The influence of executive compensation on company reputation ranking

(1)

The influence of Executive Compensation on company

Reputation Ranking

Abstract

This study predicts that the level of CEO pay is able to predict both peer-based and consumer-based reputation ranking of US based firms. Also, this relationship might be moderated by the concentration of Fortune 500 firms within the same US state, and the political orientation of the US state of origin. Using a sample of Fortune 500 companies, a correlation and regression analysis found that there was a significant relationship between CEO pay and consumer-based reputation ranking and no significant relationship between CEO pay and peer-based reputation ranking. The peer-based study did find proof for a significant relationship between company profitability, size and reputation ranking. No evidence for moderating effects was found. Future research should use primary data.

Business Administration Strategy | MSc Thesis word count 14.295

Marloes den Os 10175563

Supervisor: Dr. D.A. Waeger June 22, 2017

Subject: Internal and External Drivers of the Environmental, Social and Corporate Governance (ESG) Performance of Firms

(2)

Statement of Originality

This document is written by Student Marloes den Os, who declares to take full responsibility

for contents of this document.

I declare that the text and the work presented in this document are original and that no

sources other than those mentioned in the text and its references have been used creating it.

The Amsterdam Business School is responsible solely for the supervision of completion of the

work, not for the contents.

(3)

Abstract……….………...………...1

Introduction.……….……….………..…...5

1.1 Executive Compensation………..…..6

1.2 Performance and Reputation………..….7

1.3 Research Gap………..….8

1.4 Academic and Managerial Relevance………..11

1.5 Structure………..…12

Theoretical Framework………….………….………..…..………...13

2.1 Executive Compensation………..………13

2.1.1 Agency Theory………..….14

2.1.2 The Motivational Assumption………..………..15

2.2 Company Reputation Ranking………17

2.2.1 Reputation and Reputational Risk………..…….18

2.2.2 Executive Compensation and Reputation………..20

2.3 Moderating Effects………21

2.3.1 US State………22

2.3.2 Political Orientation……..……….………24

2.4 Conceptual Framework………..………26

Methodology……….………..27

3.1 Research Design………..27

3.2 Research Sample……….28

3.3 Variables………29

3.3.1 Dependent Variables……….29

3.3.2 Independent Variables….………33

3.3.3 Moderation Variables………34

3.3.4 Control Variables………..34

3.4 Statistical Procedure……….35

(4)

Results……….……….36

4.1 Sample Statistics……….…………36

4.2 Data Statistics………38

4.3 Correlation Results………39

4.4 Regression Results……….42

4.4.1 Peer- and Consumer-based Reputation Rankings……….42

4.4.2 Moderation by HQs within same US state………….……….43

4.4.3 Moderation by Republican or Democratic orientation..………46

Discussion………..48

5.1 General Discussion………48

5.2 Hypotheses...……….48

5.2.1 Company Concentration and Political Orientation………..49

5.3 Theoretical Implications……….50

5.4 Managerial Implications………51

5.5 Limitations and Future Research.……….………..52

Conclusion……….54

References……….55

(5)

Introduction

The role of executive compensation and how it affects firm performance has played an important role in the field of financial economics the past decades. In the 1990s the number of academic papers on CEO pay has even outpaced the remarkable increase of CEO pay itself (Murphy, 1999). The reason for the substantial interest on the topic of CEO pay can be found in many publications by journalists and research papers on the topic of CEO compensation. Examples of such articles are publications such as: CEOs Get Paid Too Much (HBR, 2014), The Highest-Paid CEOs Are the Worst Performers (Forbes, 2014), and Attacking the Corporate Gravy Train (The Economist, 2009). In the HBR article, a study is discussed that was done on a global scale asking workers from all levels within organizations how high CEO pay should be. The conclusion in the article is that the ideal pay gap between an average worker pay and CEO compensation should be smaller. The Forbes article states that the highest paid CEOs are the worst performers. Based on a study by Professor Cooper from the University of Utah, Forbes states there was a significant negative relationship between CEOs who were paid more and the lower financial results they delivered to their shareholders, compared to their peers who were paid less. On the other hand, the article by the Economist invalidates some of the critique on the height of CEO pay, but does agree with Forbes that in some cases pay-systems rewards CEOs for failure. The conclusion that can be withdrawn from the previous mentioned articles is that CEO compensation is an important factor that is looked at both in terms of how companies are performing internally as well as externally. In the academic world, CEO compensation is seen as an influence on firm performance in many ways (Wade et at., 1997). But on the other hand, published articles like this are mainly influencing how readers look at a company and evaluate its reputation. In other words, the amount, frequency and content of CEO compensation may influence how a company reputation is evaluated. An example of how CEO salary may influence company reputation, is found with acclaimed CEO of Tesla Elon Musk, who didn’t accept salary in 2014 (Fortune, 2015)

(6)

1.1 Executive Compensation

The underlying principle determining the amount, frequency and content of CEO compensation, is the relationship between a CEO and the board of directors of the company the CEO works for. The relationship between the boards of publicly listed firms and their CEOs is best described as an agency relationship. Barney and Hesterly (1996) state that agency relationships occur whenever one party to a transaction, the principal, delegates either work or authority over work, to the other party in the transaction. The other party is the agent and the welfare of the principal is affected by the choices of this agent. According to agency theory contracts need to motivate the agent to pursue the principal’s objectives. Also a contracts should share risk between the agent and the principal, since agency theory depicts that the agent is more risk averse than the principal (Eisenhardt, 1989). In other words, CEO pay should be high enough that the CEO’s objectives are aligned with the goals of the board, and that CEO pay should divide the risk between the two parties. This ensures that the CEO is willing to take on more risk when the boards would want that to happen. The agency relationship and accompanying issue of compensation packages awarded to the CEO’s of publicly traded firms is a matter that many scholars have tried to tackle over the past years (Milbourn, 2003). Research on the agency relationship and pay-for-performance principle dates back quite a few years. It dates back to one of the first publications about the relationship between level of CEO pay and firm performance. This was a study by Jensen and Murphy that was published in 1990. They were among the first to examine the predictions of the agency theory in the light of the ‘pay-for-performance’ principle. They found a small significant relationship between CEO compensation and performance (Jensen & Murphy, 1990).

