Master Thesis The effect of culture on the relationship between CSR orientation and executive compensation packages: a comparative European study

(1)

Master Thesis

The effect of culture on the relationship between CSR orientation

and executive compensation packages: a comparative European study

Thesis MSc Accountancy & Controlling, track Accountancy Academic year 2016-2017, Rijksuniversiteit Groningen

June 20th 2017

Abstract

Financial performance criteria have been dominant in executive compensation packages, and little focus has been placed on the role of non-financial performance criteria in these packages. Using a sample of 143 executives from 8 countries across Europe, this paper cross-sectionally studies the effect of the degree of CSR orientation on executive compensation packages. In particular, this paper tries to define the moderating effect of cultural differences between countries on the relationship between CSR and executive compensation. The results provide no evidence of a moderating effect of cultural differences on the previously mentioned relationship. Some possible causes are identified and discussed.

Student T. Kesler s2322129 Supervisor R.C. Trapp

Co-assessor C.A. Huijgen

(2)

1

Introduction

1.1 Background

There is evidence that financial performance criteria are dominant in executive compensatio n packages (Murphy, 2000). However, there has been little focus in the scientific literature on how non-financial performance indicators, such as environmental, social and corporate governance variables, can affect executive compensation packages (McGuire, 2003; Hong et al., 2015) and to what extent ‘triple bottom line’ criteria are reflected in these compensatio n packages (McGuire, 2003). These criteria specify that, while firms should engage in environmentally and socially responsible behaviour on moral grounds, it is also possible to simultaneously benefit financially from these activities (Gimenez et al., 2012). If these financial benefits are the result of managerial performance, they should be reflected in an executive’s compensation package and thus alter its composition (McGuire, 2003).

This paper aims to link corporate social responsibility (CSR) to the degree of long- term compensation in an executive’s compensation package. CSR has been defined by Carroll (1979) as the firm’s consideration for investors, society and the government and its responsibilities towards these stakeholders. As will be discussed in more detail below, CSR goals are usually term and thus the executive compensation package should reflect a long-term focus by providing the appropriate incentives. However, there is often a mismatc h between which goals a firm intends to incentivise and which goals it actually incentivises. For instance, Shan & Walter (2016) discuss the outrage over excessive executive compensatio n and argue that most of the outrage is actually due to design weaknesses in the contracts that exist between executives and their firms. These weaknesses cause a mismatch between what behaviour the firm is aiming to incentivise and what kind of behaviour it actually incentivises. In summary, the mismatch between promoting CSR commitments to external stakeholders and genuinely compensating and incentivising executives to make these commitments is the focal point of this paper. By attempting to research whether a more socially responsible firm compensates its executives more for aligning their actions with long-term goals and objectives, a step can be made towards further understanding the mismatch between compensation and firm goals.

Cai et al. (2011) state that, while there is abundant literature on both CSR and executive compensation, there is little literature describing the link between these two. In their paper, they find that CEOs in more socially responsible firms are paid significantly less than CEOs in more socially irresponsible firms.

(4)

3 They argue that this is due to conflict-resolution based on stakeholder theory. They explain that CSR can be used to solve conflicts between shareholders and other stakeholders. According to Cai et al. (2011) there are two ways in which CEO compensation can help resolve conflicts, namely one, that a lower level of compensation avoids conflicts with certain stakeholders and two, a CEO with higher ethical and social standards may find a lower level of compensation desirable based on virtue ethics.

This would suggest that there is a relationship between the degree to which CSR is important within a firm (CSR orientation) and the executive compensation. However, prior literature acknowledges that further research is necessary. Mahoney & Thorne (2005) also call for further research into this theme because their findings are contradictory to McGuire (2003). Mahoney & Thorne (2005) find that firms compensating their executives based on more long-term compensation show less CSR weaknesses. McGuire (2003), however, finds a significa nt positive relationship between long-term compensation and CSR weaknesses, meaning that the more long-term the compensation, the more CSR weaknesses are likely to be present in a firm. This shows that the evidence on the link between CSR and executive compensation is inconclusive at best and there is a need for additional research in the area that links CSR orientation to executive compensation. Furthermore, there is specific need for more research into the relationship between long-term compensation systems and CSR orientation as shown by McGuire (2003) and Mahoney & Thorne (2005). The financial crisis has shown that compensating executives too much for short-term performance can cause a mismatch between compensation and executive performance, and therefore firm performance. This mismatc h leads to firms not incentivising the actions which are necessary for the achievement of long-term goals, but rather the actions that maximize executive compensation in the present because executives’ present compensation is not dependent on future results (Jensen & Murphy, 1990).

1.2 Research Questions

It is important to understand the relationship between CSR orientation and compensatio n because CSR has mostly long-term goals and the board of a firm will need to know in what way to compensate executives so their actions are aligned with these long-term goals (Amihud & Lev, 1981; Holmstrom, 1979; Jensen & Meckling, 1976).

The link between CSR and executive compensation has been researched mostly within a single country or in the United States, but in different countries there may be varying factors that influence this relationship. One of these is culture.

(5)

4 Zilber (2012) states that through applying institutional theory to firm settings, the transformation of firm cultures can be better understood. In addition, he suggests that through institutional theory, the effect of outside cultures on firm cultures can be researched more properly. The differences in national cultures between countries are therefore expected to influence business cultures in these countries. Therefore, this paper aims to shed light on the influence of differences between countries regarding cultures and attitudes towards doing business through comparing their national cultures. The main dimensions that vary across cultures are individualism, masculinity, power distance and uncertainty avoidance. Their individual impact on national culture will be explained in the next section.

According to Murillo & Lozano (2006), CSR is identified using organisational culture. In their survey they found that, when they used the term ‘social practice’, organisational actors felt the need to explain this term. This explanation was based on the culture in their firm. For this reason, cultural differences between European countries, as identified according to the taxonomy by Hofstede (1983), are linked to the relationship between CSR orientation and, by extension, executive compensation packages in these countries.

As mentioned earlier, this paper aims to link CSR orientation to the degree of long- term compensation in an executive’s compensation package in order to determine whether executives are compensated for investing in long-term CSR goals. This leads to the follow ing research questions: First, what is the influence of CSR orientation on long-term executive

compensation? Second, what is the influence of country-specific culture differences on this relationship?

The research question will be empirically tested through OLS regression analysis based on cross-sectional data of 143 executives from eight European countries. The results show that the degree of CSR orientation is not significantly related to the degree of long-term compensatio n, and that the moderating effect of country-specific culture differences on this relationship is also not significant. These findings are not consistent with the expectations based on prior literat ure and institutional theory.

1.3 Relevance

This paper adds to the existing literature on the subject of CSR and executive compensation in two major ways. First, it focuses on European countries and examines the influence of cultura l differences between these countries on the relationship between long-term executive compensation and CSR orientation. The paper attempts to explain these cultural effects from an institutional theory perspective.

