The impact of national culture on firm compensation practices : does national origin affect compensation practices differences between European firms that cross-list in US and US listed firms?

(1)

The impact of national culture on firm compensation practices:

Does national origin affect compensation practices differences between

European firms that cross-list in US and US listed firms?

Name: Xi LIU

Student number: 11582421 Thesis supervisor: Peter Kroos Date: June 24th_{, 2018}

Word count: 13,595

MSc Accountancy & Control, specialization Control

(2)

Statement of Originality

This document is written by student Xi LIU who declares to take full responsibility for the contents of this document.

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

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

(3)

Abstract

In response to the call for further research to explore the relationship between national culture and firm compensation practices (Greckhamer, 2011), this paper is aimed to investigate how national culture influence firm compensation practices, specifically, how the firm’s national origin influence two dimensions of firm compensation practice, which are CEO annual compensation and the ratio of CEO pay to the pay of median worker (CEO pay ratio) respectively. Since each multinational firm would employ solutions that reflect he values of the country in which it domiciles (Shetty, 1979), this paper seeks to deepen the understanding of firm compensation practices by examining the relationship between the national country origin of firm and CEO annual compensation as well as CEO pay ratio. I conduct a comparative empirical study of US listed firm with headquarters in Europe relative to US listed firms that are headquartered in the US. However, my regression results don’t reflect a significant impact of firm national origin on firm compensation practice. This insignificant result may imply that multinational firms in a global context might have to comply with local cultural values, customs, and compensation practices in order to attract desired quality and quantity of local labor. I further discussed some reasons and implications for this insignificant result and provide some suggestions for future research directions in the final part.

(4)

1 Introduction

1.1 Background

With the separation of corporate management and ownership, agency problems arise between the management and business owners. According to the agency theory, the business owner will determine executive compensation through observable performance indicators, thus motivating and rewarding the executives for their contribution to the company, and bringing the executives’ actions in line with shareholders’ interests through effective rewarding incentives to avoid the executives’ self-interested behavior. Thus, a proper compensation distribution system in practice can be a powerful incentive mechanism.

The compensation practices within the enterprise, including the distribution of remuneration within the management team and the remuneration distribution between the management and ordinary workers, involves issues of fairness and efficiency and is becoming one of the hot issues of the media and academia in recent years. Many discussions were made on this topic and the research perspectives have become very pluralistic since the Wall Street occupation. Some scholars have conducted research and analyses on the current situation of executive pay gaps and found that executive pay is continuously growing. Through research on 72 American manufacturing companies from the year 1964 to 1981, Murphy (1985) found that vice presidents who were promoted to president or CEO received an average compensation increase of 20.9% or 42.9% respectively, indicating that the average compensation changes realized upon promotion was quite huge. Jarque (2008) summarized that the average compensation of a CEO working in one of the Top 500 firms in US has increased by six times from 1988 to 2008. According to a report by Mishel and Schieder (2017) from Economic Policy Institute, CEO annual compensation has increased a lot over the past 20 years, as shown in the Exhibit 1.

Several scholars have provided explanations for the increasing level of executive compensation. Some researchers have pointed to the increase in firm size as a factor which can influence executive compensation, because increasingly larger firms require more talented mangers to run those firms, and CEO compensation goes up accordingly. Others have looked at CEO managerial power and criticized greedy and powerful managers for the rise in executive compensation (Murphy, 2002; Fried and Bebchuk, 2003; Lin and Lu, 2009).

(7)

One way how the excessive executive compensation is manifested is the gap between CEO pay and that of ordinary employees. Consequently, in August 2015, the Securities and Exchange Commission (SEC) offered additional guidance on CEO pay ratio disclosures and determined that, starting in January 2017 under the Pay Ratio Disclosure Rule (PRDR), every US public listed firms must publish the ratio of the pay of its CEO to the pay of the median worker (referred to CEO pay ratio) annually in its proxy statement.

Exhibit 1 - CEO average annual compensation (in thousand dollars) 1

* CEO average annual compensation for the top 350 US firms ranked by revenues.

Research on the factors affecting the compensation practices of enterprises is of significance. Better understanding of the factors affecting the compensation practices can help develop a better understanding of remuneration distribution systems in reality.

Not so much research up to now has examined the impact of national culture on firm compensation practices. Based on comparative analysis of compensation packages in American firms, French firms, and Dutch firms, Pennings (1993) found that American executives understood compensation contracts differently relative to those executives employed by French firms and Dutch firms, and argued that the difference could be partially explained by national cultural differences between US and Europe. Aimed at comparing compensation contracts of car retailers in China, US, and the Netherlands, Merchant et al.’s (2011) found that Chinese and US firms tended to use more incentive compensation than

1_{Source: Lawrence Mishel and Jessica Schieder, July 20, 2017, CEO pay remains high relative to the pay of typical workers}

and high-wage earners, Economic Policy Institute.

843 1,102 1,508 2,808 5,947 20,664 19,112 10,746 12,827 12,958 15,261 15,60316,569 16,341 0 5000 10000 15000 20000 25000 1965 1973 1978 1989 1995 2000 2007 2009 2010 2011 2012 2013 2014 2015

(8)

Dutch firms did, and that Chinese firms made greater use of subjective bonus plans relative to US firms. In a cross-cultural research to study relation between compensation and culture, Greckhamer (2011) found that great pay equity between executives and ordinary employees could be attributed to national cultural and environmental factors from the macroscopical perspective. Similarly, when examining executives’ tournaments, Burns et al. (2017) found that the relation between firm performance and compensation differences within the management team could be somewhat attributed to cross-national cultural differences. The national culture of a country plays a dominant and crucial role in shaping a firm’s organizational culture in which they operate (Lindholm, 2000).

Some scholars applied Hofstede's national cultural framework to investigate the relation of national culture and compensation. Gomez-Mejia and Welbourne (1991) argued that national culture played a vital role in the development process of compensation system, and the efficacy and efficiency of remuneration strategies depended largely on the accordance with national culture, especially for multinational corporations. They further concluded that Hofstede's (1980) four cultural dimensions, which included individualism, uncertainty avoidance, masculinity/femininity, and power distance, influenced development process of firm’s compensation system. Based on cross-cultural studies in 24 countries, Schuler and Rogovsky (1998) analyzed the association between the Hofstede (1980)’s cultural dimensions and four types of compensation practices based on performance, social benefits and programs, state and employee ownership plans, and their results also indicated that national culture has played a crucial role in explaining differences of firm compensation practice in different countries. Tosi and Greckhamer (2004) also applied Hofstede’s models to investigate how national culture influence CEO pay, and they found that total annual CEO compensation, the proportion of variable pay to total compensation, and the ratio of CEO pay to the pay of the lowest level employees were all connected with power distance, and total annual CEO compensation and the ratio of variable pay to total compensation were related to individualism.