Furthermore, Aggarwal and Samwick (1999) state that the typical principal-agent model fails to recognize that the interaction between directors and managers occurs in an environment of strategic interactions between firms, in imperfectly competitive markets. According to their study, relative performance evaluation filters out common industry shock by placing a positive emphasis on the firm’s own performance while emphasizing the negative performance of the industry. This

(7)

negative sensitivity of an industry pay-for-performance principle implies that a CEO will receive higher compensation when CEO’s of other firms within the industry deliver lower returns to their shareholders (Aggarwal & Samwick, 1999). In other words, this study implies that CEO’s from well performing firms are compensated better than CEOs from worse performing companies and, moreover, that industry shocks due to economic circumstances are not taken into the equation of deciding on the height of CEO pay.

1.2 Performance and Reputation

Another concept grabbing attention within the academic world is the effects of company reputation on company performance. As a famous quote of Benjamin Franklin entails “It takes many deeds to build a good reputation, and only one to lose it” (Eccles et al., 2007). Company reputation has been recognized as a critical factor in successfully marketing a service (Thomas, 1978; Lewis and Booms, 1983). Company reputation has been found to be of significant influence on consumer’s perceptions of product and service quality as well as having a significant positive effect on consumer value and consumer loyalty (Keller, 1993). Berry (2000) examined branding in consumer service settings, which lead to evidence for the fact that the meaning customers derive from the service experiences they have is influenced by company reputation. Nguyen and Leblanc (2001) confirm that corporate reputation and corporate image are acknowledged as having an impact on customer loyalty toward the firm. Customer loyalty leads, among other effects, to price insensitivity (Nguyen & Leblanc, 2001). However, a company does not indefinitely project a unique image that represents its reputation. Rather a firm possesses various attributes that are put into a different perspective by different stakeholders. Clients, employees and shareholders may value different attributes and experience different types of contact with firms (Dowling, 1988). The difficulty within academic research is put forward here: finding a valid and reliable measure for company reputation representing the views of all the different stakeholder groups. As Nguyen and Leblanc (2001) state ‘the building of corporate image and reputation is a lengthy process which may be improved rapidly by technological

(8)

breakthroughs and unexpected achievements or, conversely, destroyed by neglecting the needs and expectations of the various groups who interact with the firm’. Thus, firm reputation must include different stakeholder views and may be impulsively created or destroyed by dynamic environmental developments or events. Another way firm reputation is build is through its credible actions. This is considered fragile, because one bad action by the firm has a much stronger effect on the consumer’s view of the firm, than that of a good action (Herbig and Milewicz, 1994). This means that a company’s reputation is strongly affected by the results of its perceived actions within a certain period. Therefore, company reputation rankings may easily differ from year to year (Herbig and Milewicz, 1994).

There are several publications of company reputation rankings that regularly produce reputational ratings of companies by assessing the evaluations from different stakeholder groups (Fombrun, 1998). Reputation rankings call attention to the activities of firms, and also have an influence on the appreciation of certain firms by consumers. On top of that, they influence the behaviors of other stakeholder groups observing companies that are rated. They can turn unknown firms into ‘celebrity companies’ and are also able to transfer famous firms into infamous firms (Rindova et al., 2006). Publishers of these reputation rankings are, among others, Business Week Magazine and Fortune Magazine, as well as newspapers like the Financial Times and The Wall Street Journal (Fombrun, 2007). Therefore, one can conclude that getting a company with a positive score on a reputation ranking list is in the interest of all stakeholders of a company.

1.3 Research Gap

Research on how company performance influences company reputation, using reputation ranking scores was performed by Fombrun and Shanley (1990), and McGuire, Schneeweis, and Branch (1990). These studies found highly significant relationships between performance and firm reputation. However, research on effects on, or from, company reputation rankings is limited to research done on

(9)

the effect of philanthropic expenditures on reputation ranking, the effects of reputation ranking on several inter firm variables such are human resources and financial firm performance, and the influence of reputation scores on the financial performance of firms (Brammer & Millington, 2005; Joo & McLean, 2006; De la Fuente Sabaté & De Quevedo Puente, 2003). Research that touches upon the subject of the effect of CEO pay on firm reputation is available in the form of the effects of CEO pay on firm performance, where performance measures are included (Devers et al., 2007).

So, there is no academic research available on the possible link between the monetary level of executive compensation and a company-specific reputation scores. One of the reasons why it is important to study the link between executive compensation and company reputation ranking is the growing consumer power in the current digital era (Labreque et al., 2013). All information a consumer wants on a company is just one click away via the internet. As mentioned, company reputation has a significant influence on consumer’s perceptions of product and service quality as well the significant positive effect on consumer value and consumer loyalty (Keller, 1993). Another reason to study the link between CEO compensation and company reputation ranking is the fact that information is better available in the digital era. Consumers are not the only stakeholders that have the power to get information on a company faster and more elaborately; this is also true for competitors or peers of a corporate. In this sense company reputation is especially important to peers whom are deciding to collaborate, invest in, or recommend working with a particular firm. Company reputation has become increasingly important to uphold in a digital era where knowledge has become more transparent and is available almost instantly (Labreque et al., 2013).

This study will find out if the amount of monetary executive compensation has an influence on consumer-based and peer-based reputation ranking. If lowering the amount of CEO compensation helps obtain a better reputation perception from both stakeholder groups, this could be of significant

(10)

importance to firm performance. Therefore, the research aims to study this gap in literature will be done by analysis of secondary data to answer the main question:

‘To what extent does the level of executive compensation of American based companies affect the peer-based and consumer-based reputation ranking of that firm?’