(6)

5 According to institutional theory, the effects of underlying cultural differences ma y influence the business culture and attitudes in a specific country (Zilber, 2012). DiMaggio & Powell (1983) described three ways in which firms experience pressure from the environme nt, namely coercive, mimetic and normative. The latter is of special interest in this paper’s setting since it describes the mechanism by which firms experience pressure from external norms and values. These external norms and values are, for an important part, shaped by national culture. This way, national culture can influence the business attitudes of firms.

Second, it uses an index to measure the degree of CSR orientation in order to add valid it y to that concept as an improved measure of CSR orientation. This index will consist of four areas, namely economic, environmental, social and corporate governance performance. Under the traditional CSR definition of Carroll (1979), CSR includes consideration for investors, society and the government and a firm’s responsibilities towards these stakeholders. The four areas adequately cover the main stakeholders from Carroll’s definition and therefore seem sufficient to act as proxies for the degree of CSR orientation of firms. Mahoney & Thorne (2005), among others (Gjolberg, 2009; Skouloudis et al. 2016), used an index of variables to measure the degree of CSR orientation but call for more research using these types of indices to enhance their validity. They also call for further research using different national contexts and new variables because, as explained in the previous section, some of their findings are contradictory (McGuire, 2003).

The remainder of this paper is organised as follows. Section II discusses the theoretical framework and elaborates on the definition of CSR. Section III describes the methodology used to measure the variables and to analyse the data. Section IV presents the results, and fina lly Section V discusses the results and their implications and limitations.

(7)

6

Theory

This section discusses the theoretical framework and definitions of the main concepts. CSR and compensation are first discussed, followed by culture. Last, hypotheses are developed.

2.1 Corporate Social Responsibility and Executive Compensation

Carroll (1979) describes CSR as the firm’s consideration for investors, society and the government and its responsibilities towards these stakeholders. Although many differe nt definitions of CSR have appeared over the years, Carroll’s definition was one of the first and also one of the most accepted and most used definitions (Spence, 2016; Gomez-Carrasco et al., 2016; Ramasamy & Yeung, 2009) which is why it will be used in this paper. It is also quite a broad definition, meaning that it includes a large number of stakeholders which is typical for CSR. There are also narrow definitions of CSR that, for instance, strictly adhere to the Environmental, Social and Governance aspects of CSR (Del Bosco & Misani, 2016; Husted & de Sousa-Filho, 2017)

Historically, the subject of executive compensation has been approached mostly from an agency perspective (Elston & Goldberg, 2003; Gu & Kim, 2009; Matsumura & Shin, 2005). However, this research has sometimes been contradictory. This may indicate the need to approach this issue from a different perspective. Therefore, this paper will address the research questions from an institutional perspective. Institutional theory, as defined by Scott (2004) seeks to explain the deeper and more resilient aspects of social structure. It mainly attends to how norms and routines are established, adopted and adapted, and eventually abandoned. The theory was developed in the field of social sciences and has two major branches, old and new institutionalism. New institutionalism is explained and defined by DiMaggio & Powell (1991). Their criticism on old institutionalism is that it does not take social context and the durabilit y of social institutions into account. This, according to their paper, is a major flaw since the world is becoming increasingly socially aware and institutions have become more influent ia l. Therefore, new institutionalism will be used to explain the effect of culture in this paper.

2.2 The Effect of Culture

According to institutional theory, rules that are codified in laws can influence the behaviour and perception of people. The law will be established, and then adopted and adapted by individuals, who will in turn take their newly adopted norms to their firm. In this way, the law influences national and firm cultures (DiMaggio & Powell, 1991).

(8)

7 Culture, however, will obviously be shaped by more than just a country’s law system because culture goes beyond regulation, just as CSR according to Sheehy (2015) goes beyond regulation. Because culture goes beyond regulation, it is logical to argue that, due to differe nt events and influential people, the cultures of countries have developed significant differe nces over time. This is line with what Meyer (1970) suggested, namely that social norms and rules together form most of the social order. Firms are recognised in institutional theory to be ‘rationalised systems’ that perform certain activities with a specific goal in mind (Scott, 2004).

However, an even more noticeable insight that Meyer and Rowan (1977) provide is that these rationalised systems are cultural systems, which were set up to guide firm actors in choosing the appropriate methods for achieving goals. This would imply that national culture influences the culture of firms through the actors within firms, and that these cultural systems guide actors in making the ‘right’ choices.

It is argued that, based on institutional theory as explained above, the differences in national cultures will have an impact on the way people think and act in business, and that therefore the relationship between CSR orientation and the degree of long-term compensatio n is affected by differences in national cultures.

Wan & Chiu (2009) describe culture as consisting of symbolic elements that are believed to be important by the members of that culture. This view of culture coincides with Hofstede’s (1983) dimensions (the ‘elements’) of national culture that describe the views that people have about what is important in their lives. To identify differences between cultures, measures must be used that can be applied to many different countries. Hofstede (1983) identified four cultura l dimensions on which national cultures can be rated. These dimensions have been used extensively in prior research (Hooghiemstra et al., 2015; Newman & Nollen, 1996; Soares et al., 2017), but have also been criticised. A common criticism is that just four (and in this paper three) dimensions cannot capture an entire culture accurately (Schwartz, 1994; Samiee & Jeong, 1994; Leung, 1989). This, however, is a criticism true of all measurements of culture. Hofstede’s dimensions are regarded as one of the most robust measures of culture (Smith et al., 1996), and is also very useful for cross-cultural examinations. Therefore, three of Hofstede’s dimensions will be used in this paper, namely uncertainty avoidance, individualism and masculinity. The rationale behind choosing these particular dimensions is discussed below.

(9)

8 The first dimension, uncertainty avoidance, indicates to what degree members of a society (in this paper, a country) are uncomfortable with uncertainty about the future. A country that scores higher on this dimension tends to have firm codes of behaviour and belief and is not open to unorthodox behaviour, whereas a lower score means that people are more relaxed in their attitude towards future uncertainty. Countries, and thus people, with more rigid codes of behaviour are expected to be more likely to set executive compensation in line with long- term goals since they dislike unpredictability about the achievement of these goals.

The second dimension is individualism versus collectivism, which indicates the importance that people attach to having a close-knit community around them. A higher score means society is more individualistic, while a lower score means that collectivism is valued. In a more individualistic country, people tend to be more inclined towards compensatio n maximisation. Therefore, it is expected that a higher score on this dimension decreases the degree to which CSR orientation is in line with the compensation package.

Third, the masculinity- femininity dimension measures whether a country emphasises values such as aggressiveness and dominance (high score on this dimension), or compassion and empathy (low score on this dimension). Highly feminine countries are characterised by a preference for a good quality of life, through cooperation and modesty. Because feminine countries care more about the quality of life than more masculine countries, it is expected that in these countries the compensation package is more in line with the degree of CSR orientatio n. The power distance dimension is excluded from this analysis. This score on this dimensio n represents the degree to which power is distributed unequally across a society. It also indicates how accepting the less powerful members of society are of the power of the more powerful members. This dimension has been excluded because no prior literature was found that was able to indicate a possible moderating effect of this dimension on the relationship between CSR orientation and compensation.