According to discussion above, one might expect that national culture can affect firm compensation practices. Given the limited research on how national cultural differences impact firm compensation practices, my thesis is aimed to study the impact of national culture on firm compensation practices.

(9)

1.2 Research question

In response to the call for more research to explore how culture influence firm compensation practices (Greckhamer, 2011), my thesis investigates the impact of national culture on firm compensation practices.

Therefore, my research question is:

RQ: How does national culture affect firm compensation practice?

1.3 Relevance of the research question

This study aims to contribute to the academia and practical business from the following three aspects.

First, it adds to the literature which studied the determinants that influence firm compensation practices. While most studies focused on individual and organizational level variables which can impact firm’s compensation practices, intuitively, it is reasonable to deem that including national culture as a macro level variable that influences compensation practices should be useful and beneficial.

Second, this study aims to contribute to prior studies which examined the effect of national culture. While some studies have looked at culture, not so many studies have related culture to compensation practices. So, my study contributes by examining how national culture specialty impact compensation practices within corporations.

Thirdly, this study also features societal contribution. Firm compensation practice plays a decisive role in practical human resource management. A good and effective compensation system will help improve the employees' job satisfaction and performance, thereby enhancing the whole company’s competitiveness and promoting its development. By examining on the relation between culture and compensation practice, it is suggested that multinational corporations take cultural differences into consideration when designing pay system. Thus, research about the determinants of compensation practice is of importance for a company, especially a multinational company.

1.4 Structure of thesis

The remainder of this paper is organized as follows. The second section reviews prior literature and develops my hypothesis. In the third section, the sample, empirical models, and

(10)

variable measurement are described, and the fourth section presents and discusses the empirical results and robustness check. The final section will provide conclusions and discussions, and suggest future research possibilities.

(11)

2 Theories and literature review

2.1 Compensation

2.1.1 Roles of compensation

The system of remuneration is one of the most important issues in Human Resource Management and it mainly serves as a way to incentivize employees to work in line with the whole company’s interests. The incentive compensation is usually expressed as salary, bonuses, rewards, prizes, and other material benefits to employees who perform well. These are all served as “positive incentive”. In contrast, for some badly-behaved employees, the organization can also employ “negative incentives” by reducing the amount of bonuses, wages and other material benefits. Therefore, compensation is argued to provide effort-inducing incentives. Lewellen et al. (1987) examined the incentive role of compensation, and they argued that the firm executive pay package is designed to reflect an attempt to prevent and control agency problems such as horizon problems and risk exposure problems, thus to lower agency costs between the principal and the agent. Similarly, applying panel data set from auto racing, Becker and Huselid (1992) proved that the tournament spread played an incentive role on individual performances. By examining the behavior of around three thousand workers across a nineteen-month period in a large auto glass firm, Lazear (2000) found that the average level of output per worker increased significantly after changing the compensation method from hourly pay (pay based on working hours) to piece-rate pay (pay based on job performance).

Compensation management can also be useful and helpful in achieving employee retention and job satisfaction. Morice and Murray (2003) investigated the teacher assessment and compensation plan for the Ladue School District in St. Louis, and they found that salary increase based on performance evaluations could improve job satisfaction and teacher’s retention rate. Osibanjo et al. (2014) explored how compensation packages influence teaching staff’s job performance and retention at a private university, and found that compensation packages were closely related to job performance and retention. Those empirical results imply that, in order to enhance employee’s job performances and to prevent high employees’ turnover, the management should endeavor to review and design appropriate compensation plans for employees at various levels.

(12)

In addition, compensation can serve as selection role. Existing research basically hold that highly-profiled, non-risk-averse employees will self-select towards well-paid jobs, while nonqualified, risk-averse employees will be more likely to leave. Ahn and Lee (2000) argued that an output-based contract offered in a labor market can influence job performance by attracting employees with certain personal attributes to the company (self-selection effects). In Godes and Mayzlin’s (2012) study of “using the Compensation Scheme to Signal the Ease of a Task”, they also explained that the firm could use a compensation package to signal to the sales people the “ease” with which his or her efforts will translate into high revenues of the firm’s products.

2.1.2 Related theories about compensation

(1) Efficient contracting theory

Efficient contracting theory, which is also generally called optimal contracting theory, refers to a compensation contract that achieves optimal level under a realistic constraint conditions. The efficient contracting theory generally hold that the board of directors should design efficient compensation contracts to attract talented executives and incentivize them to act in the firm’s best interest so as to achieve maximization of shareholder’s value.

In the theoretical analysis of the efficient contracting theory, scholars such as Mirrless (1976), Lazear and Rosen (1981), and Grossman and Hart (1983) used mathematical analysis methods to elaborate the compensation efficient incentive contracting model. The representative one, Mirrless’s (1976) study proposed a one-time contract static gaming model for how the principal should set up the reward and punishment mechanism according to the agent's action. This kind of compensation incentive mechanism is also called the explicit incentive mechanism. Fama (1980) argued that the long-term agency relationship could take advantage of the “reputation effect.” He deemed that in the competitive market, professional managers’ remuneration pricing depended on past operating performance. Even if there is no explicit incentive contract, they would also work hard to maintain accumulated personal reputation.

However, some research pointed out that the efficient contracting theory was of limited use in practice. Jensen and Murphy (1985) argued that the efficient contracting theory did not explicitly explain the coefficient magnitude of executive compensation and company performance in practical applications, concluded that although the correlation between executive remuneration and firm performance is positive, the corresponding statistical

(13)

significant magnitude is quite small: based on CEO data of the 250 major US companies from the year 1974 to 1988, they found that CEO only earned additional 6.7 cents for each 1,000 dollar increase of shareholder wealth. However, Hall and Liebman (1998) argued that the small sensitivity of pay-for-performance can still translate into large dollar values when looking at large firms. Thus, executives in large firms are still motivated to receive million dollars from improving firm performances.

(2) Managerial power theory

The efficient contracting theory assumes that the contract signed by the management and the shareholders is based on an optimal compensation incentive plan for maximizing the interests of both parties, but the effective implementation of this theory requires certain prerequisites. If certain conditions cannot be satisfied at the same time, the executive compensation cannot reduce the agency costs, but lead to the agency problem.

