.40 62.88 22.90 4.770 99.09 Social score 4,218 61.93 63.32 19.28 4.904 99.03 Governance score 4,218 59.46 61.81 20.77 5.680 98.82 CSR strategy score 4,218 64.56 73.05 27.73 0.152 99.98 CSR committee 4,218 0.581 1 0.493 0 1 CEO duality 4,218 0.728 1 0.445 0 1 Board size 4,218 10.90 11 2.251 5 36 CEO age 4,218 56.88 57 6.196 29 82 CEO gender 4,218 0.965 1 0.185 0 1 Market capitalization (in $10 bln) 4,218 3.44 1.44 6.24 0.12 8.69 Market-to-book-ratio 4,218 6.02 2.75 74.12 -1,35 3,409
19
In table 2, ESG scores are shown separately for each year in the sample. It is clear that a time trend exists, scores are growing nearly every year and companies even see a substantial increase after 2014. The average ESG score increases from 46.41 in 2014 to 53.04 in 2015, an increase of over 14% while in the 7 years preceding the total growth was around 6.3%. Looking at the median values, ESG scores increase from 43.27 to 50.26 between 2014 and 2015. This supports the view that CSR is an area of interest gaining more attention in the board room in the last few years.
23% 22% 25% 27% 24% 28% 28% 24% 24% 27% 27% 11% 11% 11% 11% 11% 12% 12% 12% 12% 12% 12% 2% 2% 3% 3% 3% 2% 2% 2% 2% 1% 2% 35% 28% 39% 40% 46% 48% 58% 55% 62% 66% 22% 30% 22% 23% 25% 22% 23% 22% 18% 20% 17% 22% $0,00 $2.000,00 $4.000,00 $6.000,00 $8.000,00 $10.000,00 $12.000,00 $14.000,00 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 CE O comp en sat ion in $ th o u san d s Fiscal year
Figure 1: Composition of CEO compensation per year
20 Table 2
Breakdown of ESG scores by year for the average of 2008-2018 of the S&P500 companies.
ESG scores N Mean Median St. Dev. Min Max.
2008 306 43.65 41.35 13.50 17.71 88.05 2009 361 44.66 42.81 15.17 17.01 87.52 2010 395 43.12 40.73 14.65 15.12 86.03 2011 397 45.42 42.19 15.76 13.89 94.27 2012 396 45.27 42.33 15.78 8.62 89.58 2013 396 45.18 41.54 16.10 12.87 90.70 2014 409 46.41 43.27 15.56 13.22 91.43 2015 431 53.04 50.26 17.41 17.49 93.24 2016 431 54.13 50.51 17.07 18.40 91.89 2017 438 54.89 50.37 16.40 19.27 91.09 2018 258 49.75 45.21 16.39 19.58 88.92
21
22
Table 3
Pearson correlations among CSR proxies, CEO compensation, company-specific variables and CEO-specific variables. ‘***’, ‘**’, and ‘*’ represent significance at 1%, 5%, and 10% levels, respectively. The definitions of the variables are described in Appendix A.
Variables ESG score E score S score G score Salary Bonus Stocks Options Total compensati on Equity compensati on Nonequity compensati on CSR Committee
23 5. Results
5.1. CSR investment and CEO compensation structure
Table 4 presents the regressions of CEO compensation on CSR, with the dependent variable being the ESG score. In column 1, I regress the ESG score on the proportion of total compensation that is nonequity-based and equity-based. Results show that companies invest significantly lower in CSR if the proportion of compensation that is non-equity (salary plus bonus) increases. A 1% increase in the proportion that is nonequity-based declines the ESG score, the proxy for CSR investment, with 0.143. Looking at the proportion of CEO compensation that is equity-based, a 1% increase increases the ESG score with 0.001, but the change is statistically insignificant. These results show some evidence consistent with hypothesis 1a, where CSR performance is positively associated with the proportion of CEOs' equity-based compensation while it is negatively associated with the proportion of CEOs’ nonequity-based compensation. It suggests that CSR is more aligned with the good governance view, CEOs invest more in CSR if compensation-based agency problems are lower, through a higher equity-based compensation proportion and lower nonequity-based compensation proportion. With equity-based compensation, the value of CEO compensation depends on firm value and CEOs will be more encouraged to engage in value-enhancing practices, more than when their compensation is nonequity-based that is not connected to shareholder wealth.
