An empirical analysis of the relation between the level of CEO pay and employee satisfaction

(1)

An Empirical Analysis of the Relation Between the Level of

CEO Pay and Employee Satisfaction

Student Name: Ravi Bhageloe Panday

Student Number: 10409149

Study: BSc Economics and Business, variant Finance and Organization

Supervisor: Ms. Jindi Zheng

(2)

2

Statement of Originality

This document is written by Ravi Bhageloe Panday who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creat ing it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

The pay gap between CEO and workforce has been increasing tremendously over the past decades. A negative relation exists between pay dispersion and employee satisfaction. As the compensation of top management including CEO is generally publicly known and more likely to be viewed in the media, workers tend to compare their compensation with them. This study investigates how the level of CEO pay affects employee satisfaction. An empirical analysis shows that the level of CEO pay is significantly negatively related to employee satisfaction.

(4)

4

Tables and Figures

Table 1 Sample sizes Main group and Control group 16

Table 2 Descriptive Statistics and Correlations 20

Table 3 Effect of the level of CEO pay on employee satisfaction 22

(6)

6

1. Introduction

Senior executive and particularly CEO remuneration have been increasing tremendously over the decades. The ratio between CEO pay and that of the average worker was administered to be 29:1 in 1980 (Phillips, 1991). That same ratio increased to 150:1 in the 1990s and most recently it was assessed to be 331:1 in 2013 (AFL-CIO1, 2014). This ever increasing pay gap did not remain unnoticed due to revised SEC2 requirements disclosure of executive compensation (SEC, 1992) and the degree of public attention that covers this matter. Over the decades, this issue has been studied from both economic and social viewpoints.

Employee satisfaction is greatly influenced by pay satisfaction (Goodman, 1974) which is negatively correlated with the wage gap (Stabile, 2002). Pay satisfaction is to a significant extent the product of comparisons of pay (e.g. Festinger, 1954; Messé & Watts, 1983; Patchen, 1961; Goodman, 1974; and Goodman & Friedman, 1971). Brickman (1975) stated accordingly: “the

value of a given reward is not absolute, but is relative to other rewards with which it is compared.” Pfeffer (1993) showed that the larger the pay gap within organizations, the lower the

workforce’s satisfaction levels. Furthermore, literature reviews performed by Leventhal (1976) and Deutsch (1985) both indicate that pay disparity is negatively related to satisfaction. Ambrose & Kulik (1988) argue a vast pay gap is detrimental since it adversely affects employees’ morale, productivity, absenteeism and turnover.

Workers compare their pay with different people of different levels in the pay hierarchy (e.g. Schwab & Wallace, 1974; Lawler, 1971; Martin & Murray, 1983; Adams, 1965; Crosby, 1976; and Ronen, 1986). Accordingly, Hyman (1942) observed that man determines his position in the compensation hierarchy by making direct comparisons with other groups in the hierarchy without taking fairness into account. Also Festinger (1954) finds no relation to fairness of justice. According to him, significant differences with executives will cause distress independent of fairness reasoned through hierarchy. In other words, workers do not see the extra responsibility that comes with a certain (higher) position, they only see the differences in pay and feel that they are underpaid.

1

American Federation of Labor and Congress of Industrial Organizations represents the largest labor union federation in the U.S.

2

(7)

7 Consequently, one can expect that the ever growing pay gap between CEO and workforce has an effect on the workers’ satisfaction. This study will investigate whether a company with a lower CEO pay does effectively have more satisfied employees. Therefore, this study aims at analyzing the effects of CEO pay on employee satisfaction. Correspondingly, the following research question will be discussed: How does the level of CEO pay affect employee satisfaction?

The main result shows a clear significant negative relation between the CEO pay and employee satisfaction and thus supports this study’s hypothesis. After controlling for various factors, the coefficient of CEO pay remains significant. The findings are consistent with prior research performed by Pfeffer (1993), Ambrose & Kulik (1988) and Wade et al. (2006).

This study contributes to empirical academic literature regarding pay disparity and employee satisfaction. More specifically, this study targets CEO pay instead of the commonly used pay disparity amongst employees. To my knowledge, this has not been examined before.

Secondly, this study is different in terms of quality and broadness of the sample used. As a measure of employee satisfaction, this study uses the published works of Great Place To Work For Institute3 (GPTWI). GPTWI conducts the most extensive employee satisfaction survey in all America. Consequently, the sample used in this study is significantly of better quality and of greater number than samples used in other studies in the field. Pfeffer (1993) conducted tests with students exclusively, Berkowitz and colleagues (1987) gathered data by telephone calls regionally, Wade et al. (2006) worked with a small sample, Caldwell & O’Reilly (1982) studied satisfaction in a laboratory setting with college students as participants and in a subsequent study they conducted in- field, people of solely one profession were surveyed. This study examines the relation with the use of the institute’s extensive dataset, which allows analysis on a national scale with a diverse group of companies.

Thirdly, labor market participants could use the results from this study to select their future employer. Likewise, companies could adjust their payment scheme to increase workforce satisfaction. Above all, the study’s findings can possibly act as a start to a public discussion surrounding this topic which can serve as a fundament for rethinking the government’s present policy in this field.

The remainder of this article is organized as follows. The following section defines how pay disparity and employee satisfaction are related and gives an overview of prior literature. This

3

Great Place To Work Institute represents an international human resources consulting, research and training organization that concentrates on employee satisfaction.

(8)

8 section ends with hypothesis development. The third section describes sources of data, sample sizes and the method of research. The fourth section discusses the results and the fifth concludes the paper.

(9)

9

2. Literature Review and Predictions

This chapter starts with the relevance of employee satisfaction as a field of study and how it is academically defined in existing literature. Subsequently, an overview of prior literature is presented regarding the relation between employee satisfaction and pay dispersion. Next, existing theories on pay disparity are discussed. The section ends with the development of this study’s hypothesis.

