Corporate social responsibility and financial performance: a contingency based approach

Corporate social responsibility and

financial performance: a contingency

based approach

Tim Jacob Herman Pieter Reimerink

Student number: 1654160

UNIVERSITY OF GRONINGEN

Faculty of Economics and Business

Master Thesis - International Business and Management

July 2012

Supervisor: Dr. A. van Hoorn

2

Abstract

3

Index

2.4THE BUSINESS CASE FOR CSR, WHAT DO WE KNOW? ... 14

3. THEORY AND HYPOTHESES ... 18

3.1THE CSP-CFP RELATIONSHIP ... 18

3.2THE CSP-CFP RELATIONSHIP AND FIRM SIZE ... 22

3.3THE CSP-CFP RELATIONSHIP ACROSS INDUSTRIES ... 23

4. DATA & METHOD ... 27

4.1DATA ... 27 4.1.1 Dependent variable ... 28 4.1.2 Key independents ... 28 4.1.3 Control variables ... 31 4.2EMPIRICAL MODEL ... 32 5. EMPIRICAL RESULTS ... 34 5.1ROBUSTNESS CHECKS ... 37

6. DISCUSSION & CONCLUSIONS ... 39

6.1LIMITATIONS AND DIRECTIONS FOR FUTURE RESEARCH ... 41

REFERENCES ... 43

APPENDIX ... 47

A1.DATASTREAM DESCRIPTION OF VARIABLES ... 47

A1.1 Total Assets ... 47

A1.2 R&D Expenses ... 47

4

1. Introduction

Over the past half century, the concept of corporate social responsibility (CSR) has evolved and developed rapidly in both the academic and professional world. It has been, and still is, a subject of continuous debate and research among management scholars and an inescapable matter for businesses around the world. Especially larger firms are very much concerned with the moral obligations to their employees, customers, society and the environment. “As many as 90 % of the Fortune 500 companies now have explicit CSR initiatives” (Luo & Bhattacharya, 2006, p.1). These initiatives come in all different forms, from donations to charity to fair trade programs to environmental friendly productions techniques. A research by Berner (2005) which was published in Business Week, reported investments by firms in CSR initiatives varying from $51.3 million dollars up to $926 million which constituted in between 2.7 and 43.3 percent of pre-tax profits of those firms. Do firms spend such amounts on CSR activities purely out of charity and good intentions? Or are there other interests at stake?

Four decades ago, a new stream within the CSR literature emerged, by some referred to as “the business case for CSR” (Carroll & Shabana, 2010). This particular stream is only concerned with the effect of CSR on the financial performance of firms and deals with questions like: what are the benefits of engaging in CSR activities and how does this pay off? Is there a return on investment to CSR? Is

corporate social performance (CSP) positively related to corporate financial performance (CFP)? Although numerous researchers have tried to answer these questions by exploring the CSP – CFP relationship, no definite consensus exists (Griffin & Mahon, 1997). There are studies that found positive relationships, studies that found negative relationships and studies that found no relationship at all. However, a meta-analysis of Orlitzy et al. (2003) pointed out that the literature on the CSP – CFP issue is more coherent and meaningful than researchers generally think. They proved that there is a statistically significant positive relationship between CSP and CFP across a wide variety of industries and study contexts. There are two possible explanations, which are not mutually exclusive, for the studies that demonstrated negative or no relationship at all; 1. methodological errors/differences (Griffin & Mahon, 1997), 2. the omission of mediating and moderating variables (Carroll & Shabana, 2010; Orlitzky et al., 2003; Rowley & Berman, 2000).

Several follow up studies that did include mediating and moderating factors into their conceptual model (Luo & Bhattacharya, 2006; Barnett, 2007), found evidence that the relationship between CSP and financial performance is indeed mediated by different factors and is dependent upon situational

5

which influence the CSP – CFP relationship and that the exclusion of these factors in previous studies explain a substantial amount of the inconsistencies found.

Although researchers have found and examined some of these moderating and mediating factors, this area of research is still in its founding stages and requires further research. There are several more factors to explore which possibly play a role in the complex relationship between CSR and financial

performance. Rowley & Berman (2000) state that researchers have often overlooked many contingencies which may cause variability in returns to CSR. In line with this, Barnett (2007, p. 813) claims that “CSR cannot financially please all of the corporations all of the time, but it can please some of the corporations some of the time, researchers should try to figure out which ones and when”. The question of which corporations and when is of course a complex one, there are probably many factors that determine whether improved CSP leads to improved CFP. The relevant question should therefore not be “does CSR pay?” but “when and for whom does CSR pay?”. In other words, under what circumstances are firms able to financially benefit from CSR and under what circumstances are they not?