The managerial power theory hold that executives can "bypass" the supervision of the board of directors through their managerial power to impact and control their own compensation. Lambert et al. (1993) concluded that CEO’s salary was higher if there were more directors appointed by CEO sitting in the board. This is because of lack of independence of board members, they tended to act in line with their own interests when they were appointed by the CEO. Boyd (1994) studied the effect of board control on executive compensation and pointed out that executive compensation was greater in firms with weaker board control. Core et al. (1999) argued that board characteristics and stock ownership structure explained a significant amount of differences in CEO annual compensation with other firm economic factors constant, and suggested that CEO annual compensation went higher if firm had greater agency problems and thus weaker governance structures. Bebchuk et al. (2002) developed a model regarding the managerial power and rent extraction on CEO pay, and suggested that CEO had power to control and impact their compensation and to use that kind of power to extract rents when board of directors were highly dependent on CEO. In Stuart and Robert (2004)’s study of the relationship between the CEO compensation and the existence of compensation committee, they found CEOs in companies that set up the compensation committee had lower salary and proved that the existence of firm compensation committee could effectively suppress the boost of CEO salary levels.

(14)

(3) Human capital theory

The human capital theory emphasizes on the agent’s skills and knowledge. In this theory, three types of human capital are distinguished, which are ordinary employee human capital, general manager human capital, and senior manager human capital respectively. Senior managers' human capital is relatively high-level human capital in the company because senior managers can provide professional management with superior quality resources for the company's production and operation. According to the human capital theory, under a reasonably functioning labor market, human capital is indeed a crucial determinant of both executives’ and employees’ compensation (Peng et al., 2015). Differences in the ability, responsibilities and required skills between executives and ordinary employees can explain their compensation gap.

With regard to research on determinants of compensation, the concern of the human capital theory is on the micro level, mainly including individual characteristics, especially one’s education, training and experience, which can influence corresponding compensation package. Becker (1964) argued that both employees’ compensation and managers’ compensation are strongly related to their education and experience. Fisher and Govindarajan (1992) investigated the impact of several human capital factors on PCM’s (profit center manager) compensation, found that PCM’s job tenure was positively correlated with their compensation, and pointed out the opinion that manager’s job tenure increased their power and expertise which lead to increased compensation. Carpenter et al. (2001) argued that CEOs who experienced international assignment performed better through their ability to control rare, valuable, and inimitable resources. By focusing on archival salary data of IT professional institutions, Ang et al. (2002) found that IT professional’s compensation is significantly dependent on their education level, training and professional experience. Similarly, based on archival salary data of 2,251 IT professionals in Singapore, Slaughter et al. (2007) found that IT professionals whose job required greater firm-specific human capital were paid higher than those whose job required less firm-specific human capital.

2.1.3 Determinants of CEO compensation

Core et al. (1999) summarized two categories of determinants of CEO compensation, which are economic factors and corporate governance factors. Economic determinants include firm performance, size, risk, and other factors; while corporate governance factors are mainly

(15)

related to board independence and ownership structure in the firm, which has been widely discussed in the managerial power theory.

Extensive research has been conducted on the relationship between enterprise performance and executive compensation. Murphy and Salter (1975) found significant positive correlation between return on asset (ROA) and executive compensation: executives’ compensation increased significantly with the increase in the company’s ROA. Based on a sample of executives in 73 large US companies, Murphy (1985) analyzed the relationship between executives’ compensation and shareholder profits and found that the executives’ compensation increases with firm’s operating income. Jensen and Murphy (1985) also found a statistical significant relation between executive compensation and firm performance. Kerr and Bettis (1987)’s study also showed that stock return was a major determinant that explained the level of executive compensation. Guillet et al (2013) also argued that the executive compensation in the US catering industry was significantly positively related to firm performance (proxied as Tobin’s Q in his paper).

Firm size is another major factor that can influence and explain executive compensation. Generally speaking, the greater the firm size is, the more resource for which CEO or senior managers can control and thus the more complicate the management issues involved, so the higher of executives’ ability requirements will be, leading to corresponding higher compensation that executives require. Moreover, large firms are informationally opaque and therefore actions of managers are more difficult to observe. Increasing moral hazard concerns evoke the need for incentive alignment by means of incentive compensation contracts. Most of the previous studies have confirmed the view that executive compensation increases with the expansion of firm size. Joseow et al. (1993) used sales scale, assets scale and the number of employees as measures for company size, and found a significant positive correlation between executive compensation and firm size. Through an empirical study of listed companies in Hong Kong, M. Firth et al. (1999) also found that firm size is a main factor determinant of executive compensation. Tosi et al.’s (2000) research on determinants of CEO compensation showed that enterprise size explained two-fifths of the CEO compensation variability, while enterprise performance just explained five percentage of the CEO's compensation variability. Guillet et al. (2013) also found that the size of enterprise was significantly and positively correlated with executive compensation.

Many empirical studies have also examined the impact of industry-level characteristics such as growth opportunity on compensation. Firms with high amount of grow opportunity are

(16)

more likely to apply incentive compensation plan because managers play an important role in selecting right investment projects. Clinch (1991) found that firms with high research and development expenditures and thereby high growth opportunity appeared to relate compensation rewards more closely with market or accounting based performance measures than did firms with low esearch and development expenditures. Smith and Watts (1992) found that companies with larger investment opportunity set (such as increased availabilities of positive net present value projects) had higher levels of executive compensation and used executive incentive compensation (stock-option) plan more extensively. Anderson et al. (2000) found that, in contrast to companies in traditional industry, companies in emerging industry with many growth opportunities and high risks give executives more incentive pay. Some other scholars investigated the impact of firm risk on CEO compensation. According to prior literature, there are typically two contradicting views concerning the relation of firm risk and executive compensation. On the one hand, it is deemed that when firms are characterized by high risk, managers would demand higher incentive compensation (risk premium) because of increased difficulty and effort put in management control. In this sense, it is expected to observe a positive correlation between CEO compensation and firm risk. Guay (1999) found that firms' stock return volatility (firm risk) was positively related to the convexity of the CEO's total compensation. Cheng et al. (2010) also found that a positive relationship exists between stock price volatility and executive residual and total compensation. On the other hand, it is argued that risk makes manager’s performance measures noisy, and thus incentive compensation decreases with the increase of firm risk because the noise in the performance measures makes incentive compensation costly. By examining a sample of 435 IPO firms, Beatty and Zajac (1994) found a “consistent inverse relationship between the levels of firm risk and the degree to which incentive compensation for top manager is used”. Similarly, based on investigation of an average of seventy-five randomly selected managers from each of 740 companies, Bloom and Milkovich (1998) implied that firms with higher risk relegated smaller percentages of total pay to bonuses and confirmed the hypothesis about the negative relationship between firm risk and the use of contingent compensation. Furthermore, some scholars argued that the relation between firm risk and executive compensation was non-linear. Miller et al. (2002) found that both the proportion of variable pay in CEO compensation packages and their magnitude were curvilinearly related to unsystematic firm risk. To sum up, there is no consensus of the

(17)

empirical results concerning this topic, but it can be ascertained that firm risk can have an influence on executive compensation.