In summary, it is expected that in countries with higher uncertainty avoidance, and lower individualism and masculinity scores, the focus of firms tends to be more on creating long-term shareholder value, whereas in countries with low, and respectively high, scores on these cultural dimensions, firms are expected to generally be more concerned with short-term results. The scores on the dimensions are reflective of specific cultural traits that are expected to influence the way people think and act in business, and influence their focus on either long-term or short-long-term goals.

(10)

9 Furthermore, because the people that determine the executive compensation package (Board of Directors) are also influenced by national culture, it could be argued that they will be more likely to set compensation in accordance with their focus on either long-term or short-term goals.

2.3 Hypothesis development

The focus on long-term goals is one of the variables that has traditionally been used to measure the degree of CSR orientation in a firm (Cai et al., 2011; Callan et al., 2011; Mahoney & Thorne, 2005; McGuire, 2003). Because the focus on long-term goals is very important for CSR within a firm, executives should be compensated in line with these goals. Thus, it could be argued that the degree of CSR orientation in a firm has an impact on the way the executives are compensated. Executive compensation structure is mainly divided into short- and long- term compensation. Short-term compensation includes cash salaries and bonuses, while long- term compensation primarily includes stocks and stock options (Gopalan et al., 2014; Livne et al., 2013).

The closer the reward is to the moment when the results of a project are observable, the fairer the executive can be compensated for the actual performance of the project. This means that, for long-term goals, compensation with a long(er) vesting period should be used. The vesting period is the time period from when the compensation is awarded until the moment when it is actually paid to the executive. The percentage of long-term compensation as compared to the total compensation is calculated and used as the dependent variable. This measure is quite common in the literature in the field (see for instance Mahoney & Thorne, 2005; McGuire, 2003; Becher et al., 2005; Brewer et al., 2004).

An article by McGuire (2003) shows that when CEO compensation is higher, the degree of CSR weaknesses also becomes higher. This points to a mismatch between compensation and CSR goals. However, Mahoney & Thorne (2005) find that when the degree of CSR increases, the level of long-term compensation also increases. This indicates a positive relationship. If a firm is more focused on CSR, it could be expected to also be more aware of the way in which it is compensating its CEO. Furthermore, McGuire (2003) argues that the CEO of a more CSR oriented firm should want to be paid less from a virtue ethics and conflict resolutio n (stakeholder theory) perspective. In addition, it is argued by Shan & Walter (2016) that compensation being exactly in line with long-term goals is rarely the case.

(11)

10 All in all, it is expected that the relationship between CSR orientation and the degree of long-term compensation is positive. The following hypothesis is presented:

Hypothesis 1: A higher degree of CSR orientation is positively related to the degree of long-term compensation.

As mentioned above, people living in a country that scores high on the uncertaint y avoidance dimension and low on the individua lism and masculinity dimensions are expected to emphasise long-term value creation due to their aversion towards uncertainty about the achievement of future goals, their sense of community responsibility and their dedication towards creating a good quality of life for everyone (Hofstede, 1983). These are all factors associated with CSR (Carroll, 1979). Institutional theory states that this focus will be adopted by individuals and firms which will cause firms to be more focused on achieving long- term goals, resulting in also rewarding executives for achieving these long-term goals.

In countries which score low on uncertainty avoidance, and high on individualism and masculinity a focus on short-term goals is expected due to people generally being more selfis h and valuing short-term financial gains above long-term gains. This leads to the second and third hypotheses:

Hypothesis 2: A short-term value creation focus will mitigate the positive relationship between the degree of CSR orientation and the degree of long-term compensation.

Hypothesis 3: A long-term value creation focus will enhance the positive relationship between the degree of CSR orientation and the degree of long-term compensation.

(12)

11

Methodology

This section discusses the research methodology. The dependent and independent variables are first introduced, followed by the control variables and the moderating variables. Next the data and analysis are discussed.

3.1 CSR and Compensation

For the independent variable, the degree of CSR orientation, an index will be used that measures performance in four major areas of CSR. These areas are economic, environmenta l, social and corporate governance performance and will be explained in more detail below. The index is the Equal Weighted Rating available in the Asset4 ESG (Economic, Social and Governmental) database (item A4IR). This database is part of Thomson Reuters’ Datastream database. As the name of the index suggests, it weighs individual firm scores on four areas equally to create a combined score per firm. In total, the combined score on these four areas (EQRATE) represents firm scores on 1.291 variables relating to CSR. This combined score can range from 0 to 100, with 100 being the highest attainable score on all areas. The Asset4 database has been used in some recent research indicating its appropriateness for use in scientific research (Daszynska-Zygadlo et al., 2016; Jitmaneeroj, 2016; Chopra & Wu, 2016). Furthermore, due to the enormous amount of time it would take to gather just a few dozen variables by hand, and because the inclusion of more variables can provide a more accurate basis for statistical analysis, this index is considered very appropriate for this paper.

The score on the area Economic consists of the average of 156 variables relating to, among others, implementation and improvement of employee satisfaction policies, whether management has publicly committed to avoid misuse of information, and whether a policy to improve financial transparency is in place. The score on the Environmental area is based on the average of 332 variables which mainly relate to whether the firm reports on its environmental impact and has policies in place to reduce its CO2 and other emissions. In the

Social area, a firm is scored on 521 variables that relate to the firm’s policies and actions concerning product responsibility, community responsibility and corruption and bribery charges. Finally, the Corporate Governance area scores companies on, for instance, the implementation and independence of an audit committee, and whether the CEO is a member of the supervisory board. The score for this area is based on 282 variables.

(13)

12 To measure the dependent variable, executive compensation and, more specifically, the long-term component of executive compensation, the amount of variable long- term compensation in comparison to the total compensation package is calculated (LCOMP). Long-term compensation is defined as all equity-based compensation (Brockman et al., 2016). This includes, among others, stock and stock option compensation. Tosi & Greckhamer (2004) use a similar definition, as does Sapp (2008). Total compensation will be measured as the combination of cash and equity compensation that the executive has received. The executive compensation data were obtained from the BoardEx database.

