The Relationship Between Criticism from
Important Countries and Firm CSR Performance
M.E. Hoornenborg 10678336
31 August 2015
Final Version Master Thesis
MSc Business Administration – Strategy Track Supervisor: Daniel Waeger
STATEMENT OF ORIGINALITY
This document is written by Student Mathijs Engelbert Hoornenborg 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.
ABSTRACT
Because of globalization, increased standard of living in developing and developed countries and the rise of new media, the influence of external stakeholders over large multinational firms is growing. Because of this, pressure for firms to engage in corporate social responsible behavior is becoming an important part of the corporate strategy of firms. Not all stakeholders have equal influence over firms, a stakeholder needs to be perceived as having power, legitimacy and urgency in order for decision makers to attend to their needs and demands. Because of bounded rationality and organizational attentional processes, firms and decision makers can only focus their attention on a limited amount of stakeholder claims. Building on this foundation this thesis investigates the influence of criticism from important countries on CSR performance of firms. The results, disproved the hypotheses that criticism from important countries have an effect on CSR performance. The proportion of criticism from the home country of the firm was related to a higher level of environmental performance in firms. This research connects the external and internal influences that decision makers face in regards to corporate social responsible behavior in firms which has implications for theory and practice.
Keywords: Corporate Social Responsibility, Stakeholder Theory, Attention Based View of the Firm, Criticism, Globalization
TABLE OF CONTENTS
STATEMENT OF ORIGINALITY ... 2
1. INTRODUCTION ... 5
2. LITERATURE REVIEW ... 7
2.1 Corporate Social Responsibility ... 7
2.2 Stakeholder Theory ... 9
2.3 Attentional Processes ... 11
2.4 Research Question and Gap ... 14
3. THEORETICAL FRAMEWORK ... 18
3.1 Organizational Attention to Stakeholder Criticism ... 18
3.2 Degree of Human Development Across Firm’s Operations ... 20
3.3 Home Country Importance and Development ... 22
4. METHODS ... 24
4.1 Sample and data collection ... 24
4.2 Measures ... 26
4.2.1 Dependent variable ... 26
4.2.2 Independent variable ... 28
4.2.3 Moderating variables ... 29
4.2.4 Control variables ... 30
4.3 Statistical analysis and results ... 31
5. DISCUSSION ... 46
5.1 Academic Relevance ... 47
5.2 Managerial Implications ... 48
5.3 Limitations and Suggestions for Future Research ... 48
6. CONCLUSION ... 50
1. INTRODUCTION
In this day and age stakeholder influence and pressure on firms have become more and more prevalent. As the standard of living of nations grows, so does the influence of social activist stakeholders (Werther & Chandler, 2005). A notable example is influence on the oil and chemical firms. In 1989, off the coast of Alaska an Exxon oil tanker struck a reef there, spilling gallons of crude oil into the ocean, causing one of the greatest man-made environmental disasters, having a devastating effect on local wildlife there. Protests from environmentalists and shareholders pressured Exxon to add an environmentalist to its board with other oil and chemical firms forced to follow suit (Teoh, Welch and Wazzan 1999).
Social activism is not confined to a single industry or firm, but rather diverse and pressure groups can be found in all markets and segments, forcing firms to react to these groups, but also to proactively try to please these social stakeholders (Porter & Kramer, 2006). Social activist stakeholders are just as heterogeneous in nature as the industries where they act in and can range from environmentalist groups, political groups to religious groups.
Pressure from external stakeholders on the firm have been shown to impact firms in different ways. Protests from communities and activist groups can impact where firms open up new stores within a country (You, Rao & Ingram, 2013), in which country a firm opens a new store (Soule, Swaminathan & Tihanyi, 2014) and which industries to invest in, or more importantly, which industries not to invest in (Weber, Rao & Thomas, 2009).
This is also the case with big multinational firms, which face multiple socio-political environments with different stakeholders per country and environment which all ask different things from the firm. Research shows that stakeholder actions can drive firms
to divest their assets in certain countries to other countries, as has been the case in South Africa during the apartheid era (Posnikoff, 1997; Teoh, Welch & Wazzan, 1999) and in Burma, during one of the longest running civil wars (Soule, Swaminathan & Tihanyi, 2014).
The influence of these groups of stakeholders of firms do not always have the same power and influence in order to achieve actual change. Most of the changes in
pro-environmental responses in firms can be accounted for by just a small group of external stakeholders, namely campaigners, regulators (Fineman & Clarke, 1996) and consumers (Mohr & Webb, 2005).
So we see that stakeholders can have an impact on the multinational firm, but not all external stakeholders are perceived as equal and have the same influence over
multinational firms. This thesis tries to explain why some stakeholder actions in certain countries have larger impact on multinational firms than other stakeholder actions from other countries. I will base my thesis on stakeholder theory and attentional processes within human decision makers and organizations to answer this question.
This thesis begins with a review of the available literature on corporate social responsibility, stakeholder theory and attentional processes. After discussing the relevant literature, a theoretical framework will be drawn and hypotheses will be formed based on de previously discussed literature. Third, research methods are designed to answer the research questions posed in the theoretical framework. Forth, the empirical results are presented and subsequently discussed. Finally the conclusions of the thesis in regards to theory and practice are being drawn.
2. LITERATURE REVIEW
In the following paragraph I will discuss the relevant literature on corporate social responsibility and decision making processes made by organizations in respect to
stakeholder criticism. I will first discuss what exactly corporate social responsibility is, secondly I will explain the concepts of stakeholder theory, thirdly I will outline the main findings in research on attentional processes within firms and lastly the research gap and research questions will be discussed.
2.1 Corporate Social Responsibility
For many years, community and social development goals were philanthropic activities that were seen as separate from business objectives; doing well business wise and doing good community wise were seen as separate pursuits (Kotler & Lee, 2005). Research in the field of Corporate Social Responsibility (CSR) originated in the 1960’s due to the belief that several hundred of the largest firms in the world held the decision making power to influence the lives of citizens on many different points (Carroll, 1999). Hence emerged the questions on which responsibilities to society should fall on the decision makers of these large firms.
In more recent times with the emergence of more and more media outlets, the consequences of decisions made by firms have become more visible. Because of this, governments, media and activists are becoming more adept at holding firms accountable for the social consequences of their activities. Currently there are a lot of organizations which rank firms on the performance of their corporate social responsibility, releasing
these rankings to the general public. As a result, CSR has become an inescapable priority for decision makers within the firm.
Because CSR is such a broad concept, it is hard to define what actions encompass CSR policies. The term CSR is often used as an umbrella term for actions taken by the firm in order to contribute to social and environmental welfare, regardless of profit or revenue maximization (McWilliams, 2000). CSR addresses issues like: human rights, unfair business practices, organizational governance, environmental aspects, marketplace and consumer issues, community involvement, social development, workplace and employee issues, and occupational health and safety issues (Leonard & McAdam, 2003). To make the definition more concrete, Campbell (2007) has defined corporations as acting in a social responsible way if they do two things. First, they must not knowingly do anything that can harm their stakeholders like employees, investors, suppliers, customers or the community in which they operate. Second if corporations do cause harm to one of the above stakeholders, they must rectify it when the harm is discovered or brought to their attention.