24 Table 4
Ordinary least-squares estimates of ESG score on percentage of CEO’s nonbased compensation, equity-based compensation, the log of market capitalization, CEO-specific variables, and company-specific variables. Where equity-based compensation could be split up in stock-based compensation and option-based compensation. Interaction variables of CEO compensation and CEO duality are included. Sample is the S&P500 in 2008-2018. Robust standard errors are reported in parentheses. ‘***’, ‘**’, and ‘*’ represent significance at 1%, 5%, and 10% levels, respectively
Dependent variable: CSR strategy score
Independent variables (1) (2) (3) (4) (5) (6) Nonequity-based share -0.143*** -0.145*** -0.138*** -0.089*** -0.068*** -0.111*** (0.015) (0.015) (0.015) (0.015) (0.014) (0.024) Equity-based share 0.001 (0.001) Stock-based share 0.005 0.005 0.000 0.001 -0.002 (0.005) (0.005) (0.002) (0.001) (0.006) Option-based share -0.005 -0.005 0.001 0.000 0.002 (0.004) (0.004) (0.003) (0.002) (0.006) ln(Market cap) 0.878*** 0.648** -1.176*** -1.147*** (0.202) (0.263) (0.294) (0.294) Market-to-book ratio -0.008*** -0.008*** -0.008*** (0.002) (0.002) (0.002) ROA 6.615* 10.629*** 10.532*** (3.668) (3.640) (3.630) CSR Committee 10.299*** 10.314*** (0.611) (0.611) Board size 0.053 0.048 (0.124) (0.125) CEO duality 0.054 -1.161 (0.594) (0.879) CEO age 0.086** 0.085** (0.043) (0.043) CEO gender 0.553 0.560 (1.371) (1.372) Nonequity-based share x CEO
duality
0.062** (0.029) Stock-based share x CEO
duality
0.003 (0.006) Option-based share x CEO
duality
25
governance view of CSR further by implying that ESG scores increase if the degree of compensation agency problems are lower. In other words, CEOs increase CSR investment when their real wealth is linked to firm value through a high proportion of stock-based compensation. However, both coefficients are not significant at any level so these results do not provide significant evidence for hypothesis H2a, The signs of the proportion of compensation that is nonequity-based remains negative and highly significant with a value of -0.145, providing further evidence supporting hypothesis H1a. In column 3 the log of market capitalization is incorporated, as bigger companies are hypothesized to have more assets at their disposal to invest in CSR. The size variable is positive and significant at the 1% level with a value of 0.878, indicating that bigger companies indeed have a higher ESG score. The firm level control variables market-to-book-ratio and ROA are incorporated in column 4 and impact the ESG score significantly with values of -0.008 and 6.615, respectively. The positive coefficient of ROA shows that companies that are profitable engage significantly more in CSR than unprofitable companies. The negative coefficient of the market-to-book-ratio is striking as it shows that ESG scores decline if future profitability is expected to be positive. The proportion of nonequity-based compensation remains significant, where a 1% increase in the percentage of CEO compensation that is nonequity-based decreases the CSR score with 0.138. In column 4 time-fixed effects and industry-fixed effects are incorporated in the model as well. It does improve the model but has no different effect on the significance and signs of all coefficients compared to column 3.