3.2 Control Variables

The control variables that will be included can be divided into two categories, namely financ ia l performance variables and structural variables. In the financial performance category are Return on Assets (ROA) (Wu & Shen, 2013) and Net Income growth (NI) (Shen & Chang, 2009; Wu & Shen, 2013). Controlling for structural differences between companies are: firm size measured as the logarithmic derivative of total sales (REVLOG) and market capitalisa t io n (MCAPLOG) (McWilliams & Siegel, 2001; Cavaco & Crifo, 2014), board independence (INDEP) as the percentage of non-executive directors on the total board and Board Size (Ghosh & Sirmans, 2005), age of the executives (AGE) in years (Ryan & Wiggins, 2001), gender of the executive (GENDER) in male or female for which a dummy variable is computed (0 = Male ; 1 = Female) (Dreher et al., 2011), and firm type (INDUST) measured through a dummy variable and sorted in accordance with the Asset4 Industry Classification (Jaskow et al., 1996) are included in the analysis. The model that will analyse the influence of only the control variables on the dependent variable is:

1. 𝐿𝐶𝑂𝑀𝑃 = 𝑎₀+ 𝑎₁𝑅𝑂𝐴 + 𝑎₂𝑁𝐼 + 𝑎₃𝑅𝐸𝑉𝐿𝑂𝐺 + 𝑎₄𝑀𝐶𝐴𝑃𝐿𝑂𝐺 + 𝑎₅𝐼𝑁𝐷𝐸𝑃 + 𝑎6𝐴𝐺𝐸 + 𝑎7𝐺𝐸𝑁𝐷𝐸𝑅 + 𝑎8𝐼𝑁𝐷𝑈𝑆𝑇 + 𝑒

To assess the relationship between the degree of CSR orientation and the degree of long-term compensation, the following model is used:

2. 𝐿𝐶𝑂𝑀𝑃 = 𝑎₀+ 𝑎₁𝐸𝑄𝑅𝐴𝑇𝐸 + 𝑎₂𝑅𝑂𝐴 + 𝑎₃𝑁𝐼 + 𝑎₄𝑅𝐸𝑉𝐿𝑂𝐺 + 𝑎₅𝑀𝐶𝐴𝑃𝐿𝑂𝐺 + 𝑎6𝐼𝑁𝐷𝐸𝑃 + 𝑎7𝐴𝐺𝐸 + 𝑎8𝐺𝐸𝑁𝐷𝐸𝑅 + 𝑎9𝐼𝑁𝐷𝑈𝑆𝑇 + 𝑒

(14)

13

3.3 Culture

The Hofstede dimensions have scores ranging from 0 to 100. On the uncertainty avoidance dimension (UAI), a higher score indicates a more long-term focus. On the individualism (IDV) and masculinity (MAS) dimension, a lower score indicates a more long-term focus. OLS regression is also used to examine the interactio n effect of UAI, IDV and MAS, resulting in the following models:

3. 𝐿𝐶𝑂𝑀𝑃 = 𝑎₀+ 𝑎₁𝑈𝐴𝐼 + 𝑎₂𝑅𝑂𝐴 + 𝑎₃𝑁𝐼 + 𝑎₄𝑅𝐸𝑉𝐿𝑂𝐺 + 𝑎₅𝑀𝐶𝐴𝑃𝐿𝑂𝐺 + 𝑎₆𝐼𝑁𝐷𝐸𝑃 + 𝑎₇𝐴𝐺𝐸 + 𝑎₈𝐺𝐸𝑁𝐷𝐸𝑅 + 𝑎₉𝐼𝑁𝐷𝑈𝑆𝑇 + 𝑒 4. 𝐿𝐶𝑂𝑀𝑃 = 𝑎₀+ 𝑎₁1𝑈𝐴𝐼 + 𝑎₂𝐸𝑄𝑅𝐴𝑇𝐸 + 𝑎₃𝑈𝐴𝐼 𝑥 𝐸𝑄𝑅𝐴𝑇𝐸 + 𝑎₄𝑅𝑂𝐴 + 𝑎₅𝑁𝐼 + 𝑎6𝑅𝐸𝑉𝐿𝑂𝐺 + 𝑎7𝑀𝐶𝐴𝑃𝐿𝑂𝐺 + 𝑎8𝐼𝑁𝐷𝐸𝑃 + 𝑎9𝐴𝐺𝐸 + 𝑎10𝐺𝐸𝑁𝐷𝐸𝑅 + 𝑎₁₁𝐼𝑁𝐷𝑈𝑆𝑇 + 𝑒 5. 𝐿𝐶𝑂𝑀𝑃 = 𝑎₀+ 𝑎₁𝐼𝐷𝑉 + 𝑎₂𝑅𝑂𝐴 + 𝑎₃𝑁𝐼 + 𝑎₄𝑅𝐸𝑉𝐿𝑂𝐺 + 𝑎₅𝑀𝐶𝐴𝑃𝐿𝑂𝐺 + 𝑎₆𝐼𝑁𝐷𝐸𝑃 + 𝑎₇𝐴𝐺𝐸 + 𝑎₈𝐺𝐸𝑁𝐷𝐸𝑅 + 𝑎₉𝐼𝑁𝐷𝑈𝑆𝑇 + 𝑒 6. 𝐿𝐶𝑂𝑀𝑃 = 𝑎0+ 𝑎1𝐼𝐷𝑉 + 𝑎2𝐸𝑄𝑅𝐴𝑇𝐸 + 𝑎3𝐼𝐷𝑉 𝑥 𝐸𝑄𝑅𝐴𝑇𝐸 + 𝑎4𝑅𝑂𝐴 + 𝑎5𝑁𝐼 + 𝑎₆𝑅𝐸𝑉𝐿𝑂𝐺 + 𝑎₇𝑀𝐶𝐴𝑃𝐿𝑂𝐺 + 𝑎₈𝐼𝑁𝐷𝐸𝑃 + 𝑎₉𝐴𝐺𝐸 + 𝑎₁₀𝐺𝐸𝑁𝐷𝐸𝑅 + 𝑎₁₁𝐼𝑁𝐷𝑈𝑆𝑇 + 𝑒 7. 𝐿𝐶𝑂𝑀𝑃 = 𝑎₀+ 𝑎₁𝑀𝐴𝑆 + 𝑎₂𝑅𝑂𝐴 + 𝑎₃𝑁𝐼 + 𝑎₄𝑅𝐸𝑉𝐿𝑂𝐺 + 𝑎₅𝑀𝐶𝐴𝑃𝐿𝑂𝐺 + 𝑎₆𝐼𝑁𝐷𝐸𝑃 + 𝑎₇𝐴𝐺𝐸 + 𝑎₈𝐺𝐸𝑁𝐷𝐸𝑅 + 𝑎₉𝐼𝑁𝐷𝑈𝑆𝑇 + 𝑒 8. 𝐿𝐶𝑂𝑀𝑃 = 𝑎₀+ 𝑎₁𝑀𝐴𝑆 + 𝑎₂𝐸𝑄𝑅𝐴𝑇𝐸 + 𝑎₃𝑀𝐴𝑆 𝑥 𝐸𝑄𝑅𝐴𝑇𝐸 + 𝑎₄𝑅𝑂𝐴 + 𝑎₅𝑁𝐼 + 𝑎₆𝑅𝐸𝑉𝐿𝑂𝐺 + 𝑎₇𝑀𝐶𝐴𝑃𝐿𝑂𝐺 + 𝑎₈𝐼𝑁𝐷𝐸𝑃 + 𝑎₉𝐴𝐺𝐸 + 𝑎₁₀𝐺𝐸𝑁𝐷𝐸𝑅 + 𝑎₁₁𝐼𝑁𝐷𝑈𝑆𝑇 + 𝑒

(15)

14

3.4 Data

Data were initially obtained on a total of 650 executives in 131_{countries, out of which 327}

datasets of individual executives (for most firms, more than one executive dataset was complete), spread over 10 countries, were complete.