Because of the focus on CSR performance across different media outlets, NGO’s and governmental organizations, large firms have set in place policies and departments to actively monitor the state of CSR performance and improve it where necessary. With good reason, because besides a moral obligation to be a responsible firm, CSR performance also has impact on consumer behavior. CSR performance has been shown to positively impact evaluation of the firm and purchase intent (Mohr & Webb, 2005). Moreover, performance on the environmental responsible domain affected purchase intent more strongly than price did. From a financial perspective, the CSR performance has been shown to have a positive impact on corporate financial performance (Ruf, Muralidhar, Brown, Janney and Paul, 2001). A positive change in CSR performance was associated with growth in sales in the
current year and next year, showing that there are short-term benefits from improving CSR performance. Motives for firms to engage in CSR are not always purely motivated by social responsibility, many firms use CSR programs as a tool to further their strategic plans, competitive advantage, or increase their legitimacy in a certain sector (Babiak and Trendafilova, 2011).
Because firms are constantly being held accountable by different groups of organizations, both distal and focal, the pressure to implement CSR programs is mostly coming from multiple groups. Which groups have the most influence over CSR programs within a firm has been the subject of a 70 study meta-analysis done by Clarkson (1995). This study had two important findings. First, that firms manage relationships with
stakeholder groups, rather than with societies as a whole and second that it is important to distinguish between social issues and stakeholder issues. In regards to this, firms tend to focus their attention on stakeholder groups. The more important these stakeholder groups are to the firm, to more likely the firm is to focus the attention on issues of these
stakeholder groups. The complex environment of stakeholder groups and stakeholder management is discussed in the next paragraph, where stakeholder theory is explained.
2.2 Stakeholder Theory
There has been a variety of research on the impact of the external socio-political environment on the actions and decisions made by the firm. Earlier, more traditionalistic views of corporate strategy had a narrow view of stakeholders as being the owners or shareholders of the firm (Clement, 2005), leading to a shareholder view of the firm. Freeman (1984) proposed that firms should not only consider the needs and demands of their shareholder, but also consider those of a wide range of external entities, or
“stakeholders” in order to be competitive. This research was the basis of the widespread theory called “stakeholder theory”, which has been expanded on over the last couple of
decennia. This research view firms as economical entities existing in a complex environment with stakeholders with divergent interests and inputs (King, 2008).
Stakeholders, like consumers, employees, shareholders and communities all make, often different, claims on firms. In order to come to acceptable corporate, strategic and social performance, firms need to appropriately manage its relationship with the various stakeholders (Freeman & Gilbert, 1987).
Not all stakeholders have equal influence over the firm and its decision making processes as claims from stakeholders may be lacking in urgency or might be competing with claims from other stakeholders relative to the firm. In order to have influence over the workings of the firm, stakeholders must possess three important characteristics; power, legitimacy and urgency (Freeman, 1994; Agle, Mitchell and Sonnenfeld, 1999). If a stakeholder has power over a firm, a stakeholder can impose its will upon the firm. Legitimacy of the stakeholder implies that the stakeholder demands comply with the prevailing norms and goals within a firm. The last characteristic, urgency is consists of two elements, the importance stakeholders give to their own demands and their sensitivity to how long it takes firms to deal with their demands. Possessing these qualities affect stakeholder salience, the degree to which managers give priority to competing stakeholder claims.
Research building on the notion of manager his perception of stakeholder characteristics shows that it is not necessary for stakeholders to have these three actual qualities in order to have stakeholder salience, but that managers’ perception of these three
qualities in stakeholders prove a better indicator of firm decision making processes (Mitchell, Agle and Wood, 1997). In practice, this stakeholder balancing exercise
necessary to satisfy stakeholders by the firm has proven difficult to enact, and research suggest that managers cannot produce every kind of social value for every kind of stakeholder and are instead constrained by their cognitive resources and bounded rationality, thus prioritizing stakeholders on instrumental or normative considerations instead of actual rational considerations (Galbreath, 2006; Jamali, 2008). Accordingly, stakeholder claims and decisions based on these claim need to be seen in the light of human and organizational constraints and attentional structures within the firm.
2.3 Attentional Processes
Organizational theory traditionally looks at managers and decision makers as boundedly rational. The decision making process is constrained by the cognitive capabilities of the decision makers (March & Simon, 1993). Because of these cognitive constraints decision makers have difficulty making an optimal, rational decision. Instead they often choose for an acceptable solution which is based on heuristics, rather than an optimal one (Simon, 1979). In order to cope with these cognitive constraints, people rely on learning processes, attentional processes and heuristics in order to increase their decision making potential (Gigerenzer and Selten, 2002).
Because managers are boundedly rational, their cognitive capabilities are restrictive in a way that managers cannot make decisions based on all the rational information, leading to satisficing decisions instead of optimal ones (Simon, 1979). The cognitive capabilities and limitations of decision makers lead to a lack of overview of the entire situation. This leads to decision making processes which are only based on a small part of the relevant
information, the part where the attention of decision makers and the firm is focused on (Cyert & March, 1963).
Attentional processes within organizations have been a subject of academic study for a long time. Decision making processes based on attention have its roots in the research done on the subject of bounded rationality where, decisions made by the firm are also limited by the cognitive limitations of the individual decision maker (Simon, 1965). Firm behavior for Simon was not only the result of structural processes placed by the firm, but also cognitive processes. So decision making in organization was a product of both the limited attentional capacity of human behavior and structural influences of organizations on individual attention.
A step forward in attentional processes within organizations comes from Ocasio (1997) where the author explicitly links structure and cognition into an attention based view of the firm. This attention based view of the firm links individual information processing to organizational attention structures present in the firm and its central
argument is that firm behavior is a result of how firms channel and distribute the attention of decision makers (Ocasio, 1997), calling this organization attention.
The attention based view of the firm consists of three theoretical principles and levels on how firms distribute and regulate attention of its decision makers. The first principle is called the focus of attention and describes the individual level of cognition. Focused attention both facilitates action and perception towards the tasks and activities which are being tended to and inhibits perceptions and actions towards tasks and activities towards that are not tended to (Kahneman, 1973; Kahneman & Tversky; 1973). The result is an enhanced mindfulness and vigilance of individuals regarding the area of interest, which generates a selective focus of attention towards an idea or object and facilitates action and perception towards that object.