To sum up, most theoretical and empirical research have investigated the influence of firm economic characteristics, corporate governance characteristics and executive’s individual characteristics on executive compensation. Some empirical research have confirmed that typical economic factors include firm performance, size, growth opportunity and risk have profound impact on executive compensation. The studies on managerial power theory have verified the effect of corporate governance characteristics upon executive compensation. Despite differences in research focus and empirical results, the research findings concerning corporate governance characteristics can be summarized as follows: in a well-structured company with reasonable board structure, board of directors can effectively exert the supervisory role on senior executives and suppress irrational growth of executive compensation. Furthermore, human capital theory has affirmed the impact of executive’s individual characteristics like job tenure on their compensation.

2.1.4 Determinants of compensation gap

Many scholars began to discuss the increasing income gap since the technological changes from the 1980’s. With regard to existing research on compensation gap, there are mainly two theoretical perspectives, which are tournament theory and behavioral theory respectively. The tournament theory was proposed by Lazear and Rosen (1981). In the early 1980’s, Lazear and Rosen (1981) proposed the opinion that different agents should be incentivized by relative performance and believed that the promotion within the organization was like a race to eliminate opponents continuously, and the final winner would receive job promotion and a much higher pay. The agent was viewed as a competitor in the promotion competition, which was deemed as a "Sequential Elimination Tournament". The theory is therefore called the tournament theory. They initially used it to explain the fact that the pay gap within executive members had been increasing with the increase in the position ranks. They believed that companies tended to use promotion to motivate people to work hard. The theory indicates that the pay gap will motivate managers to increase their efforts to obtain high-paying positions, thereby improving corporate performance. Green and Stokey (1983) pointed out that this type of tournament compensation model outperformed other compensation models when individual output was affected by common factors. Lambert et al. (1993) found that pay gaps were increasingly growing at higher managerial position ranks. Lin and Lu (2009)

(18)

examined the tournament theory and found that pay gaps are positively correlated with firm performance. Lin et al. (2011) conducted empirical research using data from 1,386 listed companies in China from the year 1999 to 2006 and found the pay gap between CEOs and other managers increased with the increase in the number of other managers. Early tournament theory was mainly used to explain the pay gap between agents at the same level (within executives), and was later expanded to explain pay gap of different levels of agents (between executives and employees).

On the contrary, the behavioral theory emphasizes more on fairness and teamwork, and advocates reducing pay gaps. The advocates of behavioral theory believe that more equitable compensation will help increase company value and increase internal cooperation, while excessive pay gaps will lead to lower employee satisfaction and demotivate employees. There are also some empirical research that validate the behavioral theory. Lazear (1989) pointed out that a high internal company pay gap could result in more fierce competition among employees, which imposed a negative impact on firm performance. Pfeffer and Davis-Blake (1992) studied the relationship between pay gap and turnover rate and found that the turnover rate of executives with lower pay levels was higher when the pay gap went larger. Similarly, Pfeffer and Langton (1993) studies the compensation gap in academia, and found that the greater the pay gap was, the lower the researcher’s job satisfaction and research productivity would be. Based on a study of the relative salary levels of employees at all universities in California, Card (2012) found that when employees were able to know their colleagues' compensation levels, employees with lower income levels were more likely to be influenced by their peers' compensation information.

Some researchers also argued that the tournament theory and behavior theory are complementary to explain the pay gap. Henderson and Fredrickson (2001) argued that although economic theory such as tournament theory was a better predictor to explain the magnitude of CEO pay gaps, there also existed a balance between the tournament perspective and behavioral view to predict and explain firm performance. It is believed that these two theories were established under different circumstances, that is, the relationship between firm performance and pay gap has a range effect.

2.1.5 Trends in compensation

It is conventionally held that CEOs in US firms are paid significantly more than CEOs in their foreign counterparts. Conyon and Murphy (2000) showed that, after controlling for

(19)

industry, enterprise size, and other aggregate factors and CEO individual characteristics, US CEOs earned nearly twice more than UK CEOs in 1997.

However, by 2003, Conyon, Core, and Guay (2010) showed that the US CEOs pay premium over UK CEOs had fallen to 40%, and potentially disappeared after adjusting for the risk associated with undiversified CEO equity portfolios. Fernandes et al. (2012) also challenged the opinion that US CEOs were overly paid by using CEO compensation data in 1,648 US companies and 1,615 non-US companies in their selected fourteen countries, and they found that the US pay premium is economically modest after controlling both for firm-specific characteristics such as industry, firm size, firm risk and firm performance, growth opportunities and for stock ownership and board governance which are systematically differ across.

It is globally accepted that CEO compensation is growing fast over the past several years. Using the US top 500 enterprises as samples, Jarque (2008) summarized that the CEO average pay has increased six-fold from 1988 to 2008. And he found that this dramatic pay increase is accompanied by increasing ratio of performance-related pay such as stock grants and stock option grants in CEO’s compensation package.

According to Economic Policy Institute’s report (2017), it is indicated that the trend of US top 350 firms’ CEO compensation has historically corresponded with that of the stock market as measured by the S&P 500 index, as shown in exhibit 2. Apart from a severe drop of CEO compensation as well as S&P 500 index in 2008 due to the world-wide financial crisis, the US CEO compensation witnessed a continuous overall increase.

And as seen from Exhibit 3, CEOs received considerably more pay than the average US worker, and this gap was continually increasing after the 2008 financial crisis. By 2015, CEOs in US top 500 firms earned an average of 248 times as much as the annual average pay of the median US worker did.

To summarize, CEO compensation remains constantly high and the executive-employee pay gap is escalating accordingly.