5.2. CSR investment and company-specific and CEO-specific variables
In column 5, several company-specific and CEO-specific variables are included that could potentially affect the level of ESG scores. With this extended model, the coefficient of the proportion of compensation that is non-equity is -0.068 and keeps its significance, although the absolute value has declined compared to the previous models. The coefficients of the equity-based compensation variables remain insignificant with values of 0.001 and 0.000. Looking at the company-specific variables, we see some significant effects on the ESG score. First, the board size has a positive effect, if board size increases by 1, ESG score goes up by 0.053. These results are striking, however, as previous literature suggested that the characteristics of a smaller board fit
Constant 50.307*** 50.271*** 29.491*** 30.408*** 57.691*** 57.909*** (0.380) (0.389) (4.772) (7.951) (8.516) (8.529)
Time-fixed effects No No No Yes Yes Yes
Industry-fixed effects No No No Yes Yes Yes
Observations 4,218 4,218 4,218 4,218 4,218 4,218
26
more with having higher CSR investments. But no inferences can be made as the coefficient is not significant at any level. The next company-specific variable, the presence of a CSR sustainability committee, has a significantly positive effect on ESG score. If a committee is present in the company, CSR investments increase with 10.299, significant at the 1% level, supporting the hypothesis that having a CSR sustainability committee has a positive effect on CSR investment. Next the CEO-specific variables are discussed. Results in column 5 show that the gender dummy variable has a positive effect on the ESG score. Since the variable takes a value of 1 if the CEO is male and 0 if the CEO is female, it shows evidence inconsistent with previous literature that say females are more inclined to invest in CSR than males. If the CEO is male, average CSR strategy scores is 0.553 higher than in companies where the CEO is female. However, the coefficient is not significant at any level. The second CEO-specific variable, executives age, has a positive coefficient of 0.086 that is significant at the 5% level. It suggests that companies with older CEOs have higher ESG scores than companies with relatively younger CEOs, which provides no support to existing literature. Finally, to test whether CEO duality strengthens the relationship between CSR investment and CEO compensation, I will look at the duality variable and the interaction variables with the compensation variables. The coefficient of the duality variable in column 5 is insignificant with a value of 0.054 so the increase in CEO power does not causes CEOs to have, on average, a higher ESG score. Adding the interaction variables in column 6, the extended model does not provide additional explanation for the ESG scores. The interaction variables have absolute values that are very close to zero, with only one variable that has a significance of 10% (0.062), the interaction variable with the share of nonequity-based compensation. Therefore, these results provide little evidence support the hypothesis regarding CEO duality, the negative association between CSR and nonequity-based compensation and the positive association between CSR and equity-based compensation does not become more pronounced with CEO duality.
5.3. Robustness checks
27
not found in table 4 when the ESG score was used, both the option-based coefficient and the stock-based coefficient were not significant. Dissecting the ESG score into the three pillars does find evidence for hypothesis H2a. Appendix D, looking at the environmental score, shows that the coefficient of option-based compensation is positive and significant at the 5% level in column 5, with a value of -0.004. In other words, if option-based compensation increases with 1%, the environmental score of the company declines with 0.004. This negative relationship provides evidence for hypothesis H2a as this result is aligned with the good governance view. When compensation agency problems increase through an increase in option-based compensation, the environmental score declines. Looking at the social governance score, we find further evidence for hypothesis H2a, as the coefficient of stock-based (option-based) compensation is significantly positive (negative) at the 1% level with a value of 0.005 (-0.005). These coefficients strengthen the view of good governance, by showing that the social governance scores are higher when the agency problems are lower, that is, with a higher proportion of stock-based compensation and a lower proportion of option-based compensation.
In conclusion, aggregating the three pillars in one score, ESG score, some significant and interesting effects of compensation and CEO-specific and company-specific variables on the score cancel out. However, because of the small coefficients, economic significance is to be questioned.
28
lower for younger CEOs if CEO compensation consists of nonequity-based compensation. Following hypothesis H1a, a negative association with nonequity-based compensation for both age groups suggests that engagements in CSR are lower when consumption-based agency problems are higher.