According to the agency theory, the optimal compensation package is one that provides the agent with incentives to align its goals with those of the principal. One effect that contributes to designing the optimal compensation package is the social comparison mechanism (Festinger, 1954). Due to social comparison, the CEO compensation package is supposedly higher for companies that originate from US states where a lot of other comparable companies have their headquarters as well. Therefore, the number of companies within the same US state might moderate the relationship between CEO compensation and reputation ranking. Another factor that moderates this relationship is the political orientation of the US state of origin. Borghesi and Chang (2016) state in their research that Republicans tend to accept bigger wage gaps than their Democratic counterparts. This statement is confirmed by the fundamental belief of Republicans that wages should be determined by the free market (Hawkings, 2017). On the contrary, Democrats root for minimum-wages, higher taxes for the wealthy and a smaller wage-gap (Waltman & Pittman, 2002). This leads us to believe that the political orientation of the state a firm is headquartered in will have a moderating effect on the relationship between the level CEO compensation and company reputation ranking. Thus, the following sub questions will be formulated to explore moderating effects on the main question:

 To what extent does high or low clustering of companies within a particular US state influence the effect of executive compensation on peer-based and consumer-based firm reputation

 To what extent does company US state of origin’s political orientation influence the effect of executive compensation on peer-based and consumer-based firm reputation

(11)

In order to include the views of different stakeholders in this research, two dependent variables will be used to test the hypotheses on. The first dependent variable will be a peer-based reputation score, and the second dependent variable will be a consumer-based reputation score. Both these measures will be based on multiple sub-criteria of reputation. By peer-based measurement is meant that the reputation score will be given by peers working within large companies within the US or experts on the topic of corporate reputation. By consumer-based reputation measurement is meant that the measure represents the reputation scores as perceived by the general public in the US. 1.4 Academic and Managerial Relevance

Existing literature describes the effects of executive compensation on overall firm performance (Murphy, 1999). However, no research has yet touched upon the possible effect of executive compensation on firm reputation ranking. First, this study will elaborate on existing research on the positive effects of company reputation ranking on overall company performance. Once this is established, this study will enrich existing theory by studying the effects of executive compensation on the level of firm reputation ranking, both peer-based and consumer-based. Furthermore, this research will find out if social comparison due to the number of firm headquarters within the same US state affects the relationship between CEO compensation and company reputation. Also, the effect of political US home-state orientation has an influence on the relationship between CEO compensation and company reputation. All effects will be accounted for from a peer-based point of view and a consumer-based point of view. Examining the results of these points of will provide new insights on the relationship between executive compensation and reputation ranking of companies. Finally, this study will provide a deeper insight into the relationship between executive compensation and company reputation ranking by evaluating moderating effects of high or low concentration of comparable companies within the same US state and the political orientation of the US state of origin.

(12)

Based on the excessive growth in literature that studies variables that possibly affect firm performance, there can be concluded that there is a managerial interest in finding out if there is a relationship between executive pay and company reputation ranking (Fombrun, 2007). Specifically, through this study, managers will be aware of the effect that CEO compensation has on company reputation scores. This is important because reputation scores affect customer value and customer loyalty (Keller, 1993). Therefore, managers need to stay informed on company reputation rankings and how these could possibly be improved through adjusting executive compensation plans. Also, managerial conclusions may be drawn from the results of this study by members of boards of directors. The results of this study have the intention to give boards support in building a compensation plan for managers and CEOs that will yield the best results.

1.5 Structure

Following this introduction, there will be a literature review where existing literature on CEO pay, agency theory, corporate reputation ranking and the effect of CEO pay on firm reputation is analyzed and elaborated on. In this section, also the moderating variables are discussed and linked. First the possible moderating effect of high and low concentration of comparable companies with the same US state of origin is discussed, and secondly the moderating effect of Republican or Democratic orientation of US state of origin is examined. In this chapter, the hypotheses will be drafted. In the following chapter, the methodology section, the research design is explained and the research sample and data are thoroughly discussed. This section is followed by the results of the data analysis. In the results section correlation and regression tables with the outcomes of the statistical analysis are presented, and answers are given to all proposed hypotheses. Finally, in the discussion section, the results will be explained and discussed using existing literature. Paragraphs explaining the theoretical and managerial implications of the results will be presented here. The limitations of this study and suggestions for future research are also presented in this section.

(13)

Theoretical Framework

Evaluating existing literature will provide a better understanding of what executive compensation entails and how its level is generally determined. In the following section a summary is given of the most relevant studies touching upon the subjects that are studied in this article. On the basis of this existing literature, hypotheses are developed. The following subjects will be discussed in this section: executive compensation, agency theory, reputation ranking, influence of reputation, influence of number of firms in a particular state, and influence of political orientation. These variables will lead to the hypotheses that are studied in this research, as well as form the conceptual model at the end of this chapter.

2.1 Executive Compensation

The level of executive pay has increased substantially in the past decades. Existing research shows evidence for a significant growth in executive pay within the US during the period 1993-2003, while this is not explained by changes in firm size, performance, and industry classification (Bebchuk & Grinstein, 2005). Therefore, scholars have looked at the justification of the level of executive pay. A study was done examining how compensation committees of a sample of US corporations from the S&P 500 justify their compensation practices to shareholders. The study found that companies that pay their CEOs large base salaries are also more likely to be large firms that have to deal with opinions from many different stakeholders (Wade et al., 1997). This creates a relationship between an organization and its environment that is characterized by contradictory claims creating a symbolic battlefield in which variables, such as organizational reputation, are constantly challenged (Bettman and Weitz, 1983; Salancik and Meindl, 1984). CEOs are in place for managing firm reputation in such a symbolic environment, and the height of their pay should reflect their effort, but we often find that the level of CEO pay exceeds the effort and performance that is shown (Fombrun, 1998).

(14)

The height of the compensation package of CEOs is one of the contexts in which this organizational justification with relation to firm reputation is particularly interesting. Some scholars suggest that high CEO salaries are justified by the impact, and specifically financial impact, that top managers have on companies (Jensen & Murphy, 1990). However, many studies have also found no significant relationship between the performance of a company and the level of CEO pay (Holman & Jenkins, 2002). Nonetheless, the level of executive pay in the United States has become abnormally high, but the question remains: why and how is this justified? In order to adequately understand the landscape of executive compensation and how compensation contracts are designed, first the origin of the problem needs to be discussed. Among financial economists the most dominant approach towards studying executive compensation is that managers pay arrangements are seen as a partial remedy to the agency problem (Bebchuk & Fried, 2003). In this study also, the approach towards analyzing the level executive compensation will happen in the light of the underlying agency problem. 2.1.1 Agency theory

Agency theory tries to determine which contract between a principal and an agent is the most efficient (Eisenhardt, 1989). As mentioned in the introduction, the relationship between principal and agent may be applied to the situation where the board of directors acts as principal and the CEO acts as an agent. Eisenhardt (1989) states that Agency theory is concerned with resolving two problems that can occur in agency relationships. The first problem that arises within agency relationships is that the desires or goals of the principal and the agent may be in conflict with each other and that it can be difficult or expensive for the principal to verify what the agent is actually doing. The second issue that arises is with regard to risk adversity. Agency theory depicts that the agent is more risk adverse that the principal. The reason that these different attitudes towards risk may result in inefficiencies is that agents may prefer to undertake different actions or decisions because of their conflicting goals and different attitudes toward risk (Eisenhardt, 1989). Jensen and Meckling (1976) argue that the agency problem characterizes the corporate governance choices of the board of directors of firms and the