2.1 Relevance of employee satisfaction

Employee satisfaction has been a topic of interest for several decades since it evidently affects a great number of organizational aspects. According to academic literature, employee satisfaction is related to accounting profit (Pfeffer, 1993), stock performance (Bloom, 1999), cost of equity (Chen et al., 2013) and workers’ commitment (Stabile, 2002). Additionally, Hellman (1997) found that employee turnover is influenced by job satisfaction. As he added, unsatisfied employees are more likely to leave the organization than satisfied ones.

Several definitions exist of employee satisfaction. Spector (1997) defined employee satisfaction as “the overall feeling about the job or as a related set of attitudes about different

facets of the job.”, or how Locke (1976) construed it : “the possible emotional state resulting from the appraisal of one’s job or job experiences.” Taunton et al. (1997) gave a more

comprehensive definition: “Job satisfaction is a global attitudinal construct that encompasses

several components such as work or task, pay and benefits, coworkers, status and administration.” This study adopts GPTWI’s approach to employee satisfaction. Essentially,

GPTWI measures employee satisfaction by reviewing to what extent workers trust their executives, have pride in their doings and have qualitative relationships with their colleagues. Since this study makes use of the institute’s data set, this study adopts its approach accordingly. More details are to be found in chapter three.

2.2 Pay comparison in relation to employee satisfaction

Employee satisfaction is greatly influenced by pay satisfaction (Goodman, 1974). Several studies examined this effect of pay comparison on satisfaction. Interestingly, workers judge the fairness of their pay through social comparison (Festinge r, 1954; Major & Forcey, 1985; Messé & Watts,

(10)

10 1983; Patchen, 1961; Goodman, 1974; and Goodman & Friedman, 1971). As a result of judgment through social referents (O’Reilly and Caldwell, 1979), individual pay as well as the distribution of pay across the organization are of high relevance. Brickman (1975) stated accordingly “the

value of a given reward is not absolute, but is relative to other rewards with which it is compared.” Pfeffer (1993) shows that the larger the pay gap within organizations, the lower the

workforce’s satisfaction levels. Furthermore, literature reviews performed by Leventhal (1976) and Deutsch (1985) both indicate that pay disparity is negatively related to satisfaction. In addition, Ambrose & Kulik (1988) argue a vast pay gap is detrimental since it adversely affects employees’ morale, productivity, absenteeism and turnover, which are all directly related to employee satisfaction.

There exist significant evidence that workers compare their pay with different people of different levels in the pay hierarchy (Dornstein, 1989; Schwab & Wallace, 1974; Martin & Murray, 1983; Adams, 1965). Correspondingly, Hyman (1942) observed that man determines his position in the compensation hierarchy by making direct comparisons with other groups witho ut taking fairness into account. Likewise, Festinger (1954) finds no relation to fairness of justice in the process of social comparison. According to him, significant differences with equals or executives will cause distress independent of fairness reasoned through hierarchy. In other words, workers do not see the extra responsibility that comes with a certain (higher) position, they only see the differences in pay and feel that they are underpaid.

This is supported by the relative deprivation theory which states that feelings of inequality are a consequence of the pay gap between two dissimilar groups (Martin, 1982). Workers’ perception of fairness of pay is thus influenced by remuneration levels from top management (Cook & Hegtvedt, 1983). Interestingly, Ambrose & Kulik (1988) show that availability of information determines what constitutes a reference group. In other words, workers generally do not discuss compensation with peers and at the same time no disclosure regulation exists around nonexecutive remuneration. Due to public attention usually given to senior executive and in particular CEO remuneration, workers are inclined to select them as a comparison group (Stabile, 2002).

However, other studies found that not comparison but interna l standards largely affect judgments of what is a fair pay (Gurr, 1970; Davies, 1962; Jaques, 1961). Jaques indicates that

(11)

11 wage workers base their perception of fairness on job characteristics4. Also from internal standards perspective, Dyer & Theriault (1976) and Finn & Lee (1972) discuss that individual job performance and all other input, such as experience and training, set the individual’s pay rather independent of what others receive in the firm. This corresponds to Homans’ (1961) concept of equity, in which a fair pay is determined when a worker’s “investments” correspond to his “profits”. Equity is reached when the proportion between profits and investments of one individual is similar to that of others (Finn & Lee, 1972). Accordingly, Bebchuk et al. (2006) sees judgment of pay fairness as a result of workers’ comparisons as an unjustly restrictive perception. As he explained: “After a delicious meal, the world can seem a very satisfying place, particularly

if the good food is accompanied by good wine; it is quite unnecessary to check that one’s own meal was better than anyone else’s.” In other words, comparison will not better (or worsen) your

situation and therefore does not affect your feelings about it.

In contrast to the above, Berkowitz et al. (1987) found that pay comparisons contributed nearly nothing to pay satisfaction. He stresses that nonpay aspects are significant ly better predictors of pay satisfaction. He observed intrinsic job satisfaction as a contributing factor to wage satisfaction. He also found the individual’s psychology as a significant influence on pay satisfaction.

In sum, the majority of the studies show that workers make comparisons with others regardless of equality whilst not account for whether someone’s wage was fair. Consequently workers experience disappointment when their earnings are less.

2.3 CEO pay inflation

Figure 1 viewed in the appendix shows the exponential development of the CEO-average worker pay ratio over the period 1960-2000. Several studies have provided various economical explanations for the increasing pay gap between CEO and workforce. Economists argue the increasing gap is due to the fact that the labor market for executives is distinct from the one for lower employee classes (Galbraith, 1998), the change in historical hiring norm5 (Frank & Cook,

4_{Jaques (1961) uses the five “core” characteristics as defined by job characteristics theory, namely skill variety, task identity, task}

significance, autonomy and feedback.