These executives include not only CEOs but also COOs, CFOs and other C-level executives. The data used concerns the most recent complete set which covers the financ ia l year 2015. The CSR data were available for only 143 of these executives, spread over 49 unique firms and 8 countries2_{. It should be noted that the 143 observations were derived from an}

original sample of 650 observations, meaning that around 78% of the original sample was discarded. This is due to incomplete datasets caused by the very limited availability of the CSR data. The consequences of this bias are discussed in more detail in the Discussion section. Due to the already small amount of observations, the data were not winsorized. While this would have increased the degree to which the data approximate a normal distribution, this would also have resulted in the removal of even more observations.

The Hofstede scores on all three dimensions in all eight countries are available. The data on return on assets and net income growth, as well as firm size, board independence, executive age, executive gender and firm type have also been obtained for all companies and/or executives.

3.5 Analysis

First, a correlation matrix is comprised in order to assess possible multicollinearity issues in the model. This matrix is backed up with an examination of VIF scores, which provide another layer of depth in the multicollinearity analysis. The actual analysis of the sample will start with a T-Test in order to examine whether there are significant differences in mean score or variance between countries. This will be done by comparing each individual country to the mean score and variance of the complete sample. Then the regression analysis will take place follow ing the model equations that have been introduced earlier.

1_{The original thirteen countries are: Belgium Denmark, Finland, France, Germany, Ireland, Italy, Luxembourg ,}

The Netherlands, Spain, Sweden, Switzerland and The UK.

2_{The final eight countries are: Belgium, France, Germany, Ireland, The Netherlands, Spain, Switzerland and The}

(16)

15 As a robustness check, another regression analysis is performed in which the CSR score is calculated by taking the average of the Environmental, Social and Governance pillar scores (CSRESG), effectively extracting the effect of the economic performance score from the total CSR score. Since the literature is not consistent on whether a broad or narrow definition of CSR should be used, using both a broad and narrow definition in the analysis may provide additional insights. By excluding the economic pillar from the CSR score (EQRATE), only the literal ESG (Environmental, Social and Governance) variables are considered in the analysis (CSRESG variable).

(17)

16

Results

In this section the results of the empirical analysis are shown. The descriptive statistics of the CSR scores are presented first, followed by the statistics on compensation and the Hofstede dimensions. After that, the multicollinearity analysis and the regression models are presented.

4.1 Descriptive statistics

This paper uses cross-sectional data, the descriptive statistics of which are presented in table 1 through 3. Table 1 shows the descriptive statistics of the ESG scores. The mean score on the Equal Weighted Rating is approximately 85, and the most notable mean is that of the Corporate Governance Score which is the only mean that is not in the eighties. The standard deviation of the CG Score is also the highest of the four areas. The lower mean and higher standard deviation indicate that firms score very diverse on the CG area, and thus have very diverse levels of corporate governance.

Table 2 shows that the maximum amount of Equity Linked Compensation is much higher (almost three times) than the maximum amount of Direct Compensation, showing that the average executive receives more long-term than short-term compensation. Also notable are the standard deviations of the equity linked and direct compensation. The standard deviation of the equity linked compensation is approximately twice as large as that of the direct compensatio n. This indicates that the amount of equity linked compensation fluctuates quite a lot more, while the amount of direct compensation is relatively stable. The average total compensation per executive is just over €4 million, while the top earning executive receives annual compensatio n of almost €34,5 million.

Descriptives Economic Score Environmental Score Social Score Corporate Governance Score Equal Weighted Rating

N 143 143 143 143 143

Mean 81,48 83,35 83,30 67,09 85,18

Std. Dev 15,95 17,13 15,32 20,88 11,33

Min 7,07 20,88 28,30 26,17 25,48

Max 96,53 95,14 96,28 97,70 95,41

(18)

17 The age of the executives varies between 39 and 67 years, with 55 being the most common age. In the current sample, males represent 82,5% of the executives whilst only 17,5% are female.

The final descriptive statistics concern the country scores on the three Hofstede dimensions, which are presented in Table 3. As a higher UAI score, and lower IDV and MAS scores indicate a more long-term oriented country, Table 3 shows that Spain, Belgium and France are assumed to be more long-term oriented. In contrast, the UK and Ireland are clearly more short-term oriented countries. Switzerland and Germany are quite neutral countries. The Netherlands are a mixed country, with a high score on IDV but a low score on MAS and an average score on UAI.

4.2 Multicollinearity

Table 4 shows the correlation matrix for all main variables and control variables. From the matrix, it becomes clear that REVLOG and MCAPLOG are quite strongly correlated (β =

0,697 ; p < 0,01). This is also expected as these variables both measure firm size, albeit in a

slightly different way.

Descriptives Direct Compensation Equity Linked Compensation Total Compensation

N 143 143 143

Mean 2,006 2,010 4,016

Std. Dev 1,795 3,516 3,999

Min 0 45 335

Max 13,140 33,500 34,414

Table 2 - Descriptive Statistics Compensation (x€1000)

Descriptives UAI IDV MAS

Belgium 94 75 54 France 86 71 43 Germany 65 67 66 Ireland 35 70 68 The Netherlands 53 80 14 Spain 86 51 42 Switzerland 58 68 70 The UK 35 89 66

(19)

18 IDV also seems to be correlated with both REVLOG (β = -0,538 ; p < 0,01) and MCAPLOG (β = 0,481 ; p < 0,01). UAI and IDV are also quite strongly correlated (β =

-0,579 ; p < 0,01), indicating that high UAI scores may lead to low IDV scores. In individualis t ic

societies, people are more focused on achievement and success (Hooghiemstra et al., 2015). In less uncertainty avoidant societies people are more ready to take risks (Bryan et al., 2012), which may be necessary for individualistic persons to succeed in their society. Therefore, these variables could be negatively related. To further assess possible multicollinearity issues, VIF scores were computed for the complete model. A VIF score higher than 10 would mean that adaptation of the model is necessary, however all VIF scores in the complete model are below 4 which is well within the margin. This indicates that multicollinearity is not an issue in the model and that the regression analysis can be executed with all variables that are included below.

4.3 Country comparison

Table 5 shows the country comparison, based on a mean comparison test. As is immedia te ly noticeable, the amount of observations in some countries is extremely low (for instance, Switzerland and Belgium). The average amount of base Direct Compensation also varies significantly, from €75.000 in Belgium to more than €2.8 million in Germany. The same can be said for the amount of Equity Linked Compensation and the percentage of Equity Linked Compensation compared to Total Compensation.