Table 5
Ordinary least-squares estimates of CSR strategy score on percentage of CEO’s nonequity-based compensation and equity-based compensation. Where equity-based compensation is split up in stock-based compensation and option-stock-based compensation. Sample contains the S&P500 in 2008-2018 and is split in 2 CEO age groups, Q1 contains ages up to 57, and Q2 ages 57 and higher. Robust standard errors are reported in parentheses. ‘***’, ‘**’, and ‘*’ represent significance at 1%, 5%, and 10% levels, respectively.
Dependent variable: CSR strategy score
Independent variables Q1 Q2 Nonequity-based share -0.086*** -0.084*** (0.018) (0.025) Stock-based share -0.000 0.005 (0.002) (0.006) Option-based share 0.008 -0.004 (0.018) (0.005) ln(Market cap) 0.671* 0.173 (0.360) (0.425) Market-to-book ratio -0.011 -0.007*** (0.013) (0.002) ROA 5.967 12.922** (5.195) (5.850) Constant 36.827*** 71.081*** (9.752) (0.940)
Time-fixed effects Yes Yes
Industry-fixed effects Yes Yes
Observations 2,249 1,969
Adjusted R-squared 0.300 0.283
29
and -0.004 respectively, but are not significant at any level. The results are striking that the coefficients of older CEOs are not in line with previous literature. Most evidence here seems to be against the hypothesis that younger CEOs are more engaged with CSR compared to older CEOs.
30
using different proxies for CSR investment. However, there exist differences in the significance of those coefficients, where the effects of stock-based and option-based compensation are insignificant when ESG scores are used as a proxy.
With using another proxy in table 6, what is striking is that some CEO-specific and company-specific variables have gained significance. With the ESG score as a proxy in table 4, board size, the presence of a CSR committee and CEO age had a significant effect on the score. The presence of a CSR sustainability committee still has a positive effect on the CSR strategy score, with a coefficient of 35.181 (see table 6, column 5). The coefficient of board size is still positive, but the coefficient is now significant at the 1% level with a value of 1.044, indicating that increasing the board size by 1, CSR strategy score increases, on average, with approximately 1. This contrasts previous literature as smaller boards have characteristics that match those of CSR investments. The dummy variable CEO gender now has a significant positive effect, meaning that CSR strategy score is 2.467 higher when the CEO is male. Table 4 reports an insignificant effect of CEO gender on ESG score. CEO duality still reports a positive effect on the proxy of CSR investment, but is now significant at the 5% level with a coefficient of 1.158, compared to an insignificant coefficient of 0.054 when ESG score is used as a proxy. The coefficient of CEO age now loses its significance compared to table 4.
The effects of CEO-specific and company-specific variables do not differ considerably between the outcomes of the regressions, all variables have a positive effect on the CSR investment proxy that is used, but some differences in significance exist.
Table 6
Ordinary least-squares estimates of the CSR strategy score on percentage of CEO’s nonequity-based compensation, equity-based compensation, the log of market capitalization, CEO-specific variables, and company-specific variables. Where equity-based compensation could be split up in stock-based compensation and option-based compensation. Interaction variables of CEO compensation and CEO duality are included. Sample is the S&P500 in 2008-2018. Robust standard errors are reported in parentheses. ‘***’, ‘**’, and ‘*’ represent significance at 1%, 5%, and 10% levels, respectively.
Dependent variable: CSR strategy score
Independent variables (1) (2) (3) (4) (5) (6)
31
Finally, as Mehran (1995) concludes, the structure of compensation has a greater impact on motivation of executives to improve firm value than the level of the compensation. As seen in appendix I, the level of different compensation components are (close to) zero or insignificant. For example, the total level of consumption has no effect on the ESG score with an insignificant coefficient of 0. It seems that the level of CSR is mostly affected by the structure of CEO compensation as well, rather than the level of compensation.