5

Until the 1990s CEOs were workers from within the organization that worked their way up on the corporate ladder. External hiring was an uncommon event that was also known as the “anti-raiding norm” (Frank & Cook, 1995). Today, the majority of CEOs are hired from outside who require a premium in pay to lure them away from currently held positions, not to mention the necessity to compensate even more a company’s own CEO to prevent them from being taken away by another company. An interesting example covers the case of Ben & Jerry’s Homemade Inc., a company famous for its ice creams. Until 1995, the

(12)

12 1995), the applied Tournament Theory6 (Stabile, 2002), “winner-takes-all markets”7 (Frank & Cook, 1995), differences in job security8 (Kay, 1998), the decline in unionization9 (Baker & Fung, 2001).

In contrast, several studies reason market failure in the determination of executive pay as the main problem and therefore see the above stated economic explanations as insufficient. They discuss as explanation for the ever widening gap the insufficie ntly independent remuneration committees10 (Bryant, 1997; Bebchuck et al., 2002), CEO influence on the nomination of compensation committee members independent of his involvement in its activities11 (Perry & Zenner, 2000), benchmarking12 (Stabile, 2002; Bryant, 1998), CEO influence on hiring and firing of compensation advisors13 (Bebchuk et al., 2002; Loomis, 2001), and the shareholder primacy norm14 (Stabile, 2002).

company’s CEO was compensated fivefold of the organization’s lowest -paid full-time worker. However, in order to be able to attract a new external CEO in 1995, the hiring committee had to let go of the salary ratio. As Weiss (1995) concludes, compet itive concerns impede companies to maintain their own compensation limit.

6_{Tournament Theory was firstly introduced by Lazear & Rosen (1981). It views CEO compensation as the tournament prize that}

is in fact not based on a CEO’s actual productivity, but merely functions as an incentive for competition.

7

It represents a compensation model in which the position and thus not the person determines the level of compensation. They explain that the CEO is most likely only marginally more talented than the person occupying the position beneath him, but nevertheless will be paid disproportionately.

8

According to Kay (1998), ten to twenty percent of corporations replaces their CEO every year. He also shows an average CEO tenure of three years in major companies. M ore than ever, a CEO is prone to losing his position when company results do not match shareholders’ expectations (Blair, 1995; Denis & Denis, 1995; Lederer & Weinberg, 1995).

9

However, empirical research performed by M artin (1982) did not clearly indicate to what extent the decline in unionization is responsible for the increasing pay dispersion.

10

O’Reilly et al. (1988) argues that the compensation of the remuneration committee predicts a CEO’s pay more accurately than the above described economic theories. He noted that social comparison, especially amongst senior executives, is a more accurate indicator for CEO compensation than the firm performance.

11

According to Anderson & Bizjak (2000), it is for this reason that a CEO’s pay does not differ depending on whether or not he sits in the compensation committee.

12_{Due to the absence of any reference of CEO pay internally the only referent is external, also known as competitive}

benchmarking (Stabile, 2000). Competitive benchmarking is the process of companies hiring compensation consultants to perform market research in order to generate a compensation proposal. Stabile (2000) argues benchmarking as a significant cause of the upward spiral of CEO pay. Bryant (1998) adds that only a few companies are willing to express that their CEO is worthless than the average. Kah an (2001) found boards strive to compensate their CEO no less than the fiftieth to seventy -fifth percentile of what similar firms pay. Correspondingly, Bizjak (2000) indicates that competitive benchmarking substantially contributed to t he enormous increase in CEO compensation in the 1990s.

13_{For almost all companies the compensation survey performed by consultants is designed such that it selects only peer}

companies all at or above the industry’s average compensation level (Bizjak, 2000). This clearly indicates a CEO’s influence on consultants. Additionally, consultants deal with the demand to keep management satisfied as reason of consulting firms having other profitable exchanges with the hiring company.

14_{Shareholder primacy norm is related to corporate governance. It theorizes shareholder (economic) interest as primary relative to}

all other company stakeholders. Stabile (2002) argues it is responsible for the vast pay gap due to the resulting massive surge in the use of incentive pay as part of executive compensation. M ehran & Tracy (2001) noticed new stock option grants as part of a CEO’s compensation are worth on average 250% of the paid base salary and bonus.

(13)

13

2.4 Hypothesis development

The hypothesis tested in this research examines whether the level of CEO pay affects satisfaction experienced by a company’s employees. Goodman (1974) shows that employee satisfaction is greatly influenced by pay satisfaction. Festinger (1954), Major & Forcey (1985), Messé & Watts (1983), Patchen (1961), Goodman & Friedman (1971) show comparisons of pay have a great influence on pay satisfaction. Correspondingly, Pfeffer (1993) shows workforce’s satisfaction levels are directly related to the pay gap within organizations. More specifically, Ambrose & Kulik (1988) argue that a vast pay gap is detrimental since it adversely affects numerous aspects of employee satisfaction, including employees’ morale, productivity, absenteeism and turnover.

Dornstein (1989), Schwab & Wallace (1974), Martin & Murray (1983) and Adams (1965) show that workers compare their earnings both with peers and non-peers, without taking fairness into account (Festinger, 1954). In other words, comparisons are not bound by hierarchy and therefore are made with different people of different levels of hierarchy. This corresponds to the relative deprivation theory describes that an existing pay gap between two dissimilar groups lead to feelings of inequality (Martin, 1982).

Interestingly, Ambrose & Kulik (1988) show that availability of information determines a worker’s reference group. In practice, employees generally do not discuss compensation with their colleague’s whilst no disclosure regulation exist around remuneration below top management. Simultaneously, media attention usually given to senior executive and in particular CEO remuneration, resulting in workers tending to adopt them as a reference group (Akerlof & Yellen, 1990). Cook & Hegtvedt (1983) argue in accordance that remuneration levels of top management influence workers’ perception of fairness of pay.