The second principle is the based on situated attention and describes the importance of social context on what decision makers focus their attention to. This principle is
grounded on research findings in social psychology. According to this research the focus of attention of decision makers is based on the characteristics of the situation that the decision maker finds himself in and this situated attention then in turn directly shapes the behavior of the decision maker (Ross & Nisbett, 1991). Because of this, decision makers will vary their focus of attention depending on the situation and the individual behavior is dependent more on the characteristics of the situation rather than the characteristics of the individual (Belk, 1975). In the organizational context, according to the principle of situated attention, the focus of attention and action of decision makers is shaped by the organizational and environmental context.
The third and last principle of the attention based view of the firm is the principle of structural distribution of attention and is one of distributed nature of attention. The
attention processing of a firm is not a shared or collective activity, but one that is
distributed throughout the various procedures and the social structure of the organization (Hutchins, 1995). This principle proposes that the specific context that decision makers find themselves in and how they attend to this context is dependent on how the
organization distributes and controls the allocation of issues, answers and decision makers into specific activities and procedures (Ocasio, 1997).
An attention based view of the firm in this way accounts for the individual differences between firms in attention processing and attention setting, guided by
differences within these three principles of attention, namely the individual differences in the focus of attention of individual decision makers, the social and external context which these decision makers find themselves in and the way that attention is distributed within the organization.
2.4 Research Question and Gap
The current research tries to explain why stakeholder actions from some countries have stronger impact on multinational firms than stakeholder action from other countries by combining stakeholder theory with an attention based view of the firm. Previous
research on stakeholder influence shows that in order to have impact on firms, stakeholders must possess three qualities, namely power, legitimacy and urgency (Freeman, 1994; Agle, Mitchell and Sonnenfeld, 1999). Decision makers base their stakeholder management decision based on the perceived salience of these stakeholders.
There has been some research on stakeholder salience in relation to where stakeholder action is coming from. A lot of this research has been done on stakeholder action from developing countries. Hart & Sharma (2004) noted that perceived stakeholder salience of farmers in developing countries was low, with decision makers rating the power and legitimacy low, although the urgency of these farmers was high. Because these
developing countries are relatively unimportant to multinational firms in regards to sales, the urgency for change is lower for firms (Reed, 2002). In secondary stakeholders, stakeholders which do not have a direct contractual obligation to the firm, urgency has been proven to elicit the least corporate action in large American firms in regard to environmental issues (Eesley & Lenox, 2006).
This previous research looked at countries as a strict dichotomy; on the one hand you have developed countries, which have implicitly been coded as being an important country to the firm, and developing countries, which have implicitly been coded as being relatively unimportant to the firm. This dichotomy is not in place in practice, not all developed countries are automatically important to the firm just like not all developing countries are unimportant to the firm.
How attentional processes play a role in impaction a firm and its decisions based on CSR performance has been less researched. Behavioral theory suggests that decision makers do not make optimal decisions based on all the information presented, but rather have bounded cognitive abilities, making satisficing decisions, rather that optimal ones (Simon, 1979). Human cognition can only process so much at a time, so attention setting is important to process the relevant information and dismiss irrelevant information (Ocasio, 1998). To what kind of stimuli a decision maker focusses his attention on is implicitly and explicitly guided by structures and procedures within the firm (Hutchins, 1995).
A firm structures its information stream in a way that decision makers pay attention to information which is important to the survival and success of the firm. In the case of stakeholder criticism originating from a certain country, it should be particularly sensitive to criticism originating from this country which is important to the firm, because being seen as an illegitimate firm in that country would have strong consequences for the firm. Also, when a country is important to a firm, it will pay more attention to criticism coming from that country than to criticism coming from another country; hence, it will be more likely to process the information.
Previous research done by Bouquet and Birkenshaw (2008) on attentional processes in multinational enterprises and their foreign subsidiaries combined both stakeholder theory and the attention based view of the firm. They investigated how foreign subsidiaries gained attention from the corporate headquarters. The decision makers at the corporate headquarters have to allocate their limited attention in a strategic manner in order to achieve successful decisions. They found that attention decisions are partially based on the importance of the subsidiaries within the corporate structure. Secondly, they also found that foreign subsidiaries could also attract attention from the corporate headquarters
entities which are bound by the attention structures in place within a multinational enterprise. They can also play an active part in attracting attention like voicing criticism and critique. This action taking in the form of taking initiative and voicing criticism will increase the perceived salience of the stakeholder, as voiced criticism is an indicator of urgency in the demand of the stakeholder (Clement, 2005).
In conclusion, we see that stakeholder criticism can raise stakeholder salience by increasing stakeholder urgency (Bouquet & Birkenshaw, 2008). But stakeholder criticism from developing countries, although urgency was high, is being perceived as lacking salience because of the lack of power and legitimacy, and thus will not have a large impact on the firm (Hart & Sharma, 2004). It seems that in order to be effective, stakeholder salience is dependent on the other two factors, power and legitimacy. This is where organizational attention begins to play a part. Legitimacy is achieved when stakeholder demands are in line with corporate norms and goals. According to the third principle of the attention based view, attention processing is distributed throughout the various formal and informal procedures of the firm together with the various social structures within the firm (Hutchins, 1995). Attention is then dependent on how the organization distributes and controls the allocation of attention throughout specific activities and procedures. (Ocasio, 1997). In this way attention is being guided towards stakeholders and criticism which is congruent with organizational goals, needs and wants. These goals, needs and wants are usually centered around production, sales, revenue and other factors which are of strategic importance to the firm. Claims and criticism from countries which are important to the firm will therefore be perceived as being higher on legitimacy than claims and criticism from countries which are less important to the firm and will be processed quicker and with greater detail, as they are congruent with the corporate goals.
Taking all the previous research into account, we see two distinct influences that shape decision making in firms. First we have the external influences, which consist of stakeholders, which apply pressure on the decision making processes of firms. Secondly, the internal influences, consisting of structural attentional processes which shape and direct the attention and decisions made by decision makers. Stakeholder salience is therefore dependent on organizational attention toward these stakeholder claims and criticism, and criticism from important countries will be higher in perceived legitimacy, having a greater effect on the firm. This research tries to investigate the importance of the origin of
criticism in relation to the impact which it has on the firm, without using the previously used dichotomy of developed countries versus developing countries. The research question which follows is:
Does criticism, which originates from a country which is important to the firm have a stronger impact on the firm than criticism, which originates from a country which is less important to the firm?
3. THEORETICAL FRAMEWORK
In this chapter the theoretical framework is created and hypotheses are posed based on the literature which is available. The first section discusses organizational attention processes in regards to stakeholder criticism. The second section investigates the role of the degree of human development in countries that a firm is active in. The last introduces the concept of the unique role which a home country plays for large multinational
corporations.
3.1 Organizational Attention to Stakeholder Criticism
One way to look at organizational attention is by use of the attention based view of the firm. The attention based view of the firm views decision makers and organizations as being boundedly rational which make decisions based on imperfect processing of available information (Cyert & March, 1963) instead of making perfect decision based of all
available information. What information decision makers and organizations use in order to make their decisions is largely based on where they focus their attention on.