(0.001) Stock-based share 0.012 0.009 0.002 0.003** 0.004 (0.008) (0.006) (0.004) (0.001) (0.009) Option-based share -0.011 -0.010* -0.003 -0.003* -0.003 (0.008) (0.006) (0.004) (0.002) (0.008) ln(Market cap) 10.618*** 10.240*** 3.216*** 3.197*** (0.314) (0.410) (0.326) (0.324) Market-to-book ratio -0.001 -0.000 -0.000 (0.003) (0.002) (0.002) ROA -22.371*** -4.925 -4.791 (5.773) (3.662) (3.681) CSR Committee 35.181*** 35.118*** (0.680) (0.683) Board size 1.044*** 1.048*** (0.145) (0.145) CEO duality 1.158* 2.954*** (0.596) (0.935) CEO age 0.056 0.053 (0.045) (0.045) CEO gender 2.467* 2.527* (1.324) (1.325) Nonequity-based share x CEO duality -0.068** (0.034) Stock-based share x CEO duality -0.001 (0.009) Option-based share x CEO duality -0.033** (0.016) Constant 68.757*** 68.670*** -182.643*** -150.671*** -34.780*** -34.434*** (0.651) (0.683) (7.534) (9.700) (7.773) (7.758)
Time-fixed effects No No No Yes Yes Yes
Observations 4,217 4,217 4,217 4,217 4,217 4,217
32
6. Conclusion
This paper investigates how the structure of CEO compensation is associated with CSR investments in a company. Research finds compensation structures to be a way to mitigate the conflict of interest between managers and shareholders, called agency problems. One structure that mitigates these conflicts is to link CEO compensation to firm performance. I find that firms that have higher investments in CSR tend to have a high proportion of equity-based compensation and a low proportion of nonequity-based compensation, suggesting that CEOs invest in CSR when compensation-based agency problems are lower. These results are in line with existing literature that view CSR as a manifestation of good governance. As these results suggest that social activities of the company are positively linked to CEO compensation structures that induce CEOs to maximize firm value by linking their compensation to firm performance, it shows that CEOs investments are a value-maximizing activity for the company. Further looking into equity-based compensation and subdividing it, reveals that CSR is positively associated with stock-based compensation and negatively associated with option-based compensation. These results provide further evidence for hypothesis H2a, supporting the good governance view that investments in CSR are increasing shareholder value. With stock-based compensation CEOs are more motivated to increase shareholder value than with option-based compensation, option-based compensation allows CEOs to invest in personal benefits that may destroy shareholder value without decreasing their own wealth in the process. Furthermore, when segregating the ESG score, the proxy for CSR investment, more evidence in favor of the good governance view is found than with the general ESG score. Where the positive (negative) effect of stock-based (option-based) compensation is insignificant when the ESG score is used as a proxy, these effects do become significant when looking at the environmental and social governance score separately.
33
CSR investment and is found to be statistically significant. After splitting the total sample in different age groups as a robustness check, results still show that older CEOs do invest more in CSR, which contradicts existing literature. After controlling for all CEO-specific and company-specific variables, all associations between CEO compensation and the CSR proxy remain robust. Finally, the large, positive coefficient of ROA is an important determent of the ESG score, suggesting that engaging in CSR is higher when firms are profitable.
Looking at the coefficients of the different compensation variables, some have little economic significance with values close to zero or insignificant values, particularly the equity-based compensation coefficients. It remains questioned whether CEO compensation schemes are a crucial influence on investment decisions, in this case CSR investment decisions. CEO-specific and company-specific variables, however, do seem to have a significant influence on this decision. As an example, the presence of a CSR sustainability committee is accountable for an ESG score of about 10, which is substantial considering the average ESG score is 47.99. However, it is to be noted that the installation of such a committee is a decision which often is influenced by the CEO. By using the CSR strategy score as dependent variable, I check the robustness of the relationship between compensation constructions and a proxy for CSR investment. While most signs of the compensation-based coefficients remain the same, the equity-based signs now gain significance. Stock-based compensations remains to have a positive effect on the CSR proxy, but is now significant at the 5% level, compared to when ESG scores are used as a proxy. Option-based compensation becomes significant too when CSR strategy scores are used. These outcomes with the new proxy strengthen the evidence for hypothesis H2a further. The significance of nonequity-based compensation remains present with this robustness check. In short, many (in)significant coefficients are not robust when other proxies for CSR investments are used.