In addition, the pay gap has been widening exponentionally over the years. On average a CEO’s pay amounted to 30:1 in 1980 to that of the average worker (Phillips, 1991), 150:1 in 1990 Crystal (1991) and nowadays that is equivalent to 331:1 (AFL-CIO, 2014).

Along with the literature described above and the tremendous increase in pay disparity (Crystal, 1991; AFL-CIO, 2014), this study expects the level of CEO pay to be negative ly related to employee satisfaction. The following hypothesis is derived in order to empirically assess whether the level of CEO pay has a negative effect on employee satisfaction:

(14)

14

3. Research Methodology

This section embodies three paragraphs. The section starts with a description of the various sources of data used for the analysis. Subsequently, the sample selection is outlined in detail. The section ends with a detailed description of the regression model used in this study.

3.1 Data

3.1.1 Employee satisfaction

In this study it is chosen to use the works of Great Place To Work Institute (GPTWI). This institute annually publishes a list of the hundred best companies to work for. In other words, companies that have the highest employee satisfaction amongst their workers. The first publication of the list was in 1998 and it has been published every year since, resulting in 18 editions. 368 individual companies15 out of every industry all were part of the list at least once.

GPTWI conducts the most extensive emp loyee satisfaction survey in USA. It scopes 81,000 employees of all industries. A company needs to be in operation for at least seven years if it wants to be part of the survey. Also, a thousand or more U.S. workers need to be employed.

Trust represents the center of the institute’s measure of employee satisfaction. It reviews companies to what extent their workers trust their executives, experience pride whilst working and have meaningful relationships with their colleagues. GPTWI conducts elaborate tests in order to produce the employee satisfaction score. Roughly 65% of a company’s satisfaction score is based on a survey which is sent to a minimum of 400 randomly selected employees. The remainder of the score is derived by the ‘Culture Audit’, which includes questions about demographics, economic benefits, and on philosophy and communication (Great Place To Work Institute, 2015).

15

Companies are likely to be on the list for more than one year. On average, a company appearing on the GPTWI list will return four more times (1800/368 = ±5).

(15)

15

3.1.2 CEO remuneration

Data on CEO remuneration have been acquired from COMPUSTAT’s subsection ExecuComp16. The period of which this data is retrieved parallels GPTWI list’s years of publication, starting from 1998 to 201417. The data comprise individual CEO remuneration of each year for each company. CEO remuneration is the total sum of base salary, bonus, other annual payout, restricted stock grants, long-term incentive payouts, value of options granted valued by the Black-Scholes formula18 following Bebchuk et al. (2006) and Fredrickson & Davis-Blake (2010), and forms of pay classified as ‘all other’. All pay variables are adjusted for inflation to a constant dollar basis. The resulting data base totals a pool of 43,234 observations from 501 number of companies (see Table 1).

3.1.3 Control variables

All data on control variables have been retrieved from COMPUSTAT’s annual fundamentals. Similarly, all data on the variables are acquired from the period 1998 to 2014.

3.2 Sample selection

The set up of this study comprises two samples. One representing the Main group and the other representing the Control group. Companies in the study’s Main group are solely those which have been published on GPTWI’s annual list of companies with the highest employee satisfaction score.

In total 368 single companies were listed over all eighteen years filling 1800 accounts (see Table 1). In other words, companies on the list are likely to return for ± 5 years (1800/368). Unfortunately, not all companies on the GPTWI list are public. In general, non-public companies are not obliged to disclose executive remuneration. As not sufficient information is available of 198 companies, these companies have been excluded from the main sample. Consequently, the study focuses on public firms. Of the remaining firms, from 47 companies not enough

16

ExecuComp is a subsection of COM PUSTAT, which is part of Standard & Poor’s that provides data on U.S. corporate executive remuneration.

17

Despite GPTWI published the list already for 2015, all annual company data is not complete for 2015. Therefore, the year 2015 is excluded.

18

(16)

16 remuneration data are available to have them part of the sample. As a result, the Main group totals 123 companies (see Table 1)

Remarkably, 80 out of 123 (65%) companies from the Main group are S&P50019 companies. Therefore, Control group companies are chosen from the S&P500. Essentially, control group consist of companies of the S&P500 which have not appeared on the GPTWI list in any of the years. As a result, Control group selected companies match the Main group in terms of size and kind. Moreover, the S&P500 broadly represents the U.S. market and therefore provides the same variety in the Control group present in the Main group.

Table 1 Sample sizes Main group and Control group

Main group

GPTWF total 368

-non public companies (198)

-public companies without data (47)

Total 123

Control group

S&P500

-companies part of Main group -companies without complete data

Total 500 (80) (42) 378 Study total 501 19

The Standard & Poor’s 500 is an U.S. stock market index that groups the largest companies on the NYSE or NASDAQ determined by market capitalization.

(17)

17 3.3 Research method 3.3.1 Econometric model where

LISTED A dummy variable which equals 1 if the company appears on the GPTWI list and 0 otherwise

logPAY Average total CEO remuneration of each company over the years 1998 to 2014 transformed to a log scale

logMCAP Average total market capitalization of each company over the years 1998 to 2014 transformed to a log scale

logROA Average return on assets of each company over the years 1998 to 2014 transformed to a log scale

GENDER Proportion male and female CEO of each company over the years 1998 to 2014

S1 to S10 A dummy variable which assigns each company to an industry

Any firm-specific characteristics that affect the variables are controlled for by using data from individual companies, and not industries. All variables used for regression purposes are averaged for each individual company from year 1998 to 2014. Consequently, use of original panel data is avoided which prevents disturbances of correlation across the years for the same firm. For analysis purposes, a binary logistic regression is used. Output in the form of coefficients in combination with a z-score are chosen to indicate any significance instead of the default likelihood-output of a binary regression. Factors that predict employee satisfaction are not included in the regression in order to prevent endogeneity.