This attention based view of the firm links individual information processing to organizational attention structures present in the firm and its central argument is that firm behavior is a result of how firms channel and distribute the attention of decision makers (Ocasio, 1997), calling this organizational attention.
Firms channel and distribute the attention processes of their decision makers based on values relevant for the firm. Often sales, production and revenue are the main drivers of organizational strategy. This means that, implicitly and explicitly, firms lead the decision making process in a certain way by guiding the attention of decision makers to maximize
those values that are important to the firm (Ocasio, 1998). In most cases this will lead to decisions with the outcome of maximizing profits, production and sales.
In regards to stakeholders and stakeholder theory, the amount of attention given to stakeholders is related to stakeholder salience (Freeman, 1994), with stakeholders which are perceived as having much power, legitimacy and urgency receiving the most attention from decision makers. It is to be expected that stakeholders which are in line with the corporate attentional structures are being perceived as being higher in power and legitimacy and they will most likely receive the most attention.
Large multinational firms will have to deal with many different stakeholders from many different countries and some countries will be more important to firms than others in terms of revenue, sales and production and will receive more attention from the
multinational firm. Because of this, stakeholders from these countries will automatically be perceived as having more salience and influence over the firm. Criticism from these
stakeholders will then also have a larger influence over and stronger impact on the firm. This will lead to our first hypothesis and is as follows:
H1. The more a firm is exposed to CSR-related criticism from countries that are important to that firm, the stronger this firms CSR performance.
H1 (+) Figure 1. Hypothesis 1 Extent of CSR related criticism from important countries CSR Performance
3.2 Degree of Human Development Across Firm’s Operations
Large multinational corporations operate in a large number of countries, which all have their own level of human development. The level of human development of a country influences the motivations and needs of its citizens. Maslow hypothesized that peoples’ motivations are based on a hierarchy of needs (Huitt, 2004). Each individual is motivated to achieve the needs on a certain level in order for his motivation to shift to higher level needs. These level of needs range from the most basic human need, physiological, the need to eat and sleep to the highest of human need, the need of self-actualization, creating a hierarchical structure of needs. If basic survival needs are not met, and an individual is starving, he is less motivated to take actions to achieve self-actualization.
A less developed country in general has more people tending to their physical needs, without fewer priorities on increasing self-esteem or self-actualization. On the other hand, people from a developed first world country have all of the physiological and safety needs met, creating opportunity to focus on higher level needs, like self-actualization and interest in politics, not just for their own interest, but also for the public good.
As a result of these needs being met, people in countries with a higher level of human development are more sensitive towards corporate social responsibility. Different researchers (Liu and Sibley, 2011; Arbuthnott, Devoe and Lawrie (2011) hypothesized that the amount of human development in a certain country positively influences the concern for climate change. They found that in a country where the Human Development Index (HDI) was high, so was the amount of concern about climate change. Corporate social responsibility is thus more ingrained in cultures with higher human development. Firms which operate in countries where there is a high level of human development will also be subject to these ingrained mores of firms and will be more sensitive towards CSR criticism
then if a firm is operating in countries with lower levels of human development. As a result, if a firm operates in countries with higher level of human development, criticism will be taken more into account and will affect the firm’s CSR performance and its will to change the CSR performance more than if a firm operates in countries with a lower level of human development, leading to a moderating effect and the second hypothesis:
H2. The level of human development in the countries a firm operates in moderates the relationship between CSR-related criticism from important countries and CSR
performance, such that this relationship is stronger when a firm operates in countries with higher human development scores
H1 (+)
H2 (+)
Figure 2. Hypothesis 1 and 2 Extent of CSR related
criticism from important countries riticism from country importance
CSR Performance
3.3 Home Country Importance and Development
As argued in section 3.1 the country where criticism originates from is important for the in predicting the actions of the firm. Criticism from more important countries have a larger impact on CSR performance than criticism from countries less important to the firm. The home country of a firm is often one of the most important countries to the firm in terms of attention (Patel & Vega, 1999). Multinational firms usually have a greater
influence in the home country, where the firm has adapted itself to the specific country environment, leading to greater opportunities and profitability (Le Bas and Sierra, 2002). Also the amount of sales, revenue, influence and size of the firm will often be largest in the home country (Hawawini, Subramanian and Verdin, 2004)
Over time, the firm has evolved in the external home country environment, leading to specific informational and attentional processes mostly relevant to this home country (Madhok and Phene, 2001). The attention structures in place will often be the result of years of co-evolving together with the home country, leading to greater fit of attention structures in the home country in relation to other countries.
Because of this special status of the home country for the firm, firms will process criticism from their home country with greater scrutiny than criticism from other countries. The criticism from their home country will have a greater impact on the firm name, status and profits than criticism from other countries. It is therefore expected that criticism from the home country will have a larger impact on corporate social responsibility than criticism from other countries, which in turn will lead to the third hypothesis for this thesis:
H3. The more a firm is exposed to CSR-related criticism from its home country, the stronger the firm’s CSR performance
The same argument for the level of human development can be made when looking at the home country. The home country of a firm will receive the most attention from decision makers within the firm (Patel & Vega, 1999). Ringov & Zollo (2007) found that country culture has a significant effect on corporate social performance. If the home country would be a country where the level of human development is high, sensitivity towards CSR related issues would also be high in the country, but also within the firm, leading to a higher expected CSR performance of the firm. The level of human
development is expected to have a positive moderating effect on the relationship between proportion of criticism originating from the home country and CSR performance, with all other things staying the same. Therefore, the fourth and final hypothesis posed in this thesis is as follows:
H4. Ceteris paribus, home country level of HDI positively moderates the relationship between criticism from country importance to the firm and CSR performance.
H3 (+)
H4 (+)
Figure 3. Hypothesis 3 and 4. Proportion of criticism from home country
CSR Performance
Home country level of HDI
4. METHODS 4.1 Sample and data collection
This study is has a cross sectional research design in order to determine the effect of the origin of criticism country on firm CSR performance. For the purposes of the present thesis, a unique dataset is developed by combining three data sources: Covalence
EthicalQuote, the Carbon Disclosure Project (CDP), Thomson Reuters Worldscope. Covalence EthicalQuote tracks online news on topics relevant to corporate social responsibility (CSR) based on 50 CSR-criteria developed by the Global Reporting Initiative (EthicalQuote, 2015). The analysts at Covalence EthicalQuote code each news item in terms of whether it constitutes criticism or praise directed towards the firm, the source of the news item (the firm itself or non-firm sources, such as the media), which country the news item has been originally published in and which country is concerned with the incident described in the news article. The CDP sends out each year a very detailed questionnaire to firms regarding their greenhouse gas emissions. Particularly relevant for the present study, the CDP asks multinational corporations (MNCs) both to report their total greenhouse gas emissions and to break down these total emissions by country. Finally, the Thompson Reuters Worldscope database reports the financial and other relevant firm-specific information used for the present study.