34
characteristics that emphasize a long-term view of the CEO, which seems to be suited for CSR investments as these investments often contain longer-term aspects.
35
Reference list
Ahmed, K., Hossain, M., & Adams, M. B. (2006). The effects of board composition and board size on the informativeness of annual accounting earnings. Corporate governance: an international review, 14(5), 418-431.
Barnea, A., & Rubin, A. (2010). Corporate social responsibility as a conflict between shareholders. Journal of business ethics, 97(1), 71-86.
Bear, S., Rahman, N., & Post, C. (2010). The impact of board diversity and gender composition on corporate social responsibility and firm reputation. Journal of Business Ethics, 97(2), 207-221. Beiner, S., Drobetz, W., Schmid, F., & Zimmermann, H. (2004). Is board size an independent corporate governance mechanism?. Kyklos, 57(3), 327-356.
Bonn, I. (2004). Board structure and firm performance: Evidence from Australia. Journal of Management & Organization, 10(1), 14-24.
Byron, K., & Post, C. (2016). Women on boards of directors and corporate social performance: A meta‐analysis. Corporate Governance: An International Review, 24(4), 428-442.
Carlsson, G., & Karlsson, K. (1970). Age, cohorts and the generation of generations. American Sociological Review, 710-718.
Cespa, G., & Cestone, G. (2007). Corporate social responsibility and managerial entrenchment. Journal of Economics & Management Strategy, 16(3), 741-771.
Cheng, S. (2008). Board size and the variability of corporate performance. Journal of financial economics, 87(1), 157-176.
Cheng, I. H., Hong, H., & Shue, K. (2013). Do managers do good with other people's money? (No. w19432). National Bureau of Economic Research.
Cone Communications (2017). 2017 Cone Communications CSR study. Retrieved from https://www.conecomm.com/research-blog/2017-csr-study.
Dalton, D. R., Daily, C. M., Johnson, J. L., & Ellstrand, A. E. (1999). Number of directors and financial performance: A meta-analysis. Academy of Management journal, 42(6), 674-686.
Deng, X., Kang, J. K., & Low, B. S. (2013). Corporate social responsibility and stakeholder value maximization: Evidence from mergers. Journal of financial Economics, 110(1), 87-109.
Eisenhardt, K. M. (1989). Agency theory: An assessment and review. Academy of management review, 14(1), 57-74.
36
Fabrizi, M., Mallin, C., & Michelon, G. (2014). The role of CEO’s personal incentives in driving corporate social responsibility. Journal of Business Ethics, 124(2), 311-326.
Fernandez-Feijoo, B., Romero, S., & Ruiz, S. (2012). Does board gender composition affect corporate social responsibility reporting. International Journal of Business and Social Science, 3(1), 31-38.
Fukukawa, K., Shafer, W. E., & Lee, G. M. (2007). Values and attitudes toward social and environmental accountability: a study of MBA students. Journal of Business Ethics, 71(4), 381-394.
Goss, A., & Roberts, G. S. (2011). The impact of corporate social responsibility on the cost of bank loans. Journal of Banking & Finance, 35(7), 1794-1810.
Hemingway, C. A., & Maclagan, P. W. (2004). Managers' personal values as drivers of corporate social responsibility. Journal of Business Ethics, 50(1), 33-44.
Hillman, A. J., & Keim, G. D. (2001). Shareholder value, stakeholder management, and social issues: what's the bottom line?. Strategic management journal, 22(2), 125-139.
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. (2001). Value maximization, stakeholder theory, and the corporate objective function. Journal of applied corporate finance, 14(3), 8-21.
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.