(15)

CEOs resulting behavior. According to the agency theory CEOs will seek to increase their own utility at the expense of the firm due to the goal misalignment and risk adversity between both parties (Jensen and Meckling, 1976).Furthermore, the board of directors is unable to completely monitor all the actions of CEOs and this makes it difficult to discover if CEOs act in ways harmful to the firm resulting in a bad reputation (Bosse & Phillips, 2016). In order to solve these agency problems, motivational systems are in place with the main goal to let the agent, or in this case company executive, act in the best interest of the principal, or in this case the board of directors. The type of motivational system that is most frequently used to align goals between CEOs and boards of directors is monetary compensation (Bosse & Phillips, 2016). Incentives that motivate agents according to agency theory are all monetary and include: commissions, stock, options, and salaries (Eisenhardt, 1989).

So the answer of why monetary compensation as a motivational system is in place is to manage the agency relationship between CEOs and boards of directors are the underlying assumptions of agency theory (Eisenhardt, 1989). Motivational theory depicts that if agents are not sufficiently compensated, they will shirk, thus affecting company performance and indirectly endangering company reputation. The motivational assumption therefore emphasizes the importance of designing the right compensation package in order to prevent CEOs from taking decisions that are harmful for the reputation of the firm.

2.1.2 The Motivational Assumption

The agency theory views the right monetary executive compensation system as a remedy to the agency problems, motivating CEOs to act in the best interest of the board of directors. Bebchuk and Fried (2003) bring forth two approaches through which the motivational assumption draws the best monetary compensation system. The first approach they form is the ‘optimal contracting approach’. Within the optimal contracting approach boards of directors are expected to design a pay scheme that provides executives with sufficient incentive to maximize shareholder value. The optimal contracting

(16)

approach states that it is possible to draw the perfect contract through a pay-for-performance model (Bebchuk & Fried, 2004). However, in the sheer magnitude of CEO pay that has no relation with the actual performance of CEOs depicts that in reality the optimal contracting model and pay-for-performance principle is not working within the executive landscape as it should (Edmans & Gabaix, 2009; Holman & Jenkins).

The second approach introduced by Bebchuk and Fried (2003) is the ‘managerial power approach’. Within the managerial power approach, executive compensation is viewed both as an instrument for solving the agency problem but also as a part of the problem itself. Bebchuk and Fried (2003) state that because of the executives’ managerial power, the CEO has an influence over the board which gives him or her significant influence on the nomination process of board of director members. Therefore, the board of directors will generally be inclined to approve CEO pay arrangement that are higher than they should be, as long as this compensation package remains within the range of what can be justified. The managerial power approach also suggests that if board a of director members gets known for bartering with executive compensation appeals, they are less likely to be invited to join other companies’ boards (Jensen & Murphy, 1990). Besides the managerial power approach, directors have another reason to have little personal motivation to fight the CEO on executive compensation matters. Directors usually only have nominal equity interests in a company, so even if a member of the board of directors doesn’t place much value on keeping his or her seat on the board, the nominal equity is a reason to go along with CEO pay arrangements (Jensen & Murphy, 1988). Also, the aspect of information asymmetry on the part of the board of directors, due to a lack of easy access to independent information and advice on executive pay matters, makes it very difficult to effectively argue against a CEO pay plan (Murphy, 1999). On the one hand this means that the board of directors is in charge of designing an executive compensation plan that motivates the CEO to act in their best interest, but on the other hand the CEO's managerial power position puts him or her in a position to decide the level of their own pay as well.

(17)

So, even though the motivational assumption depicts that executive compensation should reflect firm performance accurately, this isn’t true in many cases. Consequently, CEO compensation packages often end up being much higher than they should be (Bebchuk & Grinstein, 2005; Edmans & Gabaix, 2009). Executive compensation packages that are out of proportion have negative consequences such as stakeholders giving low firm reputation. The following paragraph will elaborate on the importance of good reputation ranking.

2.2 Company Reputation Ranking

Good corporate reputation is a key factor to a firm because of their potential for value creation. The intangible character of corporate reputation makes it a competitive advantage since replication of reputation by competing firms is considerably more difficult (Roberts & Dowling, 2002). Firms that possess a competitive advantage, have resources that are valuable, rare, non-substitutable and inimitable (Barney, 1991; Grant, 1991). A growing amount of studies confirm that good corporate reputations create strategic value for the firms that are allied to these corporate reputations, and that reputation may be considered a competitive advantage (Dierickx & Cool 1989). Also, Wilson (1985) agrees that a firm’s reputation is an asset that is able to generate future rents. An example of the strategic importance of reputation may be found in economic game theory. In general, available information about a firm might include its cost function, plant capacity, location, marketing plans, management abilities, R&D expenditures, company values and so on. Each of the firms within a market will know their own capabilities and strategies, but don’t know those of other players in the market. Therefore, players within a market gather data based on historic performance to form beliefs about what rivals are dangerous to their own firm. When these data are stable for all players, firm reputation of rival firms gives clues about their future performance (Weigelt & Camerer, 1988). Company reputation reflects both the history of its past actions, as well as affecting consumers’ expectation of products or services (Rosenthal and Landau, 1979).

(18)

So, good company reputation is generally viewed as a company asset that can enhance buyers’ expectations about the company’s offerings. Also, a positive reputation is able to mitigate uncertainties about products or services a company offers (Yoon et al., 1993). Moreover, Montgomery (1975) found during his study that company reputation is equally important as the uniqueness and promotional effectiveness of a product or service. On top of that, company reputation has been recognized as a critical factor in successfully marketing a service (Thomas, 1978; Lewis and Booms, 1983). Landon and Smith (1997) examined the influence of reputation in their study regarding consumer decision making. They found not only that consumers incorporates past company reputation into their decision-making process, but also refer to collective reputation. In other words, company reputation plays a big role in the decision-making process when consumers are deciding whether or not to buy a certain product. Also in support of the importance of reputation, many empirical studies have found that price is not a good indicator of product quality (Zeithaml, 1988; Geistfeld, 1988). In conclusion, overwhelming evidence exists for the importance and positive effects of company reputation. But how is company reputation currently managed and is CEO compensation one of the factors that affect company reputation?