The original sample consisted of the 541 firms tracked by Covalence Ethicalquote for 2011. The year 2011 was chosen because it constitutes the most recent year full data from Covalence Ethicalquote was available. After combining the different data sources, the original sample of 541 firms was reduced to 240 firms, for which data was available.
These firms were situated among 22 different countries, with the most originating from the United States (39,00%), the United Kingdom (10,37%), Japan (10,37%), France
(6,22%) and Germany (5,39%). A full list of firms and their home countries can be found in table 1 below.
Table 1.
Frequency and percentages of firms in a certain country Country Frequency Percentage
USA 94 39,17% Japan 25 10,42% UK 25 10,42% France 15 6,25% Germany 13 5,42% Canada 10 4,17% Australia 9 3,75% Netherlands 7 2,92% Spain 7 2,92% Switzerland 7 2,92% Italy 5 2,08% South Africa 4 1,67% Finland 3 1,25% Norway 3 1,25% South Korea 3 1,25% Sweden 3 1,25% Brazil 2 0,83% Belgium 1 0,42% Denmark 1 0,42% Luxembourg 1 0,42% Malaysia 1 0,42% Taiwan 1 0,42%
4.2 Measures
4.2.1 Dependent variable
Because CSR programs encompass a large number of practices, from reducing emission from factories, to compensation workers with an honest wage to making sure the board of directors is just as diverse as the employees within the firms, defining a global CSR construct is inherently difficult (Gjølberg, 2009). It seems that looking at one CSR performance score does not adequately represent the relationship between variables and the CSR performance score must be entangled.
According to Cheng, Ioannou and Serafeim (2014) there are three global pillars of CSR; environmental performance, social performance and corporate governance.
Environmental performance consists of practices in place to focus on resource reduction, emission reduction and product innovation. Social performance on the other hand looks at employment quality, health and safety, training and development, diversity, human rights, community and customer and product responsibility. Lastly, corporate governance
encompasses board structure, compensation policy, board functions, shareholder rights and vision and strategy.
The dependent variable is the performance of the firm on Corporate Social
Responsibility. This value is constructed through information in the Worldscope ASSET4 database. CSR performance value consists of 16 categories divided among the three previously mentioned pillars; Environmental Performance, Social Performance and Corporate Governance. Specially trained research analysts collect 900 evaluation points per firm, where all the primary data used must be objective and publically available (Cheng, Ioannou and Serafeim, 2014). Table 2 shows all the categories of belonging to said pillars. The values of these three pillars (Environmental, Social and Corporate
Governance) are then added to create a global CSR score. As there are fundamental
differences between these three pillars making up CSR performance, all hypotheses will be tested not only against the CSR performance, but also against the individual pillars.
Because these variables are negatively skewed, with the mean being lower than the median, some data transformation is necessary to achieve normality in the variables. To achieve this, Bartlett (1947) proposes to subtract the score of a firm on CSR performance from the highest score on that variable. Then a logarithmic transformation was done over the highest score minus variable score. After the transformation, all variables fell within the acceptable skewness and kurtosis range of -1 and 1 (Shrivastava, 1984)
Table 2.
Pillars and Categories of Corporate Social Responsibility Performance
Pillars Categories
Environmental Performance Resource Reduction
Emission Reduction Product Innovation Social Performance Employment Quality Health and Safety
Training and Development Diversity
Human Rights Community
Customer / Product Responsibility
Corporate Governance Board Structure Compensation Policy Board Functions Shareholders Rights Vision and Strategy
4.2.2 Independent variable
The first independent variable used to answer hypothesis 1 in this thesis is the criticism from important countries variable. This was operationalized by using the data provided by the CDP project. To assess, which countries a firm is active in and how
important these countries are for the firm, data from the CDP was used. When filling in the CDP questionnaire, the firms report their total greenhouse gas emissions and they break these emissions down by the countries they are active in. The amount of greenhouse gas emissions of a particular country was divided by the total amount of greenhouse gas emissions. You then have the proportion of emission for that country for that firm. The amount of emission a firm produces in a particular country is used as a proxy for country importance, where the more a firm is active in a country, the more important it is.
The second part of the computation was creating a weighted country criticism index score. This was done by multiplying the country emission score by the amount of criticism a firm received in a specific country. In the last part of the computation I added all the weighted country criticism scores, generating the firm's exposure to criticism weighted by importance of countries in which the criticism was voiced.
This variable did not have a normal distribution, so the variable had to be transformed in order to achieve normality. The variable had a positive skewness, and according to Bartlett (1947) a logarithmic transformation has to be done in order to transform this variable. Because a lot of the variables have a score of zero, a logarithmic transformation over those zeros cannot be computed. So therefore a logarithmic
fell between the accepted range of skewness and kurtosis (between -1 and 1) (Srivastava, 1984)
This variable thus combines two pieces of information: the amount of criticism a firm receives and how important the countries, in which the criticism was voiced, are for the firm. A higher score indicates that a firm received more criticism and that this criticism came from more important countries. A lower score indicates that a firm received less criticism and that this criticism came from less important countries.
In order to test the third hypothesis we look at the proportion of criticism the firm received from the home country. This variable is operationalized by dividing the amount of criticism received in the home country by the total amount of criticism received by the firm.
4.2.3 Moderating variables
There are two moderating variables, both having to do with human development in a country and are operationalized by using the Human Development Index (HDI). The HDI is a composite variable of life expectancy, education and per capita income and is used to rank countries in four tiers of human development. The HDI is a score from 0 to 1, where a higher score indicates a higher level of human development.
To determine the degree to which a firm is exposed to a free press throughout the countries a firm is active in, data from the HDI was combined with data from the Carbon Disclosure Project. In order to determine the importance of each one of the countries for the focal firm, the amount of greenhouse gas emissions the firm emitted in a country was divided by the total amount of greenhouse gas emissions of the firm. This information was then combined with the Human Development Index scores for 2011. For each country a
firm is active in, the country’s Human Development Index score was multiplied by that country’s importance for the firm. This resulted in a weighted Human Development Index
score for each country the firm is active in. Once this weighted Human Development Index score had been computed for each one of the countries the firm is active in, these scores were added up. This resulted in the variable ‘exposure to a human development’. A lower score indicates exposure to lower human development and a higher score indicates
exposure to higher human development across the firm’s operations. This variable was negatively skewed, with the mean being higher than the median. After transforming the variable normality was still not achieved, so any results have to take this information into account
The second moderating variable is the rate of human development in the home country of the firm and also operationalized using the HDI. The HDI score of the home country is used, where a higher score indicates a higher level of human development in the home country of the firm and a lower score indicates a lower level of human development in the home country of the firm. As with the other human development variable, this variable did not have a normal distribution, even after transforming the variable, so caution has to be excised when interpreting the results.