(20)

Exhibit 2: Trends in the S&P 500 index and CEO compensation

Exhibit 3 - S&P 500 CEO-to-Median U.S. Worker Pay Ratio

Source: Equilar. Executive Compensation & Governance Outlook, 2017

2.2 National culture

2.2.1 Definition of national culture

National culture is a distinctive set of values, beliefs, customs, and assumptions generally held by people in one country. National culture defined by Hofstede is “a collective programming of the mind which distinguishes one group from another”. Bame-Aldred et al. (2013) defined national culture as a dominant culture within a country’s political boundaries. All major social systems (such as religious beliefs, education, family relationship, politics, law, and economy) are closely related to the national culture.

Morden (1999) summarized three kinds of national culture frameworks in existing literature: which are historical-social frameworks, single dimension frameworks, and multiple dimension frameworks respectively. Historical-social frameworks are originated from historical and social opinions on national culture. Single dimension frameworks use one

0 500 1000 1500 2000 2500 0 5000 10000 15000 20000 25000 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 CEO annual compensation (in thousand dollars)* S&P 500

188 178 217 225 226 246 247 248 150 170 190 210 230 250 2008 2009 2010 2011 2012 2013 2014 2015

(21)

dimension or variable to depict national culture. For example, Hall (1976) differentiated culture by categorizing it as high context or low context, in which “context” was defined in terms of how people and their society seek information and knowledge. Multiple dimensions frameworks define culture by applying several dimensions or variables. Lessem and Neubauer (1994) categorized national culture under four dimensions, which are pragmatism, idealism/wholism, rationalist, and humanist. Hampden-Turner and Trompenaars (2011) categorized culture based on the resolution of seven value dilemmas. The most famous and wide-adopted multiple dimension model, however, is the one developed by Hofstede (1980). I will illustrate this in detail as follows.

2.2.2 Hofstede's cultural dimensions theory

In the late 1970’s, the Dutch scholar, Geert Hofstede, analyzed the cultural characteristics of human populations in different countries based on the form of questionnaires, and used 4 national cultural dimensions – which were power distance, uncertainty avoidance, individualism versus collectivism, and masculine versus femininity – to describe the national cultural differences. Later, two dimensions were added into the national cultural framework, which are long versus short term orientation and indulgence versus restraint respectively. The power distance refers to the degree to which people will accept the power distribution inequity within an organization. The power distance index not only manifests itself in the subordinate's obedience to superiors, but also in social status, educational level, and other respects. The higher the power distance index, the superiors are more dictatorial to the subordinates, and the subordinates are also more willing to accept this inequality. Hofstede attributed gaps in power distance not only to differences in social norms, family and education related issues, but also to differences in regional climate environments.

Individualism versus collectivism index explores the relationship between individuals and groups in a society. Individualism emphasizes on individual efforts and personal rewards, and value personal needs, while collectivism emphasizes on group efforts and group rewards, and value collective needs.

Uncertainty avoidance refers to the level of threat that people feel when faced with uncertainties and ambiguous situations. Generally speaking, in a society with low tolerance to uncertainty, people tend to be more risk-averse, will always try to prevent threat from equivocal and ambiguous prospective situations, and form a high sense of initiative and an inner impulse to work hard. On contrast, people in a society where the score of uncertainty

(22)

avoidance is low tend to encourage more risk-taking activities and to relax attitudes towards life.

The masculinity versus femininity dimension was proposed by Hofstede according to the division of different gender roles. Men are regarded as ambitious, tough, and assertive, while women are viewed as gentle, modest, and caring about life quality. Therefore, masculinity refers to preference for success, adventure, and social status. While femininity emphasis on interpersonal relationships, quality of life, and cooperation. People in a femininity dominated society attach great importance to cooperative relations and advocate collective decision-making. Interestingly, the higher the evaluation of “masculinity” in a society is, the greater the differences between male and female values are.

The long-term versus short-term dimension refers to whether people take a long-term view of strategy implementation. The difference between a short-term oriented culture and a long-term oriented culture lies in the difference in strategic perspective: focusing on past and present results or making future long-term decisions. Furthermore, the long-term orientation emphasizes thrifty and pragmatic problem-solving; people in countries that score highly for this dimension tend to be modest and pragmatic. While the short-term orientation values tradition and social responsibilities; people in countries which are more short-term focused tend to be religious and nationalistic.

The indulgence versus restraint dimension is defined as the degree to which a person's basic needs and the desire to enjoy life can be tolerated. The higher the score on this dimension is, the more tolerant the society is to allow self-indulgence.

Hofstede's national cultural framework has become a paradigm for comparing national cultures and has been deemed to be important and useful for depicting national cultural differences (Hofstede, 2011).

2.2.3 Related Management Control research

Culture can help synergize management control system and facilitate the operation of the system under certain circumstances (Ouchi, 1979). Much empirical research has verified different aspects of relation between national culture and management control. Murphy (2003) took students from public universities in Mexico and US as research objects to compare and analyze their differences in preferences for management control, and his results showed that, compared to Mexican students, American students preferred the inter-personal assessment and individual-based compensation; in contrast, Mexican students preferred

(23)

team-based compensation over American students. Rodrigues and Kaplan (1998) argued that uncertainty avoidance measure can be applied as a predictor of the degree of an organization’s formalization and studied the direct relationship between uncertainty avoidance and organization formality, and they expected a positive correlation between uncertainty avoidance and the degree of organizational formalization. Harrison (1993) studied the relationship between accounting performance indicators and low relative work pressure as well as high job satisfaction by comparing Singaporean managers with Australian managers. As a result, he found that Singapore managers preferred this relationship because Singapore managers hold perceptions of relatively low individualism and high power distances. By applying Hofstede's five national culture dimensions, Newman and Nollen (1996) investigated the financial performance of European and Asian subsidiary units in one cross-national corporation and found that congruence between management measures and national culture will improve the financial performance in one subsidiary unit. Jansen, Merchant and Van Der Stede (2009) studied incentive compensation systems in US companies and Dutch companies and strongly suggested that the national setting did matter in explaining similarities and differences in uses of incentive compensation systems, in particularly, they found that the Dutch firms were much less tended to offer incentive compensation to managers than US firms were. Similarly, by comparing organizational control between Japanese firms and US firms, Chow et al. (1999) supported the argumentation that variances in national culture influence individuals’ preferences for and reactions to controls at the profit center level, and that both the composition and tightness of the management control system might need to be adjusted to fit the nation’s culture.