LCOMP UAI IDV MAS EQRATE ROA NI REVLOG MCAPLOG INDEP AGE GENDER

LCOMP 1 UAI 0,017 1 IDV 0,193** -0,579*** 1 MAS -0,326*** -0,348*** -0,104 1 EQRATE -0,029 0,301*** -0,117 -0,107 1 ROA 0,164* -0,095 -0,035 -0,333*** -0,139* 1 NI -0,171** 0,067 -0,299*** 0,103 0,064 0,377*** 1 REVLOG -0,316*** 0,394*** -0,538*** 0,030 0,304*** -0,304*** -0,195** 1 MCAPLOG -0,079 0,445*** -0,481*** -0,231*** 0,409*** 0,156** 0,041 0,697*** 1 INDEP 0,234*** 0,132 0,030 -0,143* 0,037 0,018 -0,099 0,211** 0,297*** 1 AGE -0,004 0,256*** -0,146* 0,146* 0,108 -0,238*** -0,031 0,388*** 0,144* 0,169** 1 GENDER -0,144* 0,290*** -0,253*** -0,126 0,148* -0,082 0,136 0,114 0,164** -0,119 -0,204** 1

*Significance at the 10% level **Significance at the 5% level ***Significance at the 1% level

(20)

19 The Equal Weighted Rating, on the contrary, is quite stable between 80 and 90 out of 100 across countries. These findings indicate that, although the level of compensation varies greatly between countries, this does not appear to be a consequence of strong variation in the degree of CSR alone.

Country Variable Name Mean Std. Deviation N Belgium Direct Compensation 75 0,00 1

Equity Linked Compensation 761 0,00 1

% Long Term Compensation 91,03% 0,00 1

Equal Weighted Rating 90,52 0,00 1

France Direct Compensation 1306,00 1068,38 21

Equity Linked Compensation 2910,29 2134,43 21

% Long Term Compensation 70,41% 16,61% 21

Germany Direct Compensation 2803,87 2147,64 63

Ireland Direct Compensation 1503,81 1125,87 21

Netherlands Direct Compensation 1359,36 1088,39 11

Spain Direct Compensation 843,86 1100,77 7

Switzerland Direct Compensation 2694,50 1424,82 2

UK Direct Compensation 1463,71 1193,64 17

Total Direct Compensation 2005,99 1794,97 143

(21)

20

4.4 CSR and executive compensation

The control variables have been submitted to a regression analysis, the results of which are shown in Table 63_{. The results indicate that NI (β = -0,004 ; p < 0,01), REVLOG (β = -0,222}

; p < 0,01), MCAPLOG (β = 0,117 ; p < 0,05) and INDEP (β = 0,729 ; p < 0,01) are

significantly correlated to the degree of long-term compensation. However, it is interesting to point out that, while market capitalisation is positively correlated to long-term compensatio n, revenue is negatively correlated. This is remarkable because both variables proxy for firm size. Board independence is significantly and positively correlated with long-term compensatio n, which is in line with prior literature (Ghosh & Sirmans, 2005).

3_{The industry variable is not represented in tables 6 and 7 but has been included as a control variable in the}

analysis.

Variable Name Coefficient Estimate Std. Error

Constant 0,071 0,268 ROA 0,001 0,002 NI -0,004*** 0,001 REVLOG -0,222*** 0,044 MCAPLOG 0,117** 0,053 INDEP 0,729*** 0,228 AGE 0,006 0,004 GENDER -0,040 0,053 N F-Statistic Adjusted R2

Dependent variable: LCOMP *Significance at the 10% level **Significance at the 5% level ***Significance at the 1% level

Table 6 - OLS Regression on Control Variables

0,276 8,745***

(22)

21

4.5 Culture, CSR and executive compensation

The first hypothesis predicts that a higher degree of CSR orientation (a higher score on the Equal Weighted Rating) will have a positive effect on the percentage of long- term compensation that an executive receives. Model 1 in Table 7 shows the model used to test this hypothesis. The results indicate that a higher degree of CSR orientation is not significa nt ly correlated with the percentage of long-term compensation. No evidence was found that indicates that a higher degree of CSR orientation is positively related with the percentage of long-term compensation.

IDV (β = -0,011 ; p < 0,01) and MAS (β = -0,005 ; p < 0,01) are both significantly and negatively correlated with LCOMP, but their coefficients are incredibly small meaning they cause almost no change in the degree of long-term compensation received. As the Hofstede dimensions were expected to moderate the relationship, and not have a direct influence on LCOMP, these findings are conform expectations.

Independent variables Coefficient Estimate Std. Error Coefficient Estimate Std. Error Constant 1,672*** 0,519 1,672*** 0,519 UAI -0,002 0,002 -0,002 0,002 IDV -0,011*** 0,004 -0,011*** 0,004 MAS -0,005*** 0,002 -0,005*** 0,002 EQRATE 0,002 0,002 0,002 0,002 UAI x EQRATE 0,000 0,000 IDV x EQRATE 0,000 0,000 MAS x EQRATE -0,000** 0,000 Control Variables ROA -0,002 0,002 0,000 0,002 NI -0,005*** 0,001 -0,003*** 0,001 REVLOG -0,329*** 0,056 -0,248*** 0,052 MCAPLOG 0,111* 0,064 0,122* 0,065 INDEP 0,710*** 0,214 0,551** 0,219 AGE 0,011*** 0,004 0,007* 0,004 GENDER -0,045 0,052 -0,042 0,053 N F-Statistic Adjusted R2

Dependent variable: LCOMP *Significance at the 10% level **Significance at the 5% level ***Significance at the 1% level

Table 7 - Moderated OLS Regression on Long Term Compensation

143 143

0,346 0,303

Model 1 Model 2

(23)

22 The second and third hypotheses predict interaction effects caused by cultural differenc es between countries, measured on Hofstede’s cultural dimensions. Model 2 in Table 7 shows that a significant negative relation was found between MAS x EQRATE and LCOMP (β = -0,000

; p < 0,01), which is in line with expectations. It should be noted however that the explanatory

power of this variable is incredibly small. However, no significant relations were found between moderator variables UAI x EQRATE and IDV x EQRATE, and the dependent variable LCOMP. The control variables NI, REVLOG, MCAPLOG and INDEP are once again significantly correlated. In addition, executive age is also significantly and positively related to long-term compensation.

Based on the results hypotheses 2 and 3 are not supported, and therefore it is concluded that the effect of cultural differences is not significantly correlated with the relations hip between CSR orientation and the degree of long-term compensation received by executives.

4.6 Robustness

All regression models, in both Table 6 and 7, were repeated with a different proxy for the degree of CSR orientation (CSRESG) as the independent variable. The findings from those regressions are consistent with the findings presented in the tables above, and thus confirm the conclusions that were previously reached.

(24)

23

Discussion

This section discusses the main findings of the paper and the implications these results may have for researchers and practitioners. Last, the limitations of the current paper are discussed.

5.1 Summary of findings

This paper has aimed to link corporate social responsibility to the degree of long- term compensation in an executive’s compensation package. It has also attempted to identify the influence of cultural differences between European countries on this relation. Through the concepts of institutional theory, national cultures were argued to be representative of organisational cultures in specific countries. This paper provides new insights as previous research has mostly been done in a single country and mostly the US. This paper also used an index to measure CSR orientation in order to increase the validity of CSR indices. It was expected that the more CSR oriented a firm is, and by extension its executives are, the higher the percentage of long-term compensation in the total compensation package would be. In addition, it was expected that in countries with high scores on the uncertainty avoidance dimension and low scores on the individualism and masculinity dimensions this relations hip would be stronger, while in countries with opposite scores this relationship would be weaker. This expectation was formed based on institutional theory, which states that through a process of adoption and adaptation, national cultures will influence the establishment of norms and routines in a firm and thus influence organisational culture.