4.2.4 Control variables
There are three control variables being used in this research. The first control variable is firm size, which is operationalized as the logarithmic value of the number of employees, which was gathered from the Thomas Reuters Worldscope Database. The number of employees is a commonly used as a proxy for firm size. The number of employees is usually skewed in firms, with the median number of employees in a given
industry being considerably less than the mean number of employees in this industry. To account for this skewedness we use the logarithmic value of the number of employees. The second control variable that is being used in this thesis is the firm profitability which was also gathered from the Thomas Reuters Worldscope Database, which is operationalized by looking at the return on assets (ROA). The third and last control variable is the type of industry (industrial, utility, transportation, banking, insurance and other firm).
4.3 Statistical analysis and results
The descriptive statistics of the dependent, independent and control variables used in this study are presented in table 3 on page 33. As discussed in the theoretical framework, CSR performance is composed of environmental performance, social performance and corporate governance. As CSR performance is composed of those variables, the values of these three pillars correlate significantly with CSR performance with environmental performance being 0.750, social performance being 0.7 and corporate governance being 0.572. We discussed the possibility that these three pillars are significantly different from each other, measuring a different construct. The correlations between these three pillars show mixed results, the correlations between environmental performance and social performance is significant, with a correlation of 0,589 as is the correlation between social performance with corporate governance (0,418). The relationship between environmental performance and corporate governance is not significant however, with a correlation of 0,027, meaning that there is no relationship between both pillars.
The correlation between criticism from important countries and the proportion of criticism from the home country are moderately related (0,237), as is the relationship between exposure to human development of the firm’s and the human development in the
home country of the firm (0,248). This is to be expected, as the home country of an MNE is often a relatively important country for that firm.
In order to test for multi-collinearity, a collinearity test is done for all the relationships between all the independent and control variables. The Variance Inflation Factor (VIF) is often used to test for multi-collinearity. VIF scores of 10 or higher are mostly regarded as being seen as an indicator of multi-collinearity (O’brien, 2007). All the independent and control variables have values of 4 and less in relation to each other, showing that there is no indication of multi-collinearity.
Table 3.
Descriptive Statistics and Correlations Among Variables in the Study
Mean St. Dev. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Dependent variables 1.CSR Performance 226,912 30,21 1 2.Environmental performance 81,98 14,14 ,750** 1 3.Social Performance 74,03 17,99 ,767** ,589** 1 4.Corporate Governance 67,29 19,046 ,572** ,027 ,418** 1 Independent variables 5.Criticism from important countries 4,56 12,45 ,083 ,080 -,056 ,123 1 6.Proportion of
criticism from home country 0,31 0,35 ,058 -,129* -,030 ,285** ,237** 1 7. Home country human development 0,89 0,044 -,064 -,054 -,072 ,055 ,043 -,018 1 8. Exposure to human development 0,81 0,14 -,085 -,024 -,015 ,017 ,059 -,012 ,248** 1 Control variables 9.numberofemployees 96437,65 165759,46 ,023 ,071 -,018 ,027 ,771** ,120 ,027 ,026 1 10.returnonassets 6,91 5,73 ,055 -,167* ,041 ,210** ,016 ,072 -,081 -,059 -,039 1 Industry Classification 11.Industial 0,76 0,43 ,013 -,019 ,082 -,003 -,004 ,033 ,009 -,048 -,010 ,432** 1 12.Utility 0,095 0,29 ,019 ,045 ,042 ,019 ,024 -,085 -,084 -,013 -,005 -,156* -,577** 1 13.Transportation 0,012 0,11 ,002 -,022 ,039 ,005 -,008 -,024 ,016 ,049 ,065 -,106 -,199** -,036 1 14.Banking 0,071 0,26 ,108 ,128 ,066 ,066 ,016 ,072 ,024 ,013 ,061 -,271** -,489** -,089 -,031 1 15.Insurance 0,037 0,19 -,114 -,110 -,098 ,031 -,064 -,043 ,017 ,000 -,052 -,211** -,350** -,064 -,022 -,054 1 16.Other 0,025 0,16 -,135* -,114 -,321** -,178** ,024 ,020 ,063 ,101 -,051 -,126 -,284** -,052 -,018 -,044 -,031 1 ** Correlation is significant at the .01 level. *Correlation is significant at the .05 level
Hypothesis 1 is tested by using a multiple linear regression with the CSR performance score as dependent variable, criticism from important countries as the independent variable and the number of employees, return on assets and industry
classification as control variables. Model 2, with the independent variable together with the control variables, accounts for 0,9 percent of the CSR performance, being not significant with an F score of 1,254 and a significance level of 0.269. The relationship between criticism from important countries and CSR performance was also nog significant (t = -0,956, p = 0,340)
In order to test hypothesis 2, a new multiple linear regression model is created with the CSR performance score as dependent variable, criticism from important countries and exposure to human development score as the independent variables and the number of employees, return on assets and industry classification as control variables. The interaction between criticism from important countries and exposure to human development was created from the multiplying the mean centered values of the independent variables. Model 3 explains 1,8 percent of the score on CSR performance and with an F value of 1,418 and a significance score of 0.174 is not significant. In addition the interaction effect between weighted criticism country score per criticism and exposure to HDI is also not significant (t =,703, p = 0,483), therefore we can reject H2.
The third hypothesis is tested by using a multiple linear regression with the CSR performance score as dependent variable, proportion of criticism in home country as the independent variable and the number of employees, return on assets and industry
classification as control variables. Model 4, with the independent variable together with the control variables, accounts for 0,5 percent of the CSR performance, being not significant with an F score of 1,153 and a significance level of 0,329. The relationship between
proportion of criticism and CSR performance was also not significant (t = -0,373, p = 0,710) providing no support for H3.
The last hypothesis, hypothesis 4 is tested by a multiple linear regression model, with the CSR performance score as dependent variable, criticism in home country and human development in home country as the independent variables and the value of number of employees, return on assets and industry classification as control variables. The
interaction between criticism in home country and human development index in home country is computed by multiplying both centered variables. Model 5 explains 0 percent of the score on CSR performance and is with an F value of 0,952 and a significance score of 0.487 not significant. The interaction effect between the proportion of criticism in home country and human development in home country is also not significant (t=-,158, p=,875) and H4 can be rejected. A full summary of the results can be found in table 4 on the next page.
Table 4.