2.2.4 Related Human Resource Management research

In the context of economic globalization, much existing research have studied the cross-cultural Human Resource Management in multinational corporations (MCNs), while not so many scholars have focused on the impact of national origins of multinational corporations on the Human Resources Management (HRM) practices. Bomers and Peterson (1977) argued that the multinational corporation’s country or region origin had influence on the bargaining strategies and European based multinational corporations tended to decentralize the intimidating bargaining tactics more than American based corporations did. Upon studying the differences of the use of expatriates in US, European, and Japanese multinational corporations, Tung (1982) found that Japanese companies experienced a relative lower expatriate failure rate as Japanese companies were found to place greater important in human

(24)

resource management and use more rigorous procedures to select and train expatriates. Based on the surveys of 116 American subsidiaries and 38 European subsidiaries in UK, Young et al. (1985) found that, relative to European multinational firms, American multinational firms tended to be more centralized and standardized in terms of personnel management and were more tended to use financial targets. Bartlett and Ghoshal (1989) found that the headquarters of US multinational corporations would influence HRM practices such as remuneration system and employee welfare policies to be more formalized and centralized. Liu (2003) argued that differences in Human Resource Management between European MNCs and US MNCs come from different psychological and sociological principles. He further argued that HRM in US MNCs were rooted in psychology and its primary concern was how to motivate workers to work, and this leaded to focus on individuals, analyzing the needs of employees, reward systems, and the joy of work. In contrast, he argued that HMR in European MNCs evolved more from sociological principles, which indicated more attention to the relationship between social systems, economic and political environment, and employees’ rights were not only protected by the law but also actively protected by labor unions.

Further, other scholars have argued that multinational corporations are not only affected by the culture origin of parent country, but also affected by the culture values in the host country. Some research on cross-cultural management of multinational corporations have investigated the multinational corporations’ behaviors when faced with a balance between the home country culture and the host country culture. For example, Buller et al. (1999) put forward a decision tree model for coordinating and managing cross-cultural conflicts for multinational corporations. In this process from MCNs completely ignoring the culture of the host country (i.e., insisting on the culture of the home country) to fully accepting the culture of the host country, they carefully identified and analyzed six strategies to deal with cross-cultural conflicts, which were avoidance, compulsion, training, trade-offs, cooperation, and adaptation respectively, and they further argued each strategy can be optimal for solving cultural conflicts in certain situations. Similarly, Adler proposed three solutions to solve the cultural conflicts faced by multinational corporations: which were dominance, compromise, and synergy respectively.

2.3 Hypothesis Development

As discussed above, most of studies have investigated firm compensation practice by focusing on corporate economic factors, governance factors, and agent’s individual

(25)

characteristics. Limited research has investigated the impact of national culture on firm compensation practices.

Ngo et al. (1989) studied the impact of national origins on human resources management practices by comparing corporations operating in Hong Kong but originated from US, UK, Japan, and Hong Kong respectively, and found significant differences in their compensation practices, as well as training policies. Townsend and Scott (1990) conducted a field survey of 28 countries across five different cultural circles to study the difference of compensation design under different national cultural backgrounds, and found that national culture had the greatest impact on the changes of proportion of wages and benefits in overall compensation practice. Kim et al (1990) compared reward allocation preferences across the United States, Japan, and South Korea, and found that the groups from the United States and Japan exhibited a stronger preference for equity than the group from South Korea. This analysis leads them to imply that the tendency to use fairness standards in pay distribution is of greater concern in highly individualistic countries than in low individualistic countries. Through a comparative analysis of the United States and Japan, Gomez-Mejia and Welbourne (1991) found that national culture has a significant impact on the compensation strategy of multinational corporations, and argued that multinational corporations must take the influence of national culture into consideration when designing compensation strategies. Graham and Trevor (2000) took organizational justice theory as their theoretical basis, and proposed a basic strategy for compensation system design through studying changes in the concept of fairness under different cultures. Tosi and Greckhamer (2004) related four cultural dimensions (uncertainty avoidance, power distance, individualism, and masculinity-femininity) developed by Hofstede to several dimensions of CEO compensation, and concluded that CEO compensation is highly correlated to power distance and individualism. Based on the arguments above, I further proceed with my hypothesis that:

(26)

3 Methodology

3.1 Sample

In order to investigate whether national origin will influence firm compensation practices, I examine compensation practices of US listed firms headquartered in the US (referred to US based firms hereafter) relative to US listed firms that are headquartered in Europe (European based firms). It is basically deemed that the number of European based firms are limited and much smaller than that of US based firms, thus I decide to initially construct a list of European based firms with all available data, then match these European based firms with US based firms.

I start with collecting all US listed firms from the Compustat - Capital IQ database and subsequently extract European based firms based on the available variable named ‘ISO Country Code – Headquarters’, which contains the standard 3-letter abbreviation code2_that identifies the country where the firm is headquartered. I firstly derived a full list of 312 European based firms by applying 20 different European country codes3_{, which refer to} twenty main countries in Europe. Then I try to find all data for these European based firms. I examine two proxies for firm compensation practices, i.e., CEO pay levels and the CEO pay ratio. With regard to the CEO pay ratio, however, CEO pay ratio data is only available in the 2018 proxy statement because the disclosure requirement for the ratio of CEO pay to that of median worker just came into effective in 2017. Therefore, my focus is on the cross-sectional data in the fiscal year 20174_.

2_{The country codes are established by the International Standards Organization (ISO).}

3_{These 20 European country codes are AUT (AUSTRIA), BEL (BELGIUM), CHE (SWITZERLAND), CZE (CZECH}

REPUBLIC), DEU (GERMANY), DNK (DENMARK), ESP (SPAIN), FIN (FINLAND), FRA (FRANCE), GBR (UNITED KINGDOM), HUN (HUNGARY), IRL (IRELAND), ISL (ICELAND), ITA (ITALY), LUX (LUXEMBOURG), NLD (NETHERLANDS), NOR (NORWAY), PRT (PORTUGAL), SVK (SLOVAKIA), SWE (SWEDEN).

4_{In order to obtain more samples with regard to the CEO annual compensation, I also have tried to use the data of the fiscal}

year 2016. But I found that, for most of the 312 European based firms, data of CEO annual compensation that is missing for the fiscal year 2017 is also missing for the fiscal year 2016 from the Execucomp database. Thus, I decide to use all data of fiscal year 2017 to make the data collection process much easily as well as consistent.

(27)

As the requirement for disclosure of CEO pay ratio is quite new and not available in any public database, I manually collect this data on the 2018 proxy statements one by one5_, leaving 66 European based firms that already issued their proxy statements with the new CEO pay ratio disclosure6_{. These 66 European based firms are headquartered in Belgium,} Switzerland, Germany, Denmark, Spain, Finland, France, United Kingdom, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, and Sweden accordingly. Data for CEO annual compensation and firm economic characteristics is available for most of firms from the Execucomp database and for the rest I collected data from 2018 SEC proxy statements. I manually obtained data for CEO pay ratio and variables concerning board characteristics from 2018 SEC proxy statements.