(18)

18

3.3.2 Measures

LISTED – Dependent variable

Companies are recorded as listed when they appeared at least once on the GPTWI’s list of highest employee satisfaction. LISTED functions as a dummy variable which equals 1 if the company appears on the GPTWI list at least once and 0 otherwise.

logPAY – Independent variable

Data on CEO remuneration are retrieved from ExecuComp. logPAY represents a firm’s average over the period of 1998 to 2014. All data on CEO remuneration are transformed to a log scale. logPAY is the total sum of base salary, bonus, other annual payout, restricted stock grants, long-term incentive payouts, value of options granted valued by the Black-Scholes Model following Bebchuk et al. (2006) and Fredrickson & Davis-Blake (2010), and forms of pay classified as ‘all other’. All pay variables are adjusted for inflation to a constant dollar basis.

Control variables

Following Bloom & Michel (2002) both firm size and firm performance are controlled for. The log of the total market capitalization is used to proxy firm size. In order to control for firm performance, this study uses the log of the return on assets. This measure is used to proxy firm performance in the studies of Carpenter & Sanders (2002) and Henderson & Fredrickson (2001). Following Pfeffer (1993), this study uses return on assets calculated as annual operating income after depreciation divided by the company’s total assets. According to Wang et al. (2011), transformational leaders are most likely to be female. Additionally, transformational leaders20 are positively associated with employee satisfaction (Wang et al. 2011). The included variable

GENDER functions as an average which measures the weight of male CEOs for each of the

companies over the period 1998-2014. A value of 1 shows there were exclusively male CEOs at the company over the period, whereas 0 tells there were exclusively women seated as CEO. All numbers in-between display the proportion of male and female CEOs of the company over the years. Finally, following Pfeffer (1993) industry is controlled for. In order to assign each

20

Transformational leadership is considered to be a form of leadership where the leader guides the organization through all stages of change, from determining the needed change to shaping and finding support for a vision to implementing the change in accordance with all stakeholders.

(19)

19 company to an industry, this study uses GICS21 as industry taxonomy. In total ten different sectors are assigned and are transformed to dummy variables for regression purposes.

21

The Global Industry Classification Standard represents an industry classification method set up by Standard and Poor’s that distinguishes ten sectors.

(20)

20

4. Empirical Results

This chapter presents the results from the analysis performed. Descriptive statistics and the intercorrelation are presented in Table 2. Table 3 displays the results of the three regressions performed.

Table 2 provides an overview of the means, standard deviations and a correlation matrix of the variables. The mean LISTED of 0.25 only indicates the weight of GPTWI companies in the total sample and consequently does not reveal anything statistically. CEO pay, Market Cap and Return on assets are all three transformed to a logarithmic scale. Therefore, the mea n statistics of these variables cannot be interpreted. Gender shows that on average 0.93 out of 1 a male CEO operated over the period 1998-2014. Finally, the industry dummies are not included in Table 2 since there were not any significant correlations wit h other variables and the means solely indicate the weight in the sample.

Interestingly, the adjacent correlation matrix shows that Market Cap positively correlates to CEO pay. This means that the larger the company the larger the CEO compensation. Furthermore, Market Cap significantly correlates to CEO pay with a ratio of 0.69. This obviously has consequences for the performed regressions due to multicollinearity. Therefore, a separate regression (3) is performed with the exclusion of market capitalizatio n.

Table 2 Descriptive Statistics and Correlations

Variable Mean Std. dev. 1 2 3 4 1. LISTED 0.25 0.43 2. CEO Pay 15.98 0.64 -0.13 3. Market Cap 23.16 1.12 -0.08 0.69 4. ROA 2.00 0.86 0.08 -0.04 0.04 5. Gender 0.93 0.08 -0.06 0.02 0.06 -0.14

The sample consists of 43,234 observations for 501 firms collected from the period 1998-2014. All variables are averaged for each company over the period 1998-2014. LISTED represents companies that appeared at least once on the GPTWI’s list of highest employee satisfaction. LISTED functions as a dummy variable which equals 1 if the company appears on the GPTWI list at least once and 0 otherwise. CEO pay is the total sum of base salary, bonus, other annual payout, restricted stock grants,

(21)

long-21 term incentive payouts, value of options granted valued by the Black-Scholes Model, and forms of pay classified as ‘all other’. All pay variables are adjusted for inflation to a constant dollar basis. Market Cap represents the absolute market value of equity. ROA represents return on assets which is defined as operating profit after depreciation over total assets. Gender shows the proportion male and female in operation as CEO of each company over the period 1998-2014. The variables CEO Pay, Market Cap and Return on assets are transformed to a log scale.

Table 3 presents the estimates of three pooled binary logistic regressions of the model introduced in the previous chapter. The model regresses employee satisfaction on the level of CEO pay along with control variables. Significance is measured by the z-score shown between parentheses. The default output of a binary regression are odds ratios. However, this study chooses to report coefficients which is easier to interpret compared to odds ratios. The various levels of significance are indicated in the table for all var iables.

Column 1 displays the pooled regression results excluding control variables. The coefficient of CEO pay (logPAY) is negative and significant at the 1% level. The result is in line with this study’s hypothesis which expects that the level of CEO pa y is negatively related to employee satisfaction.

Column 2 reports the effect of CEO pay on employee satisfaction after controlling for market capitalization (logMCAP), return on assets (logROA), the gender proportion of CEO in operation (GENDER) and industry (S1 to S10). This study finds that the coefficient on CEO pay remains negative and continues to be highly significant. Employee satisfaction is positively associated with market capitalization similar to Carpenter & Sanders (2002) and Henderson & Fredrickson (2001). In correspondence to Pfeffer (1993), employee satisfaction is positively associated with return on assets. Furthermore, in accordance with Wang et al. (2011), employee satisfaction is negatively associated with male CEOs who are less likely to be transformational leaders. Industries’ coefficients are mostly positive. However, the coefficients of the control variables do not show significance to any degree. Finally, S8 to S10 are omitted from the regression because of collinearity.