Jensen, M. C., & Murphy, K. J. (1990). CEO incentives—It's not how much you pay, but how. Journal of Applied Corporate Finance, 3(3), 36-49.
Kassinis, G., & Vafeas, N. (2002). Corporate boards and outside stakeholders as determinants of environmental litigation. Strategic management journal, 23(5), 399-415.
Krüger, P. (2015). Corporate goodness and shareholder wealth. Journal of financial economics, 115(2), 304-329.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (1998). Law and finance. Journal of political economy, 106(6), 1113-1155.
Lakhal, F. (2005). Voluntary earnings disclosures and corporate governance: Evidence from France. Review of Accounting and Finance, 4(3), 64-85.
Lee, D. D., & Faff, R. W. (2009). Corporate sustainability performance and idiosyncratic risk: A global perspective. Financial Review, 44(2), 213-237.
37
Liao, L., Luo, L., & Tang, Q. (2015). Gender diversity, board independence, environmental committee and greenhouse gas disclosure. The British Accounting Review, 47(4), 409-424. Mehran, H. (1995). Executive compensation structure, ownership, and firm performance. Journal of financial economics, 38(2), 163-184.
Rao, K. K., Tilt, C. A., & Lester, L. H. (2012). Corporate governance and environmental reporting: an Australian study. Corporate Governance: The international journal of business in society, 12(2), 143-163.
Sanders, W.G. (1999). Incentive structure of CEO stock option pay and stock ownership: the moderating effects of firm risk. Managerial Finance, 25(10), 61-75.
Schipper, K., & Thompson, R. (1981). Common stocks as hedges against shifts in the consumption or investment opportunity set. Journal of Business, 305-328.
Vroom, V. H., & Pahl, B. (1971). Relationship between age and risk taking among managers. Journal of applied psychology, 55(5), 399.
38
Appendix
Appendix A. Variables and definitions used in the regressions
Variables Definitions
% of compensation in current compensation (nonequity-based share)
Percentage of total CEO compensation in a reported period that is awarded in current compensation
% of compensation in equity awards (equity-based share)
Percentage of total CEO compensation in a reported period that is awarded in equity awards
% of compensation in options awards (option-based share)
Percentage of total CEO compensation in a reported period that is awarded in option awards
% of compensation in stock awards (stock-based share)
Percentage of total CEO compensation in a reported period that is awarded in stock awards
Board size The number of directors in the company’s board Bonus awards
(Bonus)
CEO compensation in a reported period that is awarded in bonusses
CEO age The age of the CEO
CEO duality The situation when the CEO also holds the position of the chairman of the board
CEO gender A dummy variable taking the value of 1 if the CEO is male and 0 if female CSR investment See CSR strategy score
CSR strategy score A score ranging from 0 to 100 that reflects a company’s practices to communicate that it integrates the economic, social, and environmental dimensions into its day-to-day decision-making processes, a proxy for CSR investment
CSR sustainability committee (CSR committee)
A dummy variable taking the value of 1 if the company has a CSR committee or team and 0 if the company does not have such a committee Current compensation CEO compensation in a reported period that is awarded in salary and
bonusses Environmental
expenditures
Total amount of environmental investment and expenditures for environmental protections or to prevent, reduce, control environmental aspects, impacts, and hazards
Equity awards (Equity)
CEO compensation in a reported period that is awarded in options and stocks
Environmental score (E score)
A score measuring a company’s impact on living and non-living natural systems, including the air, land and water, as well as complete ecosystems. ESG score An overall company score based on the reported information in the
environmental, social and corporate governance pillars. Governance score
(G score)
39
Market capitalization (Market cap)
The value of a company that is traded on the stock market, calculated by the total number of shares times the current share price
Option awards (Option)
Fair value of all options awarded during the year to the CEO.
Valuation is based upon the grant-date fair value as detailed in FAS 123R Salary awards
(Salary)
CEO compensation in a reported period that is awarded in salary Social score
(S score)
A score measuring a company’s capacity to generate trust and loyalty with its workforce, customers and society, through its use of best management practices.