Subsequently, I tried to match these 66 European based firms with US based firms on basis of the firm size and industry. Industry is controlled by matching the 2-digit GIC (Global Industry Classification) sector numbers. GIC industry sector category is developed by Morgan Stanley Capital International (MSCI) and Standard & Poor’s (S&P). The GICS system include 11 sectors, 24 industry groups, 68 industries, and 157 sub-industries which have covered both S&P 1500 companies and non-S&P companies. The widely-known SIC system is adopted by many academic researchers despite its major limitation that the SIC is slow to reflect emerging industries since it was developed for traditional industries prior to 1970. According to Weiner (2005)’s study of the impact of industry category system on financial research, it was concluded that there was a high correspondence between SIC and GICS. Thus, in my paper, I choose to use the GIC system to conduct comparable selection of control firms.

5_{As for the data for control variables, I firstly tried to derive data from the Compustat - Capital IQ database. Data concerning}

the firm aggregate economic characteristics, namely, firm size, firm performance firm risk, and industry sector, are all directly available from the Compustat - Capital IQ database. But the information for CEO characteristics and corporate governance characteristics are incomplete. So, in the process of collecting the CEO pay ratio data, I also verified and collected data for CEO annual compensation, CEO gender, CEO tenure, CEO duality, board size, percentage of independent directors, and percentage of CEO stock ownership.

6_{The time of collecting data started in May 2018. Therefore, the sample is not random as my sample is primarily contingent}

on the end-of-year month of the respective firm and less so on how quickly firms issue their filings. Prior research found that the choice of fiscal year ends (FYE) for firms was associated with business seasonality as revenues in firms such as the agriculture firms highly depended on the season, and a firm could choose any consistent fiscal year ends at convenience. Thus, the choice of fiscal year ends is not likely to influence my findings.

(28)

For each European based firm, I randomly select a matched control firm (US based firm) in the same industry with similar firm size7_{. For the GIC sector}_{Consumer Staples}_{, only 2} European based firms are available and best-matched firms for these European firms are also not found. As such, I omitted firms in the Consumer Staples sector.

Table 1. Sample selection

Panel A: Selection process for European based firms

Number of samples

Initial avaible European based firms 312

Less: Firms that did't disclosure 2018 SEC proxy statement 246

European firms with all avaible data 66

Less: 2 European based firms with no best-matched US based firms 2

Final sample for European based firms 64

Panel B: Industrial sector of all sample firms

GIC industrial sector 2-digit GIC

number European based firms US based firms Total

Energy 10 6 6 12 Materials 15 2 2 4 Industrials 20 14 14 28 Consumer Discretionary 25 10 10 20 Health Care 35 20 20 40 Financials 40 8 8 16 Information Technology 45 4 4 8 Total - 64 64 128

7_{I proxy the firm size by using the firm total assets. For each European based firm, I randomly select a matched US based firm}

(29)

Therefore, I finally choose these 64 European based firms and match these European based firms with 64 US based firms as my sample based on their firm size and industry. My sample includes 128 firms (64 European based firms and 64 US based firms) in seven industry sectors. All data are collected for the fiscal year 2017. My sample includes most GIC industrial sectors, including energy, materials, industrials, consumer discretionary, health care, financials, and information technology. The sample selection process for the European based firms are exhibited in the panel A of Table 1. Industrial sectors included in my sample are listed in panel B of Table 1, along with their respective 2-digit GIC numbers, and the number of US based firms and European based firms sampled in each industrial sector.

3.2 Empirical model

I test my hypothesis by means of the following model:

𝐶𝐸𝑂𝑐𝑜𝑚𝑝 = 𝛽0+ 𝛽1𝑁𝑎𝑡𝑐𝑢𝑙𝑡𝑢𝑟𝑒 + 𝛽2𝑠𝑖𝑧𝑒 + 𝛽3𝑓𝑖𝑟𝑚 𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 + 𝛽4risk + 𝛽5𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑦

+𝛽6boardind + 𝛽7duality + 𝛽8boardsize + 𝛽9ownership + 𝛽10𝑔𝑒𝑛𝑑𝑒𝑟 + 𝛽11𝑡𝑒𝑛𝑢𝑟𝑒 I proxy the firm compensation practice as CEO compensation in my paper, and I focus on two dimensions of CEO compensation, which are CEO total compensation and the executive-employee pay gap respectively.

(1) CEO total compensation: this is measured by the variable ‘Total Compensation - As Reported in SEC Filings ($)’ which is directly derived from Compustat - ExecuComp database.

(2) The executive-employee pay gap: I use the CEO pay ratio which refers to the CEO annual compensation to that of the median employee to proxy for this variable.

The independent variable is measured by the location of the firm’s headquarter. This is a dummy variable which is assigned value one when the company’s headquarter is located in Europe (non-US) and zero when company’s headquarter is located in the US. The choice for the location of the headquarter as a proxy for national culture is motivated on the basis of prior literature. By comparing management style of American international companies and that of European international companies, Shetty (1979) concluded that due to culture and tradition differences, every multinational company would adopt solutions that reflect the values of the country in which it domiciled. Therefore, it is deemed that firms whose headquarters are based in Europe (Europe-based firms) will be mainly affected by European culture, while firms whose headquarters are based in US (US-based firms) will be mainly

(30)

affected by American culture. On the other hand, multinational companies have to attract employees from local labor markets and downward deviations from local employment conditions may make it difficult for firms to attract the desired labor force. Thus, it should be noted that multinational firms may have to comply with local pay levels in order to attract the desired quality and quantity of labor.

On the basis of my hypothesis, I expect that _𝛽₁ is significantly different from zero.

3.3 Control variables

Control variables in this paper can be divided into three categories. All proxies for control variables are based the operationalization of these variables in prior research.

The first category describes basic economic characteristics of the company that, amongst others, influence corporate compensation practices (Core et al., 1999).

(1) firm size, which is operationalized as the company’s total assets in the year 2017; (2) firm performance, which is measured by ROA (return on asset);

(3) firm risk, which is measured by the standard deviation of ROA over the prior seven years; (4) industry, which is measures by measured by the standard GIC sector categories.

The second category depicts corporate governance related variables as the quality of corporate governance is argued to influence the degree in which powerful CEOs can upward influence their compensation payouts.