Column 3 shows the pooled regression results exclusive of market capitalization, as market capitalization shows a significant correlation (0.69) with CEO Pay. Therefore, it is likely to cause disturbances due to multicollinearity. Accordingly, a third regression is performed without logMCAP. Consequently, by removing market capitalization from the regression CEO pay becomes even more significant. Other variables did not notably change in any way.

(22)

22 The results in Table 3 are consistent with Pfeffer (1993), Ambrose & Kulik (1988) and Wade et al. (2006) but inconsistent with Berkowitz et. al (1987) that proved pay comparisons are not contributive to employee satisfaction.

Table 3 Effect of the level of CEO pay on employee satisfaction

LISTED (dependent variable)

Variables (1) (2) (3) Constant 6.25 (2.35)** 8.27 (2.65)*** 8.53 (2.76)*** logPAY -0.46 (-2.76)*** -0.73 (-2.89)*** -0.60 (-3.39)*** logMCAP 0.11 (0.74) logROA 0.15 (0.82) 0.17 (0.96) GENDER -1.41 (-1.15) -1.45 (-1.14) S1 0.71 (1.17) 0.68 (1.12) S2 0.79 (1.18) 0.85 (1.28) S3 -0.21 (-0.27) -0.18 (-0.23) S4 0.82 (1.26) 0.87 (1.34) S5 0.91 (1.44) 0.93 (1.46) S6 0.39 (0.62) 0.39 (0.61) S7 0.83 (1.24) 0.79 (1.18) S8 (omitted) (omitted) S9 (omitted) (omitted) S10 (omitted) (omitted) Log likelihood -275.30 -243.13 -243.40

The sample consists of 43,234 observations for 501 firms collected from the period 1998-2014. All variables are averaged for each company over the period 1998-2014. The cell entries represent the regression coefficients with next to it z-scores between

parentheses. Regression:

; in which employee satisfaction

(LISTED) functions as the dependent variable and CEO pay (logPAY) as the independent variable along with the control variables market capitalization (logMCAP), return on assets (logROA), gender proportion of CEO in operation (GENDER) and the industry dummies classifying respectively Consumer Discretionary, Consum er Staples, Energy, Financials, Health Care, Industrials, Information Technology, Materials, Telecommunication Services and Utilities. S8, S9 and S10 are omitted from the regression because of collinearity.

*p<0.10 **p<0.05 ***p<0.01

(23)

23

5. Discussion and Conclusion

This study examines whether the level of CEO pay affects employee satisfaction. In the past decades, the magnitude of the pay gap between CEO and workforce has enormously increased (Crystal, 1991; AFL-CIO, 2014). Pfeffer (1993) shows that the larger the pay gap within organizations, the lower the workforce’s satisfaction level. Dornstein (1989) shows workers directly compare their pay with different people from different hierarchal levels. More specifically, Stabile (2002) argues that workers are inclined to select top management as their reference group because of the availability of information around their pay in contrast to other employees.

The empirical results support this study’s hypothesis. The findings clearly show that the level of CEO pay significantly negatively affects employee satisfaction. This means that the larger the pay gap between CEO and workforce, the less happy employees are. Also when control variables are included a significant negative relation remains to exist.

In sum, employee satisfaction is related to the level of CEO pay which is of relevance to companies, workers, unions and the government.

6. Limitations and Recommendations

The most obvious limitation relates to the used sample which consists only of public companies. Therefore, this study’s findings cannot be generalized to nonpublic, usually smaller organizations. However, from a social comparison perspective it is most likely that these firms experience the same effects and probably even more. This is because in small firms employees from different hierarchies generally work more closely together. Any pay gap will therefore more likely lead to friction amongst workers, adversely affecting satisfaction.

In addition, since GPTWI publishes a list of only the best hundred companies in terms of employee satisfaction, one cannot know for sure that Control group companies have been subject to examination by the institute. It is quite likely that companies being part of the S&P500 have been part of the GPTWI study as the institute conducts the most extensive employee survey in the

(24)

24 USA. Consequently, this study can be advanced by validating whether all companies were part of the GPTWI survey.

A final opportunity for future research concerns with pay comparison. This study uses the level of CEO pay to estimate how the process of pay comparison affects employee satisfaction. Due to a lack of company specific remuneration data, a pay ratio between CEO and the average worker could not serve as a measure. Industry averages of compensation provided by The Bureau of Labor Statistics22 could serve as an alternative, however, prior study has never combined industry averages of pay in combination with company specific CEO remuneration. It is likely to be of too less specificity in order to make conclusions on individual company level.

7. Acknowledgements

Jindi Zheng and Indu Panday provided conducive commentary on earlier versions of this article.

22

The Bureau of Labor Statistics is part of the U.S. Department of Labor which evaluates developments on the labor market, price changes in the economy and regulates working conditions.

(25)

25

8. Appendix

Figure 1 CEO-ave rage worke r pay ratio ove r the period 1960 -2000

Source: Executive Excess 2000, the 7th Annual CEO Compensation Survey from the Institute for Policy Studies and United for a Fair Economy.

(26)

26

9. Bibliography

Adams, J.S. (1965). Inequity In Social Exchange. Advances in Experimental Social Psychology,

2, 267-300.

AFL-CIO (2015). http://www.aflcio.org/

Ambrose, M.L. & Kulik, C.T. (1988). Referent Sharing: Convergence within Workgroups of Perceptions of Equity and Referent Choice. Human Relations, 41(9), 697-707.

Anderson, R.C. & Bizjak, J.M. (2000). An Empirical Examination of the Role of the CEO and the Compensation Committee does not depend on whether the CEO actually sits in the committee. SSRN Electric Paper Collection, No. 220851.