2.2.1 Reputation and Reputational Risk

Company reputation is not an easy variable to quantify. Executives know the importance of their companies’ reputations. Companies with strong positive reputations attract better employees, they are perceived as providing more value which allows them to charge a premium to customers, and customers are more loyal and buy broader ranges of products and services (Eccles et al., 2007). A lot of companies however do an inadequate job of managing reputation and risks to their reputation. This is due to an absence of agreement on how to define and measure reputational risk. Given the lack of common standards even sophisticated firms have only the slightest idea of how to manage reputational risk (Eccles et al., 2007).

(19)

According to Loh (2007) there are three things that determine the extent to which a firm is exposed to reputational risk. The first exposure of reputational risk is when the company’s reputational value exceeds the true value of the firm. The second risk is how much external beliefs and expectations change, which can contribute to the value gap or can make this gap even wider. The third is the quality of internal organizational coordination between departments, meaning that one department can perform very well, but that others not necessarily do this as well (Loh, 2007). The management of reputational risk starts by recognizing that reputation is a matter of perceptions. The overall reputation of a company is the sum of its reputation among all the stakeholders it has, since a strong positive reputation among stakeholders across multiple categories will result in a strong positive reputation for the company overall (Eccles et al., 2007). On the other hand, organizations that actually meet the expectations of their various stakeholders may not get full credit for doing so. Here is where the CEO comes in that needs to manage the risk through objectively assessing reputation versus reality, assessing and accepting impact of changing expectation and explicitly focus on reputational risk as is displayed in figure 3.

Figure 3 Framework for Managing Reputational Risk(1)

(1)

(20)

2.2.2 Executive Compensation and Reputation

So, strong and performing CEOs should be in place in order to establish a strong and sustainable reputation. Within the management of reputational risk that needs to be done by executives, a link is seen with agency theory. Since the CEO is the one responsible for managing reputational risks and keeping up or improving existing reputation, he or she needs to be motivated to do so. In other words, an agency problem arises. A factor important to the sustainability of reputation and the solvation of the agency problem that arises is mentioned by Eccles et al. (2007). This factor is that the executive should periodically report to top management what the key reputational risks are and how they are being managed as it is up to the CEO to decide whether the risks are acceptable or too dangerous, and in the latter case need to be managed. This links to the assumptions of agency theory of information asymmetry, bounded rationality and risk aversion that predict that the CEO is bound to make different decisions than the board of directors. In other words, if companies want manage reputational risks, the board of directors needs to compensate the executive properly in order to motivate him or her to act in the best interest of the company resulting in retained or improved company reputation. So, in short, executive compensation may affect company reputation, and the solving of reputational risks, because of the agency relationship between the board of directors and the CEO.

As previously mentioned, CEO compensation packages are often not structured to maximize value for outside shareholders according to the managerial power approach (Murphy, 1999; Bebchuk & Fried, 2003). The fact that executives have a significant influence over their own compensation package suggests that no optimal contracts are produced. With regard to the assumption of self-interest, the argument can be made that a high level of CEO pay stands for a firm that has more severe agency problems. This argument is supported by the study of Core et al. (1999) who found that CEOs earn greater compensation when governance structures are less effective. The overall conclusion they found was that CEOs at firms with weaker governance have greater agency problems

(21)

and that CEOs at firms with greater agency problems receive greater compensation (Core et al., 1999). So, weak governance leads to greater agency problems resulting in high CEO compensation. In the case of the stakeholder-group ‘peers’ the high CEO compensation would let them to believe there is weak governance within the company, thus damaging the reputation of that company in their perspective. Furthermore, concerning the consumer-based view, as the articles discussed in the introduction section, high CEO pay is associated with bad performance, high wage gaps, and unjustified compensation. Therefore, we argue for both the peer- and the consumer-based reputation measure that a high level of CEO compensation has a negative influence on company reputation ranking. From this argument, the main hypothesis of this study is drawn:

Hypothesis 1a: a high level of executive pay has a negative effect on the peer based reputation ranking of the company

Hypothesis 1b: a high level of executive pay has a negative effect on the consumer based reputation ranking of the company

2.3 Moderating Effects

Following up on the main hypothesis ‘a high level of executive pay has a negative effect on the reputation ranking of the company’, this study proposes two moderating effects of the level of executive pay on company reputation ranking. The first proposed effect stems from the high or low concentration of Fortune 500 firms with the same US state of origin and will be discussed in the subsequent paragraph. The second moderating effect originates from whether the political orientation of the US state of origin is Republican or Democrat. In this second paragraph arguments will be formed on why political affiliation moderated the relationship between CEO pay and reputation ranking.

(22)

2.3.1 US state or origin

As previously mentioned, a question that is being raised by many scholars is ‘why has CEO pay increased so much?’ The study Gabaix and Landier (2008) did on this particular topic shows that the primary height of CEO pay depends for a large part on the size of the firm. That is partially why CEO pay has risen, because of the six-fold increase of market capitalization of large, US based companies in the 1980s and 1990s. Another factor influencing the height of compensation packages is the managerial power approach as was explained in the previous paragraphs. However, there is another factor that plays a significant role in the determination of executive compensation, namely the social comparison effect. Social comparison refers to the need of individuals to evaluate their opinions and abilities compared to their peers (Festinger, 1954). Academic research using social comparison as a variable that might influence the level of CEO pay found strong associations between level CEO compensation and the social compensation level of the members of the board of directors (O’Reilly et al., 1988).