(1) % of independent directors, which depicts the degree to which the board is independent of executives, and this variable is measured by the percentage ratio of independent directors to the total directors who actively sit on the board;

(2) CEO duality, the dummy variable which depicts whether the CEO is also the chairman of the board in the company;

(3) board size, which is measured by the total number of directors who actively sit on the board;

(4) CEO ownership, which is measured by the average annual CEO stock ownership percentage; according to the SEC reporting convention, this ratio is calculated as the number of shared held or exercisable within sixty days of a firm’s proxy statement filing divided by the total number of common shares outstanding.

(31)

The third category includes CEO individual characteristics variables.

(1) CEO gender, this is a dummy variable which is assigned one for a male CEO and zero for a female CEO;

(2) CEO experience, which is measured by CEO tenure, more specifically, the number of years of employment served by the CEO in the company, which is believed to be a proxy for the power of CEOs.

See Table 2 for a summary description of all variables.

Table 2 - Variable definition and description

Variable name Abbreviation Description

Dependent variable

CEO annual

compensation CEOcomp

This is measured by the total compensation as reported in SEC proxy statements. The total annual compensation is generally the sum of salary, bonus, the total value of restricted stock granted, the total value of stock options, long-term incentive payouts, and all other payments.

executive-employee

pay gap paygap

This is measured by the pay ratio of CEO compensation to that of the median employee.

Independent variable

location of

headquarter Natculture

This is a dummy variable in the first step regression, which is assigned value one when the company is headquartered in Europe (Non-US) and zero when company is headquartered in US.

Control variables

Firm economic characteristics

firm size size This is measured by the company’s market value

firm performance Firm

performance

This is measured by ROA (return on asset).

Firm risk risk The relevant proxies for firm risk are standard deviation of ROA over the prior seven years.

Industry industry This is measured by the standard GIC sector categories

Corporate governance characteristics

% of independent

directors boardind

This is measured by the ratio of independent directors to the total number of directors

(32)

CEO duality duality

CEO duality is a dummy variable indicating that the firm’s CEO also serves as the chairman of the board of directors. CEO duality is assigned one if the CEO is also the chairman, and zero otherwise.

Board size boardsize Board size is the total number of directors on the board. % of CEO stock

ownership ownership

CEO percentage stock ownership is the percentage of outstanding shares owned by the CEO.

CEO charateristics

CEO gender gender This is a dummy variable that equals one if the CEO is male, while equals zero if the CEO is female

CEO experience tenure

This is depicted by the number of years of the CEO at the firm as CEO. More specifically, this is measured by the number of years of employment served by the CEO in the company.

(33)

4 Results

4.1 Descriptive statistics

Table 3 presents descriptive statistics for the main variables used in my paper. The average CEO annual compensation is $9,370,639. The mean of CEO pay ratio is 202.16, which means that the CEO averagely earns 210 times as much as the median worker earns. The sample firms have an average size of $ 16 billion, with an average 5% annual return on assets and 13% firm risk (standard deviation of ROA). The board consists of averagely 10 members where 85% are independent directors. Mean percentage of CEO stock ownership is 1% which is common in large firms. With regard to the CEO duality, it is assigned one when CEO is also the chairman of the board and zero otherwise. As it shows in Table 3, over 50% firms separate the position of Chief Executive Officer and the Chairman position. CEO gender is also a dummy variable which is assigned one when CEO is male and zero otherwise. We can see that the vast majority of CEOs are males. As for the CEO experience, the average number of years served by CEO is 6.6 years in the company.

Table 3. Descriptive statistics

MEAN SD 10% 25% 50% 75% 90%

Panel A: dependent variables

CEO annual compensation (unit: dollars)

9,370,639 6,346,224 2,432,193 4,583,270 7,991,318 13,300,000 18,800,000

CEO pay ratio 202.16 258.56 51.00 82.50 153.00 213.50 335.00

Panel B: independent variable

HQ 0.50 0.50 0.00 0.00 0.50 1.00 1.00

Panel C: control variables

Firm size (unit: million dollars) 16,516.34 30,857.01 391.73 2,162.31 6,641.53 16,674.00 32,623.00 Firm performance 0.05 0.11 -0.07 0.03 0.08 0.10 0.16 Firm risk 0.13 0.38 0.01 0.01 0.03 0.06 0.17 % of independent directors 0.85 0.08 0.75 0.83 0.88 0.90 0.92 CEO duality 0.39 0.49 0.00 0.00 0.00 1.00 1.00 Board size 9.77 2.11 7.00 8.00 10.00 11.00 12.00 % of CEO stock ownership 0.01 0.04 0.00 0.00 0.00 0.01 0.02 CEO gender 0.98 0.12 1.00 1.00 1.00 1.00 1.00 CEO tenure (in years) 6.59 5.88 1.00 2.00 5.00 10.00 16.00

(34)

T abl e 4 . C or re la ti on Ma tr ix 1 2 3 4 5 6 7 8 9 10 11 12 (1) C EO a nnu al c om pe nsa ti on (U ni t: do ll ar s) 1.00 (2) C EO p ay r at io 0.24 1.00 (3) HQ -0.07 -0.04 1.00 (4) F ir m si ze 0.39 0.06 0.02 1.00 (5) F ir m p er fo rm anc e 0.18 0.15 -0.19 0.00 1.00 (6) F ir m r is k -0.17 -0.13 0.15 -0.15 -0.70 1.00 (7) % o f i nd epe nd ent d ir ec to rs 0.16 0.07 0.02 0.18 0.09 -0.11 1.00 (8) C EO d ua li ty 0.27 -0.06 -0.06 0.22 0.09 -0.15 0.16 1.00 (9) B oa rd si ze 0.26 0.13 -0.04 0.56 0.08 -0.22 0.19 0.19 1.00 (10) % o f C EO st oc k ow ne rs hi p 0.00 0.00 -0.14 -0.09 -0.19 0.08 -0.36 0.14 -0.12 1.00 (11) C EO g end er 0.04 -0.24 -0.13 0.04 -0.04 0.04 0.11 0.10 0.05 0.02 1.00 (12) C EO te nu re 0.14 -0.04 -0.20 0.07 0.06 -0.11 -0.09 0.33 0.06 0.33 0.03 1.00 C or re la ti ons e xp re ss ed in b ol d a re s ign if ic an t a t a 5% o r l ow er si gni fi ca nc e l eve l

The impact of national culture on firm compensation practices : does national origin affect compensation practices differences between European firms that cross-list in US and US listed firms?