The data were analysed using OLS regression and moderated OLS regression models, using data on 143 executives from 8 European countries. Two different proxies for the degree of CSR orientation were used with similar results. None of the hypotheses are supported by the results of the analysis.

The first hypothesis, linking CSR orientation to long-term compensation, may be non-significant because the executive compensation setting process is not tailored to compensate executives properly for CSR goals. If the compensation package does not reward the executive for making progress towards long-term goals, it is to be expected that no relationship was found. This information may be useful to Boards of Directors that set compensation packages. The second and third hypotheses, predicting the interaction effect of cultural differences, may be non-significant due to a false assumption that business cultures are comparable to national cultures. Societies are becoming ever more diverse, and this could cause the business culture to differ from the national culture of the country in which a firm operates.

(25)

24 People from other countries will usually bring a different culture with them to their new home country and new firm, resulting in a mismatch between society’s goals and firms’ compensation systems.

Another potential cause of these divergent results is a selection bias in the data. Due to the limited availability of CSR data, only 143 out of the original 650 observations were complete and usable for the analysis. Excluding almost 80% of the original sample may have resulted in the remaining data being only representative of firms that have a high degree of CSR, thus flawing the analysis. Also, due to the selection bias only a small number of observations remain, possibly making the sample too small to be representative.

5.2 Implications

The answers to the two research questions stated in the introduction are that there has not been found a significant correlation between CSR orientation and long-term executive compensation, and that there was also no significant moderating effect caused by country-specific culture differences. As discussed in the introduction, the literature on the relations hip between the degree of CSR orientation and the degree of long-term executive compensation is not unified. Mahoney & Thorne (2005) found that more long-term compensation led to less CSR weaknesses in the company, while McGuire (2003) found the opposite. This paper has found that there is no relationship between the two, thus not favouring either argument but actually adding a possibility to the spectrum of explanations, namely that there is no relationship between these two concepts. Also, other compensation proxies, such as total compensation (Brockman et al., 2016) and pay duration (Gopalan et al., 2014), have been used in prior research and may yield different results.

The mismatch between firm goals and executive compensation packages and incent ives that was mentioned earlier should also be addressed. In light of the results of this paper it could be concluded that, because no significant relationship was found between CSR orientation and the degree of long-term compensation, that there is indeed a mismatch between which goals a firm deems important and aims to incentivise, and which goals the firm actually incentivises.

By approaching the relationship between CSR orientation and executive compensatio n from an institutional perspective, this paper contributes to the existing understanding of this relationship. No significant moderating effect of culture was found, meaning that perhaps institutional theory is not an adequate lens for viewing the CSR orientation and executive compensation relationship.

(26)

25 In addition, if the degree of CSR orientation does not influence the way in which executives are compensated, then supervisory boards may want to become more sensitive to setting compensation in line with firm goals.

5.3 Limitations and recommendations

In defining the proxy for the degree of CSR orientation as an index, the assumption was made that all four areas weigh evenly in the total score since no justification for different weight ings was found in prior literature or theory. However, it may be interesting to study the effects of different weightings of each area in order to gain more insight into which elements of CSR are most important.

Another limitation in this paper is its sample size. It is very small, and not very balanced across the eight included countries. This paper could be repeated with more observations in more countries in order to produce more generalisable and valid results.

For future research, it would be interesting to imple ment different measurements for long-term compensation using vesting periods (such as pay duration, Gopalan et al., 2014) since this paper did not have the proper resources to study this. Using vesting periods means that the measurement of exactly how long-term the compensation is can be measured more precisely. Also, future research could investigate the link between short-term compensation and CSR. As stated before, executives may find it more appropriate to receive a lower total compensatio n due to virtue ethics. Therefore, it may be possible that short-term compensation is related to CSR orientation.

In addition, this paper has operated on the assumption that CSR orientation is (partially) an explaining factor for executive compensation packages. This is in line with prior research (Cai et al., 2011; McGuire, 2003). However, previous research also (Mahoney & Thorne, 2006; Jian & Lee, 2015) examines the relationship the other way around, assuming that executive compensation causes a certain degree of CSR orientation. Future research should address this causality issue.

Last, due to the low R-square score in the complete model it is very likely that there are omitted variables that were not included in the model but are significantly and strongly correlated with long-term compensation. These omitted variables could possibly be addressed in future research. Livne et al. (2013) introduce some variables that may be interesting for future research, including leverage, executive tenure and a measure of risk.

(27)

26

References

Academic literature

Amihud, Y. and Lev, B. (1981). Risk Reduction as a Managerial Motive for Conglomerate Mergers. Bell Journal of Economics, 12, 605-617.

Becher, D., Campbell, A.T., Frye, M.B. (2005). Incentive compensation for bank directors: the impact of deregulation. Journal of Business, 78 (5), 1753–1778.

Brewer, E., William, H.C., William, E. (2004). Investment Opportunity Set, Product Mix, and the Relationship Between Bank CEO Compensation and Risk-Taking. FRB of Atlanta, Working Paper No. 2004-36.

Brockman, P., Lee, H. S. G., & Salas, J. M. (2016). Determinants of CEO compensatio n: Generalist–specialist versus insider–outsider attributes. Journal of Corporate Finance, 39, 53-77.

Bryan, S., Nash, R., & Patel, A. (2012). Culture and CEO compensation. Fordham University

Working Paper.

Cai, Y., Jo, H., & Pan, C. (2011). Vice or virtue? The impact of corporate social responsibilit y on executive compensation. Journal of Business Ethics, 104 (2), 159-173.

Callan, S. J., & Thomas, J. M. (2011). Executive compensation, corporate social responsibilit y, and corporate financial performance: a multi‐equation framework. Corporate Social

Responsibility and Environmental Management, 18 (6), 332-351.

Cavaco, S., & Crifo, P. (2014). CSR and financ ial performance: complementarity between environmental, social and business behaviours. Applied Economics, 46 (27), 3323-3338. Chopra, S., & Wu, P. J. (2016). Eco-activities and operating performance in the computer and

electronics industry. European Journal of Operational Research, 248 (3), 971-981. Daszynska-Zygadlo, K., Slonski, T., & Zawadzki, B. (2016). The Market Value of CSR

Performance Across Sectors. Engineering Economics, 27 (2), 230-238.

Del Bosco, B., & Misani, N. (2016). The effect of cross-listing on the environmental, social, and governance performance of firms. Journal of World Business, 51 (6), 977-990. DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphis m

and collective rationality in organizational fields. American Sociological Review, 48, 147-160.

Dreher, G. F., Lee, J. Y., & Clerkin, T. A. (2011). Mobility and cash compensation: The moderating effects of gender, race, and executive search firms. Journal of

(28)

27 Elston, J. A., & Goldberg, L. G. (2003). Executive compensation and agency costs in Germany.

Journal of Banking & Finance, 27 (7), 1391-1410.