Companies that are comparable in size are subjects of comparison for boards of directors when they are determining their executive compensation plans altogether or adjustments to them. Social comparison among both CEOs and boards of directors is bound to drive the amount of the compensation packages up. This results in higher compensation packages for executives from companies with many comparable companies nearby than companies who are situated in areas where less comparable companies reside. The second category of companies will be less influenced by social comparison because there are simply fewer companies within the same state. On top of that, Belliveau et al. (1996) found in their research that the height of the amount of CEO pay also has a significant positive relationship with the number of influential people, including other CEOs, an executive had in his network. This also argues for a moderating effect for states with either a high concentration of comparable companies in the same area opposed to companies with little

(23)

comparable corporates within the same area. Within the United States there are three states that have more than 50 Fortune listed companies who have their headquarters in this state (Fortune, 2017). New York, California and Texas are these states, and consequently in these areas more CEOs know each other since more comparable companies are within the same area. Therefore, CEO pay will generally be higher in these states due to the social comparison effect (Belliveau et al., 1996). The argument here is that the relationship between CEO compensation and both peer- and consumer-based reputation ranking will be moderated by the extreme social comparison effects that take place in New York, California and Texas due to a high concentration of Fortune 500 firms there. It is likely that boards of directors from companies residing in one of these states will look at comparable companies within the same geographical area or state of the firm that has the same legislation regarding executive compensation. So, when there are a lot of comparable firms in the nearby area the level of CEO pay will be affected due to social comparison effects. The effect of social comparison is that in states where there is a high concentration of firms the level of executive compensation packages will be driven up. This will result in higher CEO pay overall for firms from this US state resulting in an extra negative influence on reputation ranking. This drafts the following hypothesis:

Hypothesis 2a: the relationship between executive pay and peer-based reputation ranking is weaker for firms who reside in New York, California and Texas where the concentration of Fortune 500 firms is high

Hypothesis 2b: the relationship between executive pay and consumer-based reputation ranking is weaker for firms who reside in New York, California and Texas where the concentration of Fortune 500 firms is high

On the other hand, there are also US states that only include just a few or even no other comparable companies within the same area. If there is little to no social comparison possible, the firm in question is left to create their own compensation package, which should paint a more accurate

(24)

picture of how the performance of the firm is going. Also, because there are no similar firms in the area, there is a decreased change of the managerial power approach taking effect. Since there are (almost) no other company boards of the same magnitude in the same area, the board of directors has no incentive to approve compensation demands of the CEO (Jensen & Murphy). In theory, this should lead to lower CEO compensation packages and a more accurate compensation in proportion with the performance of a company. Because of this lack of social comparison, CEO compensation packages will be lower resulting in a higher reputation scores. Therefore, the effect of low concentration of firms within the same US state moderates the relationship between CEO pay and reputation. For the purpose of this study the assumption is made that a low concentration of firms within the same US state is less than ten firms within the same area. This leads to the following hypotheses:

Hypothesis 3a: the relationship between executive pay and peer-based reputation ranking is weaker for firms who have a low concentration of Fortune 500 firms within the same area

Hypothesis 3b: the relationship between executive pay and consumer-based reputation ranking is weaker for firms who have a low concentration of Fortune 500 firms within the same area

2.3.2 Political Orientation

The second effect that moderates the relationship between CEO pay and company reputation ranking is the political orientation of the US state of origin from the firm. The distinction is made whether a US state has a Republican or Democratic orientation. This moderating effect is extracted from academic research by Hutton et al. (2014) that showed that political orientation affects corporate policies. Republican top corporate managers who are likely to have conservative personal ideologies were studied to adopt, on average, more conservative corporate policies than their Democratic counterparts. There was also found that firms under Republican reign tend to have less debt, thus

(25)

possibly having a positive impact on company reputation (Hutton et al., 2014). Furthermore, as previously mentioned, Republican supporters, opposed to Democratic supporters, tend to accept larger wage-gaps and believe that wages should be determined by the market (Borghesi & Chang, 2016). Taking into account both the positive influence of the Republican conservative strategy, that is more likely to occur within Republican oriented state, and the lesser sensitivity of Republican supporters, a moderating effect is seen on the main relationship. Both Republican peers and consumers will be more accepting towards high CEO compensation and will therefore be less rigorous in rating a firm lower in reputation because of this. Therefore, the relationship between CEO compensation and reputation ranking is moderated for companies originating from a Republican state of origin. This constitutes the following hypotheses:

Hypothesis 4a: the relationship between executive pay and peer-based reputation ranking is weaker for firms with a Republican oriented state of origin

Hypothesis 4b: the relationship between executive pay and consumer-based reputation ranking is weaker for firms with a Republican oriented state of origin

Companies that reside in US states that have a democratic political orientation there also exists a moderating effect. In 1992, the Democrats made CEO pay an issue in presidential campaign and even passed a law to punish cash salaries to executives over $1 million (Holman & Jenkins, 2002). On top of that, ideologically Democrats believe in minimum-salary and a smaller pay-gap (Borghesi & Chang, 2016). Consequently, in Democratic oriented areas a high CEO pay will be a sensitive issue, thus affecting reputation ranking. Therefore, also for democratic oriented peers and consumers we believe that there exists a moderating effect on the relationship between CEO pay and reputation ranking. We believe that, contrary to the Republican moderating effect, in Democratic states the relationship between high executive compensation and lower reputation rankings for both consumers and peers will be lower. This leads to the following hypotheses:

(26)

H4 + H5

H2 + H3

H1

Hypothesis 5a: the relationship between executive pay and peer-based reputation ranking is stronger for firms with a Democratic state of origin

Hypothesis 5b: the relationship between executive pay and consumer-based reputation ranking is stronger for firms with a Democratic state of origin

2.4 Conceptual Framework

In Figure 2, the conceptual framework is visualized displaying all the variables and hypotheses.

Figure 2 Conceptual Framework

CEO pay

High/low concentration of companies with same

US state of origin Political orientation US state of origin Reputation score peer-based Reputation score consumer-based

(27)

Methodology

In this chapter, the methodology section, the research method will be specified by (1) discussing the research design, (2) discussing the research sample, (3) explaining how the different variables were measured and collected, and (4) describing the statistical procedure.

3.1 Research Design

In order to answer the research question, a cross-sectional design is used. A cross-sectional design is a type of observational study that involves the analysis of data collected at a specific point in time, for example at the end of a year. The collected data originate from either a population or a representative subset. In order to test the hypotheses in this study, a cross-sectional regression is used (Olsen and George, 2004). Using regression analysis for testing the hypotheses helps to understand how the typical value of the dependent or criterion variable changes when the independent variable is varied, while the control variables are held fixed (Hazewinkel, 2001). Cross-sectional regression analysis suits the study best since the research sample contains data from many different firms within the United States in order to answer the research question ‘To what extent does the level of executive compensation of American CEOs affect the overall peer- and consumer-based reputation ranking of that firm?’ and to conclude if this relationship is affected by a high concentration of large firms within the same US state or the state of origin being democratic or republican oriented.