Baker, D. & Fung, A. (2001). Working Capital: The Power of Labor’s Pensions. Location: Cornell University Press.

Bebchuk, L.A., Fried, J.M. & Walker, D.I. (2002). Managerial Power and Rent Extraction in the Design of Executive Compensation. NBER Working Paper, No. 9068.

Bebchuk, L., Cremers, M. & Peyer, U. (2006). The CEO pay slice. Harvard Law and Economics

Discussion Paper, 574, 1-57.

Berkowitz, L., Crochan, S., Fraser, C. & Treasure, P. (1987). Pay, Equity, Job Gratifications, and Comparisons in Pay Satisfaction. Journal of Applied Psychology, 72, 544-551. Bizjak, J.M. (2000). Has the Use of Peer Groups Contributed to Higher Levels of Executive

Compensation? Unpublished Manuscript.

Blair, M.M. (1995). Ownership And Control: Rethinking Corporate Governance For The

Twenty-First Century. Location: Brookings Institution Press.

Bloom, M. (1999). The Performance Effects of Pay Dispersion on Individuals and Organizations.

Academy of Management Journal, 42(1), 25-40.

Bloom, M. & Michel, J.G. (2002). The Relationships among Organizational Context, Pay Dispersion, and Managerial Turnover. The Academy of Management Journal, 46(1), 33-42.

Brickman, P. (1975). Adaptation Level Determinants of Satisfaction With Equal and Unequal Outcome Distributions In Skill and Chance Situations. Journal of Personality and Social

Psychology, 31(2), 191-198.

(27)

27 Bryant, A. (1998, November 8). Executive Cash Machine: How Pliable System Inflates Pay

Levels. The New York Times, pp. 4-5.

Caldwell, D.F. & O’Reilly, C.A. (1982). Task Perceptions and Job Satisfaction: A question of causality. Journal of Applied Psychology, 67(3), 361-369.

Carpenter, M.A. & Sanders, W.G. (2002). Top management team compensation: the missing link between CEO pay and firm performance? Strategic Management Journal, 23(4), 367-375. Chen, Z., Huang, Y. & Wei, K.C.J. (2013). Executive Pay Disparity and the Cost of Equity

Capital. Journal of Financial and Quantitative Analysis, 48(3), 849-885.

Cook, K.S. & Hegtvedt, K.A. (1983). Distributive Justice, Equity, and Equality. Annual Review

of Sociology, 9, 217-241.

Cowherd, D.M., & Levine, D.I. (1992). Product Quality and Pay Equity Between Lower-Level Employees and Top Management: An Investigation of Distributive Justice Theory.

Administrative Science Quaterly, 37(2), 302-320.

Crosby, F. (1976). A Model of Egoistical Relative Deprivation. Psychological Review, 83(2), 85-113.

Crystal, G.S. (1991). In Search Of Excess: The Overcompensation Of American Executives. Location: W.W. Norton and Co.

Davies, J.C. (1962). Toward a Theory of Revolution. American Sociological Review, 27, 5-19. Denis, D.J. & Denis, D.K. (1995). Performance Changes Following Top Management

Dismissals. The Journal of Finance, 50(4), 1029-1057.

Deutch, M. (1985). Distributive Justice: A Social Psychological Perspective. Location: Yale University Press.

Dornstein, M. (1989). The Fairness Judgments of Received Pay and Their Determinants. Journal of Organizational Psychology, 62(4), 287-299.

Dyer, L. & Theriault, R. (1976). The determinants of pay satisfaction. Journal of Applied

Psychology, 61(5), 596-604.

Festinger, L. (1954). A theory of social comparison processes. Human relations, 7, 117-140 Finkelstein, S. & Hambrick, D.C. (1989). Chief Executive Compensation: A study of the

intersection of markets and political processes. Strategic Management Journal, 10(2), 121-134.

(28)

28 Finn, R.H. & Lee, S.M. (1972). Salary Equity: Its Determination, Analysis, and Correlates.

Journal of Applied Psychology, 56(4), 283-292.

Frank, R.H. & Cook, P.J. (1995). The Winner-Take-All Society. Location: Penguin Books. Fredrickson, J.W. & Davis-Blake, A. (2010). Sharing The Wealth: Social Comparisons and Pay

Dispersion in the CEO’s Top Team. Strategic Management Journal, 31, 1031-1053. Galbraith, J.K. (1998). Created Unequal: The Crisis in American Pay. Location: University of

Chicago Press.

Goodman, P.S. (1974). An Examination of Referents Used In the Evaluation of Pay.

Organizational Behavior and Human Performance, 12(2), 170-195.

Goodman, P.S. & Friedman, A. (1971). An Examination of Adams’ Theory of Inequity.

GPTWI (2015). http://www.greatplacetowork.com/

Gurr, T.R. (1970). Why men rebel. Location: Princeton University Press.

Hellman, C. (1997). Job Satisfaction and Intent to Leave. The Journal of Social Psychology, 137(6), 677-689.

Henderson, A.D. & Fredrickson, J.W. (2001). Top Manage ment Coordination Needs and the CEO Pay Gap: A Competitive Test of Economic and Behavioral Views. The Academy of

Management Journal, 44(1), 96-117.

Hofstede, G. (1980). Culture’s consequences: international differences in work -related values. Location: Sage cop.

Homans, G.C. (1961). Social Behavior: Its elementary forms. Location: Harcourt. Hyman, R. (1942). Marxism and the sociology of trade unionism. Location: Pluto Press.

Jaques, E. (1961). Equitable payment: a general theory of work, differential payment and

individual progress. Location: John Wiley.

Kahan, M. (2001). The Limited Significance of Norms for Corporate Governance. University of

Pennsylvania Law Review, 149(6), 1869-1900.

Kay, I.T. (1998). CEO Pay and Shareholder Value: Helping the U.S. Win the Global Economic

War. Location: CRC Press.