Ghosh, C., & Sirmans, C. F. (2005). On REIT CEO compensation: does board structure matter?. The Journal of Real Estate Finance and Economics, 30 (4), 397-428.

Gimenez, C., Sierra, V., & Rodon, J. (2012). Sustainable operations: Their impact on the triple bottom line. International Journal of Production Economics, 140 (1), 149-159.

Gjølberg, M. (2009). Measuring the immeasurable?: Constructing an index of CSR practices and CSR performance in 20 countries. Scandinavian Journal of Management, 25 (1), 10-22.

Gomez-Carrasco, P., Guillamon-Saorin, E., & Garcia Osma, B. (2016). The illusion of CSR: drawing the line between core and supplementary CSR. Sustainability Accounting,

Management and Policy Journal, 7 (1), 125-151.

Gu, Z., & Kim, H. (2009). CEO cash compensation determinants: an empirical examination of US airlines. The Service Industries Journal, 29 (7), 995-1005.

Hofstede, G. (1983). National cultures in four dimensions: A research-based theory of cultura l differences among nations. International Studies of Management & Organization, 13 (12), 46-74.

Holmstrom, B. (1979), Moral Hazard and Observability. Bell Journal of Economics, 10, 74-91.

Hong, B., Li, Z. & Minor, D. (2015). Corporate Governance and Executive Compensation for Corporate Social Responsibility. Journal of Business Ethics, 136, 199-213.

Hooghiemstra, R., Hermes, N., & Emanuels, J. (2015). National culture and internal control disclosures: A cross‐country analysis. Corporate Governance: An International

Review, 23 (4), 357-377.

Husted, B. W., & de Sousa-Filho, J. M. (2017). The impact of sustainability governance, country stakeholder orientation, and country risk on environmental, social, and governance performance. Journal of Cleaner Production, 155, 93-102.

Jensen, M. and W. Meckling (1976), Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure. Journal of Financial Economics, 3, 305-360.

Jensen, M. C., & Murphy, K. J. (1990). Performance pay and top-manageme nt incentives. Journal of political economy, 98 (2), 225-264.

Jian, M., & Lee, K. W. (2015). CEO compensation and corporate social responsibility. Journal

of Multinational Financial Management, 29, 46-65.

Jitmaneeroj, B. (2016). Reform priorities for corporate sustainability: environmental, social, governance, or economic performance?. Management Decision, 54 (6), 1497-1521.

(29)

28 Leung, K. (1989). Cross-cultural differences: Individual- level vs. culture- le ve l

analysis. International Journal of Psychology, 24 (6), 703-719.

Livne, G., Markarian, G., & Mironov, M. (2013). Investment horizon, risk, and compensatio n in the banking industry. Journal of Banking & Finance, 37 (9), 3669-3680.

Mahoney, L. S., & Thorne, L. (2005). Corporate social responsibility and long- term compensation: Evidence from Canada. Journal of Business Ethics, 57 (3), 241-253. Mahoney, L., & Thorne, L. (2006). An examination of the structure of executive compensatio n

and corporate social responsibility: A Canadian Investigation. Journal of Business Ethics, 69, 149–162

Matsumura, E. M., & Shin, J. Y. (2005). Corporate governance reform and CEO compensatio n: Intended and unintended consequences. Journal of Business Ethics, 62 (2), 101-113. McGuire, J., Dow, S., & Argheyd, K. (2003). CEO incentives and corporate social

performance. Journal of Business Ethics, 45 (4), 341-359.

McWilliams, A., & Siegel, D. (2001). Corporate social responsibility: A theory of the firm perspective. Academy of management review, 26 (1), 117-127.

Murillo, D., & Lozano, J. M. (2006). SMEs and CSR: An approach to CSR in their own words. Journal of Business Ethics, 67 (3), 227-240.

Murphy, K. J. (2000). Performance standards in incentive contracts. Journal of Accounting and

Economics, 30 (3), 245-278.

Newman, K. L., & Nollen, S. D. (1996). Culture and congruence: The fit between manage me nt practices and national culture. Journal of international business studies, 27 (4), 753-779. Ramasamy, B., & Yeung, M. (2009). Chinese consumers’ perception of corporate social

responsibility (CSR). Journal of Business Ethics, 88, 119-132.

Ryan, H. E., & Wiggins, R. A. (2001). The influence of firm-and manager-specif ic characteristics on the structure of executive compensation. Journal of Corporate

Finance, 7 (2), 101-123.

Samiee, S., & Jeong, I. (1994). Cross-cultural research in advertising: an assessment of methodologies. Journal of the Academy of Marketing Science, 22 (3), 205.

Sapp, S. G. (2008). The impact of corporate governance on executive compensation. European

Financial Management, 14 (4), 710-746.

Shan, Y., & Walter, T. (2016). Towards a set of design principles for executive compensatio n contracts. Abacus, 52 (4), 619-684.

Sheehy, B. (2015). Defining CSR: Problems and solutions. Journal of Business Ethics, 131 (3), 625-648.

(30)

29 Shen, C. H., & Chang, Y. (2009). Ambition versus conscience, does corporate social responsibility pay off? The application of matching methods. Journal of Business

Ethics, 88, 133-153.

Skouloudis, A., Isaac, D., & Evaggelinos, K. (2016). Revisiting the national corporate social responsibility index. International Journal of Sustainable Development & World Ecology, 23 (1), 61-70.

Smith, P. B., Dugan, S., & Trompenaars, F. (1996). National culture and the values of organizational employees a dimensional analysis across 43 nations. Journal of

cross-cultural psychology, 27 (2), 231-264.

Soares, A. M., Farhangmehr, M., & Shoham, A. (2007). Hofstede's dimensions of culture in international marketing studies. Journal of business research, 60 (3), 277-284.

Spence, L. J. (2016). Small business social responsibility: Expanding core CSR theory. Business & Society, 55 (1), 23-55.

Tosi, H. L., & Greckhamer, T. (2004). Culture and CEO compensation. Organization

Science, 15 (6), 657-670.

Wu, M. W., & Shen, C. H. (2013). Corporate social responsibility in the banking industr y: Motives and financial performance. Journal of Banking & Finance, 37 (9), 3529-3547. Zilber, T. B. (2012). The relevance of institutional theory for the study of organizatio na l

culture. Journal of Management Inquiry, 21 (1), 88-93.

Books

DiMaggio, P. J., & Powell, W. W. (1991). The new institutionalism in organizatio na l analysis (Vol. 17). Chicago, IL: University of Chicago Press.

Schwartz, S. H. (1994). Beyond individualism/collectivism: New cultural dimensions of values. London: Sage Publications.

Scott, W. R. (2004). Institutional theory: Contributing to a theoretical research program. In K. G. Smith, & M. A. Hitt, (Eds.), Great minds in management: The process of theory development (pp. 460-485). Oxford: Oxford University Press.