Stock awards (Stock)
Value of stock-related awards to the CEO. Valuation is based upon the value of shares that vested during the year as detailed in FAS123R. Total compensation The sum of salary, bonusses, other annual compensation, restricted stock
grants, LTIP payouts, and value of options grants
Appendix B. White test for the presence for heteroskedasticity using equation 1 with ESG score as the dependent variable. As the probability equals zero, it is rejected at all levels of significance, indicating that heteroskedasticity is present in the residuals.
White's test for Ho: homoskedasticity against Ha: unrestricted heteroskedasticity chi2(251) = 474.54
Prob > chi2 = 0.0000
Appendix C. Skewness/Kurtosis test for Normality using the residuals of equation 1 with ESG score as the dependent variable. As the probability is equal to 0, the null hypothesis of a normal distribution is rejected at all significance levels, indicating that the residuals are not normally distributed.
Variable Obs Pr(Skewness) Pr(Kurtosis) adj_chi2(2) Prob>chi2
40
Appendix D
Ordinary least-squares estimates of the environmental score on percentage of CEO’s nonequity-based compensation, equity-based compensation, the log of market capitalization, ROA, market-to-book-ratio, CEO-specific variables, and company-CEO-specific variables. Where equity-based compensation could be split up in stock-based compensation and option-stock-based compensation. Interaction variables of CEO compensation and CEO duality are included. Sample is the S&P500 in 2008-2018. Robust standard errors are reported in parentheses. ‘***’, ‘**’, and ‘*’ represent significance at 1%, 5%, and 10% levels, respectively
Dependent variable: Environmental score
Independent variables (1) (2) (3) (4) (5) (6) Nonequity-based share -0.218*** -0.220*** -0.137*** -0.100*** -0.055*** -0.106*** (0.029) (0.029) (0.021) (0.019) (0.016) (0.033) Equity-based share 0.001 (0.001) Stock-based share 0.006 0.003 -0.001 -0.000 -0.011 (0.006) (0.004) (0.002) (0.001) (0.007) Option-based share -0.008 -0.007* -0.004 -0.004** 0.006 (0.006) (0.004) (0.002) (0.002) (0.006) ln(Market cap) 10.680*** 10.436*** 6.535*** 6.586*** (0.249) (0.328) (0.337) (0.337) Market-to-book-ratio -0.003 -0.003 -0.003 (0.002) (0.002) (0.002) ROA -21.528*** -11.911*** -12.096*** (4.724) (3.952) (3.948) CSR Committee 18.906*** 18.913*** (0.683) (0.683) Board size 0.708*** 0.697*** (0.145) (0.145) CEO duality 0.582 -0.718 (0.652) (0.962) CEO age 0.071 0.067 (0.044) (0.044) CEO gender -0.129 -0.061 (1.293) (1.297) Nonequity-based
share x CEO duality
41 Option-based share x CEO duality -0.022 (0.016) Constant 65.011*** 64.982*** -148.359*** -143.441*** -92.752*** -92.696*** (0.478) (0.483) (5.516) (7.455) (7.549) (7.549)
Time-fixed effects No No No Yes Yes Yes
Industry-fixed effects No No No Yes Yes Yes
Observations 4,218 4,218 4,218 4,218 4,218 4,218
Adjusted R-squared 0.022 0.023 0.279 0.508 0.593 0.593
Appendix E
Ordinary least-squares estimates of the social score on percentage of CEO’s nonequity-based compensation, equity-based compensation, the log of market capitalization, ROA, market-to-book-ratio, CEO-specific variables, and company-specific variables. Where equity-based compensation could be split up in stock-based compensation and option-based compensation. Interaction variables of CEO compensation and CEO duality are included. Sample is the S&P500 in 2008-2018. Robust standard errors are reported in parentheses. ‘***’, ‘**’, and ‘*’ represent significance at 1%, 5%, and 10% levels, respectively
Dependent variable: Social score