Lambert, R.A., Larcker, D.F., & Weigelt, K. (1993). The Structure of Organizational Incentives.

(29)

29 Lawler, E.L. (1971). Pay and organization effectiveness: a psychological view. Location:

McGraw Hill.

Lazear, E.P., & Rosen, S. (1981). Rank-Order Tournaments as Optimum Labor Contracts.

Journal of Political Economy, 89, 841-864.

Lazear, E.P. (1995). Personnel economics. Location: MIT Press cop.

Lederer, J.L. & Weinberg, C.R. (1995, September). Harnessing Corporate Horsepower: The New Principles of CEO Compensation. Chief executive, 34, 30-32.

Leventhal, G.S. (1976). The distribution of rewards and resources in groups and organizations.

Advances in Experimental Social Psychology, 9, 92-131.

Locke, E.A. (1976). The Nature and Causes of Job Satisfaction. Location: Rand McNally. Loomis, C.J. (2001, June). This Stuff Is Wrong. Fortune. Retrieved from

http://archive.fortune.com/magazines/fortune/fortune_archive/2001/06/25/305435/ index.htm

Major, B. & Forcey, B. (1985). Social Comparisons and Pay Evaluations: Preferences For Same Sex and Same-Job Wage Comparisons. Journal of Experimental Social Psychology,

21(4), 393-405.

Martin, J. (1982). The Fairness of Earnings Differentials: An Experimental Study of t he Perceptions of Blue-Collar Workers. Journal of Human Resources, 17(1), 110-122. Martin, J. & Murray, A. (1983). Distributive injustice and unfair exchange. Location: Plenum. Mehran, H. & Tracy, J. (2001). The Impact of Employee Stock Options on the Evolution of

Compensation in the 1990s. NBER Working Paper, No. 8353.

Messé, L.A., Watts, B.L., & Manis, M. (1983). Complex nature of the sense of fairness: Inter nal standards and social comparison as bases for reward evaluations. Journal of Personality

and Social Psychology, 45(1), 84-93.

O’Reilly, C.A., Main, B.G. & Crystal, G.S. (1988). CEO Compensation as Tournament and Social Comparison: A Tale of Two Theories. Administrative Science Quaterly, 33(2), 257-274.

Parry, J. & Bloch, M. (1989). Money and the morality of exchange. Location: Cambridge University Press.

(30)

30 Pedersen, T. & Thomsen, S. (1997). European Patterns of Corporate Ownership: A Twelve

Country Study. Journal of International Business Studies, 28(4), 759.

Perry, T. & Zenner, M. (2000). CEO Compensation in the 1990s: Shareholder Alignment or Shareholder Expropriation? Wake Forest Law Review, 35, 123-136.

Pfeffer, J. & Langton, N. (1993). The Effect of Wage Dispersion on Satisfaction, Productivity, and Working Collaboratively: Evidence from College and University Faculty.

Pfeffer, J., & Langton, N. (1994). Paying the Professor: Sources of Salary Variation in Academic Labor Markets. American Sociological Review, 59(2), 236-256.

Phillips, K. (1991). The Politics of Rich and Poor. Location: Harpercollins.

Ronen, S. (1986). Equity Perception in Multiple Comparisons: A Field Study. Human Relations,

39(4), 333-345.

SEC (2014). http://www.sec.gov/

Spector, P. (1997). Job Satisfaction: Application, Assessment, Cause and Consequences. Location: Sage Publications.

Stabile, S.J. (2000). Viewing Corporate Executive Compensation Through a Partnership Lens: A Tool to Focus Reform. Wake Forest Law Review, 35, 153-230.

Stabile, S.J. (2002). One For A, Two For B, And Four Hundred For C: The Widening Gap In Pay Between Executives And Rank And File Workers. University of Michigan Journal of Law

Reform, 36(1), 115-182.

Taunton, R.L., Boyle, D.K., Woods, C.Q., Hansen, H.E. & Bott, M.J. (1997). Manager Leadership and Retention of Hospital Staff Nurses. Western Journal of Nursing Research,

19, 205-226.

Todorova, G., Bear, J.B., & Weingart, L.R. (2014). Can Conflict Be Energizing? A Study of Task Conflict, Positive Emotions, and Job Satisfaction. Journal of Applied Psychology, 99(3), 451-467.

Tolbert, P.S., & Zucker, L.G. (1983). Institutional Sources of Change in the Formal Structure of Organizations: The Diffusion of Civil Reform, 1880-1935. Administrative Science

Quaterly, 28, 22-39.

Tosi, H.L., & Greckhamer T. (2004). Culture and CEO Compensation. Organizational Science,

(31)

31 Wade, J.B., O’Reilly, C.A. & Pollock, T.G. (2006). Overpaid CEOs and Underpaid Managers:

Fairness and Executive Compensation. Organizational Science, 17(5), 527-544.

Wallace, M.J., & Schwab, D.P. (1976). A cross- validated comparison of five models used to predict graduate admissions committee decisions. Journal of Applied Psychology, 61(5), 559-563.

Wang, X., Contawan, R. & Nantusupawat, R. (2011). Transformational leadership: effect on the job satisfaction of Registered Nurses in a hospital in China. Journal of Advanced

Nursing, 68(2), 444-451.

Weiss, M. (1995, April). Ben & Jerry’s Executive Pay Scales Likely to Take an Upward Turn.

Times Union, 6, 12-14.

Weiss, H.M., & Cropanzano, R. (1996). Affective Events Theory: A theoretical discussion of the structure, causes and consequences of affective experiences at work. Research in organizational behavior: An annual series of analytical essays and critical reviews, 18, 1-74.

Wink, P.S. (1991). Two Faces of Narcissism. Journal of Personality and Social Psychology,

An empirical analysis of the relation between the level of CEO pay and employee satisfaction