To conduct the study secondary data are used. The main advantage of using secondary data is that it forfeits the use of time and costs that would be incurred while using primary data (Grady, et al., 2013). Another advantage of using secondary data is that the University gives easy access to high quality databases such as WRDS. A disadvantage of using secondary data is however, that the conductors of the research have no control over the study population, design or measurements (Grady, et al., 2013). In the case of this particular study there is a limited timeframe and no budget, therefore establishing the usage of secondary as more convenient.

(28)

3.2 Research sample

The research sample used in this study is the list of Fortune 500 companies from the year 2015. The Fortune 500 list ranks companies by total revenues for their respective fiscal years and includes the 500 largest companies in the United States in terms of total revenues. Included in the survey are companies that are incorporated and operated in the US, and file financial statements with a government agency. The list includes companies that file with a United States government agency but are owned by domestic or foreign private companies that do not file such financial statements (Fortune, 2017).

Table 1 Variable Definition, Measures, Data Sources

Measure Measurement Source

Company Code Ticker WRDS

Name Company name Fortune 500

Executive Compensation Level of CEO pay in 2015 Compustat

Reputation Ranking the Harris Poll Reputation Quotient 2017 Harris Poll US State the Company’s state of HQ Fortune 500 Political Orientation State Democrat or Republican in 2016 election CNN

Industry code Company Industry WRDS

Return on Assets Proxy for company profitability in 2015 Fortune 500 Total Revenues Proxy for company size in 2015 Fortune 500 Number of Employees Proxy for company size in 2015 Fortune 500

For the independent and moderator variables in this study, the Wharton Research Data Services (WRDS) databases are used. This database is an award-winning research platform and business intelligence tool containing 250 terabytes of data across multiple disciplines including accounting, banking, economics, ESG, finance, healthcare, insurance, marketing and statistics. Advantages of the WRDS database as opposed to other databases are that WRDS offers more data, access and support (The Wharton School, 2017). The independent and control variables used in this study will be data covering the year 2015. The year 2015 will give the most recent and complete

(29)

dataset with which the effects on the reputation rankings assembled in 2016 are visible. The overview of variables, measures and data sources used in this study can be found in table 1.

3.3 Variables

In this section, the origins of all the variables that are used to test the hypotheses are shortly summarized. In this section is also explained per variable why this particular set of data is used to test the hypotheses.

3.3.1 Dependent Variables

Because it is difficult to measure reputation, we will test all our hypotheses with two different reputation measures. The first measure is peer-based, in other words, a reputation rating based on people who themselves work in the companies that are rated or are experts on corporate reputation. The second reputation measure is based on consumer ratings, also referred to as the ‘general opinion’ of the public and therefore, will be consumer-based.

Reputation Ranking I is Fortune’s annual survey of the world’s ‘Most Admired Companies’

which is by far the most visible and frequently cited source of reputation ranking, that was published for the first time in 1982 (Fombrun, 2007). This reputation ranking list includes about 300 ranked companies both from the US and well known foreign firms that operate in the US. However, in this study only the firms with corresponding reputation scores that reside in the US are used. This reputation ranking of the Global Fortune 500, using only the US based companies, will be the first dependent variable in this study. This measure is used because of its reliability and because the measure is peer-based. The Korn Fery Hay Group conducts this research in collaboration with Fortune 500, and has published this list since 1997. The Hay Group is a global management consultancy firm (Haygroup, 2017). To determine the best-regarded companies in 51 different industries, the Hay Group asked executives, directors, and analysts to rate enterprises in their own industry on nine

(30)

criteria, from investment value and quality of management and products to social responsibility and ability to attract talent. A company’s score must rank in the top half of its industry survey to be listed in the top 100. They winnowed the assortment to the highest-revenue companies in each industry from a total of 680 firms in 28 countries of including the 100 top-rated companies that were picked from each industry. A sample of the top 300 companies including companies from all industries is the list that is published on the Fortune website (Fortune Most Admired, 2017). The reputation score used in this study is the overall score that companies assembled that was measured using the ranking that companies got in the different categories that were used to measure this overall reputation score. The nine categories in which the firms are ranked, are: (1) innovation, (2) people management, (3) use of corporate assets, (4) social responsibility, (5) quality of management, (6) financial soundness, (7) long-term investment value, (8) quality of products/services, and (9) global competitiveness. An average of the rankings is then assembled which determines the overall score a company gets (Fortune, 2017). The dependent variable score from the Fortune 500 is gathered through the ‘Worlds’ Most Admired Companies’ list from 2016. The scores are given on a continuous scale from 0.0 to 10.0, where 0.0 represents the lowest reputation score possible and 10.00 stands for the highest reputation score possible.

Reputation Ranking II data in this study originate from The Harris Poll Reputation Quotient

(RQ®). The Harris Poll RQ measures the reputations of the most visible companies in the US, as perceived by the general public. In other words, this is a consumer-based reputation measure. The Harris Poll has been conducted since 1999 and has provided leaders with decision making insights, helped CEOs to proactively manage corporate reputation, identify risks and opportunities and enable them to compete with established companies within industries. The Harris Poll Reputation Quotient is an evaluation of perceptions of corporate reputation and assesses the relationship between supportive behaviors and reputational equity. Supportive behaviors include: trust in companies to do

(31)

the right things when faced with product or service issue, willingness to say something positive, and consumer intent to purchase or recommend your products and services (The Harris Poll, 2017).

The Harris Poll Reputation Quotient research has evaluated public perceptions across 20 different attributes that are classified into six dimensions of corporate reputation: (1) Social Responsibility, (2) Vision & Leadership, (3) Financial Performance, (4) Products & Services, (5) Emotional Appeal, and (6) Workplace environment. These attributes have allowed The Harris Poll to reliably trend performance over time and identify industry and company specific drivers of corporate reputation. The 20 attributes are divided among the six dimensions of corporate reputation and can be viewed in figure 1 (The Harris Poll, 2017).

Figure 3 the Harris Poll - Six Dimensions of Corporate Reputation and Attributes(2)

(2)

(32)

The Harris Poll annual RQ study starts with a Nomination Phase that is used to identify the companies with the most “visible” reputations according to the general public. This is measured through respondents that are asked to name companies that stand out as having the best and worst reputations overall and two open-end questions are used (The Harris Poll, 2017):

 Of all the companies that you’re familiar with or that you might have heard about, which TWO - in your opinion - stand out as having the BEST reputation overall?