Regression Results of CSR performance. Dependent variable: CSR Performance Model 1 Model 2 or hypothesis 1 Model 3 or hypothesis 2 Model 4 or hypothesis 3 Model 5 or hypothesis 4 Constant 1,746 (,257)** 1,623 (,287)** 1,582 (,288)** 1,737 (,258)** 1,761 (,263)** Independent variables Criticism from important countries
-,055 (,057) -,057 (,058)
Proportion of criticism from home country
-,025 (,068) -,029 (,069)
Exposure to HDI -,110 (,060)
Home country HDI -,018 (,116)
Criticism from important countries x exposure to HDI
,096 (,137)
Proportion of criticism from home country x Home country HDI
-,047 (,297) Control Variables Number of employees -,038 (,052) -,013 (,058) -,005 (,058) -,036 (,052) -,040 (,053) Return on Assets -,005 (,005) -,005 (,005) -,004 (,005) -,005 (,005) -,006 (,005) Industry Industrial -,008 (,053) -,008 (,053) -,008 (,048) -,008 (,053) -,010 (,050) Utility -,002 (,079) ,002 (,079) ,008 (,079) -,004 (,079) ,001 (,083) Transportation ,062 (,205) ,055 (,205) ,032 (,205) ,061 (,205) ,058 (,207) Banking -,219 (,102)* -,222 (,102)* -,241 (,103)* -,216 (,103)* -,222 (,104) Insurance ,064 (,124) ,055 (,124) ,057 (,124) ,063 (,124) ,056 (,125)* Other ,216 (,179) ,240 (,181) ,216 (,181) ,221 (,180) ,212 (,182) Model fit N 225 225 224 225 222 R2 ,040 ,044 ,062 ,041 ,043 Adj R2 ,009 ,009 ,018 ,005 ,000 F-stat 1,303 1,254 1,418 1,153 ,952 p-value 0,250 0,269 0,174 0,329 0,487
To test hypothesis 1a, a multiple regression analysis of environmental performance as a dependent variable, with criticism from important countries as independent variable and number of employees, return on assets and industry classification as control variables was performed. Model 2 accounts for 9,1 percent of environmental performance and with an F score of 3,802 and a significance of 0.000 is significant. The relationship between criticism from important countries and environmental performance is not significant (t = ,352 p =,725) so hypothesis 1a can be rejected.
Next, a multiple regression model of environmental performance as a dependent variable, with the independent variables criticism from important countries and exposure to human development together with the control variables was performed and interaction effect between criticism from important countries and exposure to human development was performed. Model 3 accounts for 10,3 percent of environmental performance and with an F score of 3,557 and a significance of 0.000 is significant. The interaction effect between criticism from important countries and exposure to human development however is not significant (t = 0,323, p =0,747), disproving hypothesis H2a. Exposure to human
development was negatively related to environmental performance (B = -0,194, t= -2,180 and p = ,030), where a lower exposure to HDI leads to a lower score on environmental performance.
Thirdly, a multiple regression model of environmental performance as a dependent variable, with the independent variable proportion of criticism from home country together with the control variables was performed. This model 4, accounts for 15,4 percent of environmental performance and with an F score of 4,901 and a significance level of 0.000 is significant. The effect of proportion of criticism in home country on environmental performance is positive, with a t value of -2,798 and a p of 0,006 is significant. Hypothesis
H3a is therefore supported, where a larger proportion of criticism from home country is related to a higher score on environmental performance.
Lastly the multiple regression model of environmental performance as a dependent variable, with the independent variables proportion of criticism from home country and human development in home country together with the control variables and interaction effect between proportion of criticism from home country and human development was performed. Model 5 accounts for 12 percent of environmental performance and with an F score of 4,000 and a significance of 0.000 is significant. The interaction effect between criticism from important countries and human development in home country is not significant (t = 0,671 p = 0,503) however, so H4a can be rejected. Full results of these analyses and models can be found in table 5 on the next page.
Table 5.
Regression Results of Environmental Performance. Dependent variable: Environmental Performance Model 1 Model 2 or hypothesis 1 Model 3 or hypothesis 2 Model 4 or hypothesis 3 Model 5 or hypothesis 4 Constant 2,315 (,383)** 2,383 (,429)** 2,296 (,430)** 2,315 (,383)** 2,378 (,385)** Independent variables
Criticism from important countries
,030 (,085) ,019 (,086)
Proportion of criticism from home country
,277 (,099)** ,276 (,101)**
Exposure to HDI -,194 (,089)*
Home country HDI -,216 (,170)
Criticism from important countries x exposure to HDI
,066 (,204)
Proportion of criticism from home country x Home country HDI ,291 (,433) Control Variables Number of employees -,329 (,078)** -,342 (,087)** -,324 (,087)** -,347 (,077)** -,339 (,078)** Return on Assets ,011 (,007) ,011 (,007) ,011 (,007) ,009 (,007) ,009 (,007) Industry Classification Industrial ,030 (,083) ,030 (,083) ,039 (,082) ,026 (,083) ,025 (,084) Utility ,071 (,118) ,069 (,118) ,078 (,118) ,092 (,116) ,112 (,121) Transportation ,522 (,305) ,526 (,306 ,482 (,305) ,540 (,300) ,523 (,303) Banking -,104 (,153) -,102 (,153) -,131 (,153) -,142 (,151) ,523 (,303) Insurance ,035 (,184) ,040 (,185) ,042 (,184) ,043 (,181) ,523 (,303) Other ,279 (,267) ,265 (,270) ,223 (,269) ,222 (,263) ,213 (,266) Model fit N 226 225 224 225 222 R2 ,123 ,123 ,143 ,154 ,159 Adj R2 ,095 ,091 ,103 ,122 ,120 F-stat 4,345 3,802 3,557 4,901 4,000 p-value 0,000 0,000 0,000 0,000 0,000
The relationship between social performance as a dependent variable and criticism from important countries as an independent variable and number of employees, return on assets and industry classification as control variables is investigated using a multiple
regression model. Model 2 accounts for 1,7 percent and is not significant with an F score of 1,481 and a significance level of 0,165. The relationship between criticism from important countries and social performance was also not significant (t = 1,224, p = 0,222) Based on these results we can reject H1b.
A multiple regression model is used to investigate the relationship between social performance as a dependent variable and criticism from important countries and exposure to human development as independent variables and number of employees, return on assets and industry classification and the interaction effect between both independent variables. Model 3 accounts for 2,1 percent and is not significant with an F score of 1,488 and a significance level of 0,145. The interaction effect was also not significant (t = -0,565, p = 0,573). Based on these results we can reject H2b.
A third multiple regression model is created to investigate the relationship between social performance and proportion of criticism from home country. Proportion of criticism from home country, along with the control variables number of employees, return on assets and industry classification accounts for 1,8 percent of the score on social performance and is with a F score of 1,506 and a significance of 0,156 not significant, as is proportion of criticism from home country (t = 1,300, p = 0,195). Hypothesis 3b can be rejected based on these results.
The interaction effect between proportion of criticism from home country and home country development is also analyzed using a multiple regression analysis with social performance as dependent variable, proportion of criticism from home country as
independent variables and number of employees, return on assets and industry
classification as control variables. This model is not significant (F = 1,696, p = 0,083). The interaction effect between proportion of criticism from home country and human
development in home country is also not significant (t = 0,179 p = 0,858), disproving H4b. The effect of human development in home country on social performance however is significant (B = -0,264 t = -2,136 p = 0,034). This is a negative significant effect, where a lower score on human development index leads to a higher score on social performance. Full results of these analyses and models can be found in table 6 on the next page.
Table 6.