 Of all the companies that you’re familiar with or that you might have heard about, which TWO - in your opinion - stand out as having the WORST reputations overall?

Nominations from all interviews are combined and tallied with subsidiaries and summed up within the parent company. The final list of the 100 most visible companies in the US is measured in the RQ Ratings Section. For the current list of The Harris Poll RQ, the interviews took place between September 13-15 and October 4-6, 2016 among 4,092 US adults to identify the top 100 “most visible” companies (The Harris Poll, 2017). Nominations from all interviews are combined and tallied with subsidiaries and summed up within the parent company. The final list of the 100 most visible companies in the US is measured in the RQ Ratings Section.

The RQ Ratings survey was conducted in English and took place among the general public adults 18+ within the US. Respondents were randomly assigned to rate two of the companies that they were “very” or “somewhat” familiar with. After this was completed, the respondent was given the option to rate a second company. The interviews lasted about 20 minutes each and every company was rated by approximately 300 US adults of 18+ ages. The rating interviews took place online between November 28 and December 16, 2016 among 23,633 US adults that obtained an average of approximately 300 ratings per company (The Harris Poll, 2017). The Corporate Reputation scores assigned to the companies are on a scale of 0 to 100, where a company’s reputation score below 50 is

(33)

interpreted as having a very poor reputation and a score of 80 or higher stands for an excellent reputation score (The Harris Poll, 2017).

The independent, moderator (with the exception of the political orientation) and control variables originate all from 2015, whereas both the peer-based Fortune 500 measure and the consumer-based Harris Poll reputation measures used in this study will be taken from a dataset gathered over the year 2016. This is done to accurately reflect the effect of the level of CEO pay in 2015, on the reputation in 2016.

3.3.2 Independent Variable

Independent variable used in this research is the level of executive CEO compensation. For the

measure of executive pay the WRDS database is used (The Wharton School, 2017). Specifically, the data for the level of CEO pay come from the Compustat. Compustat is a database that includes financial, statistical and market information on active and inactive global companies throughout the world. Compustat has been collecting data since 1962 and provides a broad range of information products directed at institutional investors, universities, bankers, advisors, analysts, and asset/portfolio managers in corporate, M&A, private capital, equity, and fixed income markets (S&P Global, 2017). The measure used in this study includes the ‘total compensation’ for the year 2015 that enhances the total received salary and bonuses and other income of CEOs in this year. The specific variable is ‘TDC1 – Total Compensation’ from Compustat. This variable has as an output the amount of CEO compensation in millions of Dollars. This amount includes the salary, bonuses and other received compensation of the executive over the year 2015. The CEO compensation is represented on a continuous scale.

(34)

3.3.3 Moderation Variables

To test the moderating effects of the hypotheses formed in the previous chapter, secondary data of these variables will be used. The data for the moderation variables will be collected partly from the WRDS database and partly from different sources.

Company State of Origin data will be gathered through Fortune 500 where the largest

companies in the US are placed with its headquarters in the right US state. In order to test the hypotheses stating that a concentration of large firms within the same US state has a moderating effect on the main relationship. The firms from New York, California and Texas will be assigned a 1, and the other firms originating from different US states will be assigned a 2.

Political Orientation Company State of Origin data will be gathered through the poll display

from Cable News Network (CNN, 2016). For this moderation variable the exception is made and data from 2016 is used. This is done because the data of the previous election in 2012 are deemed not relevant enough since they are too far away from 2015 and the intermediate panel data on political orientation are not reliable enough. For statistical purposes a number is connected to the political orientation per state. Majority Republican = 1 and Majority Democrat = 2.

3.3.4 Control Variables

Control variables will be used to test whether the dependent variable, Company Reputation Ranking,

is really affected by the independent variable, level of CEO pay. Also, the variables will make sure that the relative impact of the dependent variable on the independent variable is portrayed correctly. The control variables used in this study will control for profitability, size and industry, as displayed in table 1. The control variables for size and profitability were gathered through Fortune 500 (Fortune, 2017). The Return on Asset measure which stands in as a proxy for profitability is measured by taking the profit of 2015 as a percentage of assets. The proxies for size, revenues and employees, represent the

(35)

revenues gathered over 2015 and the number of employees at the end of 2015. The industry codes were gathered through the same Compustat database using TICKER codes that was also used for CEO compensation. This industry measure from WRDS was then classified by the SIC industry range. The SIC industry range exists out of ten different broad industry classifications as defined by the US government (United States Department of Labor, 2017).

 Proxy for profitability will be return on assets (ROA) as a percentage over 2015

 Proxies for company size will be total revenue in dollars and number of employees in units

over 2015

 Company Industry Code will be the industry code according to the SIC industry range

Results

In this section, the results of the study are represented. In the first section the sample and the reliability of the sample is discussed. In the second section the data statistics are outlined and discussed, followed by the third section where the correlation results will be outlined and defined relevant to the hypotheses. This section is followed by the final part of the results where the regression results are presented and correctly linked to the hypotheses. In table 2 the results may be found of companies and how many Fortune 500 companies are in their state, how it affects this and in table 3 an overview is presented of the political orientation of US state of origin. Table 2 outlines the variable statistics, table 4 and 5 presents the complete correlation matrix, and table 6 and 7 capture the regression results.

4.1 Sample Statistics

The samples used in this study are companies included in the Fortune 500 companies list from the year 2015. Company name and US state of headquarters and control variables were found via the Fortune 500 website within the list for the year 2015. For the first peer-based dependent variable the Fortune 500 reputation ranking score was used from the year 2016. Not all companies on the Fortune Reputation Ranking list were from the US and also not all Fortune 500 companies were included in the Reputation Ranking. After eliminating the Fortune 500 companies that were not on the Fortune Reputation Ranking there was a sample of 211 Fortune 500 companies left. For 11 of these companies there was no CEO compensation data available in Compustat. Another 2 companies were not on the Fortune 500 list, so therefore the peer-based sample includes 198 companies (N=198).

For the second consumer-based dependent variable, as a sample, the companies on The Harris Poll 2017 that are also on the Fortune 500 list were used. The Harris Poll includes the reputation ranking data for the 100 most visible companies in the US which were assembled during

The influence of executive compensation on company reputation ranking