Regression Results of Social Performance Dependent variable: Social Performance Model 1 Model 2 or hypothesis 1 Model 3 or hypothesis 2 Model 4 or hypothesis 3 Model 5 or hypothesis 4 Constant 1,433(,275)** 1,604 (,308)** 1,570 (,310)** 1,474 (,276)** 1,454 (,278) Independent variables Criticism from important countries
,077 (,063) ,067 (063)
Proportion of criticism from home country
,097 (,075) ,094 (,075)
Exposure to HDI -,109 (065)
Home country HDI -,268 (,125)*
Criticism from important countries x exposure to HDI
-,084 (,148)
Proportion of criticism from home country x Home country HDI
,058 (,321) Control Variables Number of employees -,052 (,056) -,087 (,063) -,080 (,063) -,060 (,056) -,054 (,056) Return on Assets -,002 (,005) -,002 (,005) -,002 (,005) -,002 (,005) -,003 (,005) Industry Classification Industrial -,021 (,059) -,021 (,059) -,013 (,059) -,023 (,059) -,021 (,059) Utility -,031 (,087) -,037 (,087) -,036 (,087) -,025 (,087 -,010 (,089) Transportation ,020 (,225) ,030 (,224) ,002 (,225) ,026 (,224) ,012 (,224) Banking -,126 (,112) -,121 (,112) -,135 (,113) -,139 (,113) -,160 (,113) Insurance ,068 (,136) ,080 (,136) ,077 (,136) ,070 (,135) ,055 (,135) Other ,437 (,177)* ,400 (,179)* ,367 (,180)* ,420 (,177)* ,396 (,177)* Model fit N 226 226 225 226 223 R2 ,045 ,052 ,065 ,053 ,074 Adj R2 ,015 ,017 ,021 ,018 ,030 F-stat 1,475 1,481 1,488 1,506 1,696 p-value ,177 ,165 ,145 ,156 ,083
* p < .05, ** p < .01 (standard deviations in parentheses)
The last dependent variable which will be analyzed is corporate governance. To test the first hypothesis, a multivariate regression analysis is done with corporate governance as dependent variable, criticism from important countries as independent variable and number of employees, return on assets and industry classification as control variables. Model 2 is
significant (F = 4,668, p = 0,000) and explains 12 percent of the variable corporate governance. Criticism from important countries has a significant negative effect on corporate governance (B = -0,223, t = 3,860 p = 0,000). A higher level of criticism from important countries has a lower score on corporate governance. This is the opposite effect which we expected, disproving H1c.
The next analysis which is done is the multiple regression analysis with corporate governance as dependent variable, criticism from important countries and exposure to human development as independent variables and number of employees, return on assets and industry classification as control variables. Model 3 tests for interaction between criticism from important countries and exposure to human development. The model as a whole is significant (F = 4,070 p = 0,000) and explains 12 percent of corporate governance. The interaction effect however is not significant (t = 0,915 p = 0,361), providing no support for H2c.
The third analysis is a multiple regression analysis with corporate governance as dependent variable, proportion of criticism from home country as independent variable and number of employees, return on assets and industry classification as control variables. Model 4 as a whole is significant (F = 4,341 p = 0,000) explaining 10,6 percent of the variable corporate governance. Proportion of criticism from home country has a significant negative relation to corporate governance (B = -0,243, t = -3,538 p = 0,000), disproving H3c, where a positive relationship was expected.
The fourth and last multiple regression analysis looks at the interaction effect of proportion of criticism from home country and human development in the home country. Model 5 has corporate governance as dependent variable, proportion of criticism from home country and human development in the home country as independent variables. The control variables in this analysis are number of employees, return on assets and industry
classification. Model 5 explains 12,2 percent of the variable corporate governance and is significant (F = 4,096 p = 0,000). The interaction effect however is not significant (t = -0,566, p = 0,572), providing no support for H4c. A full summary of the results of all models in regards to corporate governance can be found in table 7 on the next page.
Table 7.
Regression Results of Corporate Governance Dependent variable: Corporate Governance Model 1 Model 2 or hypothesis 1 Model 3 or hypothesis 2 Model 4 or hypothesis 3 Model 5 or hypothesis 4 Constant ,782 (,260)** ,283 (,283) ,322 (,285) ,679 (,255)** ,735 (,256)** Independent variables Criticism from important countries
-,223 (,058)** -,211 (,058)**
Proportion of criticism from home country
-,243 (,069)** -,249 (,069)**
Exposure to HDI ,100 (,059)
Home country HDI ,278 (,116)*
Criticism from important countries x exposure to HDI
,125 (,136)
Proportion of criticism from home country x Home country HDI
-,167 (,295) Control Variables Number of employees ,112 (,053)* ,213 (,058)** ,205 (,058)** ,131 (,052)* ,119 (,052)* Return on Assets -,017 (,005)** -,014 (,005)** -,015 (,005)** -,015 (,005)** -,015 (,005)* Industry Classification Industrial -,007 (,057) -,008 (,057) -,012 (,057) -,002 (,056) -,004 (,056) Utility -,077 (,082) -,060 (,080) -,060 (,080) -,094 (,080) -,109 (,082) Transportation -,206 (,213) -,236 (,206) -,210 (,207) -,222 (,207) -,211 (,207) Banking -,173 (,106) -,186 (,103) -,175 (,103) -,139 (,104) -,124 (,104) Insurance -,136 (,128) -,172 (,125) -,169 (,125) -,142 (,125) -,137 (,125) Other ,057 (,167) ,165 (,164 ,196 (,165) ,099 (,163) ,113 (,163) Model fit N 227 227 226 227 224 R2 ,088 ,146 ,159 ,137 ,161 Adj R2 ,059 ,115 ,120 ,106 ,122 F-stat 3,014 4,668 4,070 4,341 4,096 p-value 0,005 0,000 0,000 0,000 0,000
5. DISCUSSION
This research tried to analyze the relationship between the origin of criticism at the firm and the CSR performance of the firm. Most of the results found disprove the
hypotheses posed in this research. The results show that the importance of the country of origin of the criticism at the firm is not related to CSR performance. Criticism from important countries does not lead to a higher CSR performance than criticism from less important countries. This is also true for environmental performance, social performance and corporate governance which make up CSR performance. The exposure to the rate of human development has no moderating effect on the country of origin of the criticism at the firm and CSR performance, environmental performance, social performance and corporate governance. Thirdly, the proportion of criticism from the home country of the firm is not related to the CSR performance, social performance and corporate governance in MNE’s. Lastly, the level of human development in the home country also has no moderating effect on the relationship between the proportion of criticism from the home country of the firm and CSR performance, environmental performance, social performance and corporate governance. When all the results are taken together, this provided little evidence for the most of the hypotheses.
There were two other interesting findings however. First environmental performance was positively related to exposure to criticism from important countries, where criticism from more important countries also means a higher level of environmental performance. Secondly a negative effect was found between criticism from important countries and corporate governance, as well as negative effect between proportion of criticism from home country and corporate governance.