The present thesis takes a contingency based approach towards the CSP – CFP nexus by focusing on three contextual factors that are most likely to moderate the effect of CSP on CFP; firm size, a firms’ proximity to the end consumer and a firms’ potential for damage to the environment and/or society. The idea behind these first two factors can be best summarized with a single quote: “consumers first need to become aware of a firms’ level of social responsibility before this factor can impact their purchasing” (Mohr et al., 2001, p. 47). For consumers to become aware of a firms’ level of CSR, a firm should be visible and it should have a certain proximity to the end consumer. The arguments underlying the third contingency are based on the idea that firms with a high damage potential generally suffer from bad reputations and have to invest higher amounts of money to achieve a certain degree of social performance. A firms´ visibility is represented by firm size and the latter two contingencies are represented by the industry in which a firm operates.

In the present study, I argue that larger firms are better able to benefit from CSR because their social performance is better visible to the outside world and thus more likely to be rewarded. Furthermore, I propose that the positive CSP – CFP link is more pronounced for firms with a high proximity to the end consumer. Or stated differently, companies operating on a business to consumer model (B2C) can expect a higher return on CSR investments than companies operating on a business to business (B2B) model. Finally, I expect a lower return on social investment for companies with a high potential for damage to the environment and/or society.

6

play a moderating role in the CSP – CFP relationship, this study can reduce a clear gap in the existent literature. Besides its contribution to the literature, the findings can have valuable implications to practice as well. If we know under what circumstances the engagement in socially responsible activities leads to improved financial performance, one can decide when CSR is a profitable investment and when not. Managers will be able to shape and adjust their CSR investments in such a way that both society and the firm itself benefit. Thus, by either proving or disproving the expected amplifying impact of the

aforementioned contingencies on relationship between CSP and CFP, we come one step closer to answering the complex question; ‘when can who do well by doing good’?

In one sentence this study sets out to discover:

Which factors moderate the relationship between corporate social performance and corporate financial performance?

In order give a well funded answer to the main research question, the following sub questions should be raised and answered:

Is there a positive relationship between CSP and CFP? Is the relationship CSP and CFP moderated by firm size?

Is the relationship between CSP and CFP dependent upon a firms´ proximity to the end consumer? Is the relationship between CSP and CFP moderated by a firms´ potential for damage to the environment and/or society?

1.1 Thesis outline

7

8

2. The background of CSR

2.1 What is CSR?

Before turning to the history and development of CSR I will attempt to provide a clear definition of the concept first, which is not as easy as one would think. The scientific literature runs many definitions of the concept (Barnett, 2007) and even the term ‘CSR’ itself is not consistently used by researchers. Although the term CSR is still in popular use, substituting terms as corporate citizen, sustainability, business ethics, corporate social performance, stakeholder management, internalization of externalities and corporate philanthropy are used as well (Carroll & Shabana, 2010). From this wide variety of terms, corporate social performance (CSP) is frequently used and has become well established in the CSR literature.

CSR is a broad concept, it is therefore not surprising that researchers provided us with numerous different definitions over the years (Mohr et al., 2001). A very broad but often used definition of CSR is: the corporate obligations that extent beyond the economic and legal responsibilities of the firm (Carroll & Shabana, 2010). In other words, CSR means doing more than what the legal system prescribes you to do. This is of course very broad and leaves room for multiple interpretations. It is therefore not an explicit definition of CSR but rather a fundament for more specific definitions. The idea behind this is that the legal rules and regulations set by our governments are not able to constrain everything that society perceives as being unfair or unacceptable. For example, society is generally against child labor and perceives this as unacceptable. At the same time, there are many countries in the world where child labor is still not prohibited by law. One of the reasons for this discrepancy between laws and the ethical standards of society is that societal processes are dynamic and that legislative bodies do not always work efficiently (Dam, 2008). Perceptions of fair and unfair change and it can sometimes take over 20 years for laws to change accordingly. Since societal perceptions will be continuously changing and legislative bodies need time to catch up with the pace, the setting of optimal laws is practically impossible. Therefore, conforming to society’s ethical standards is not the same as complying with rules and regulations. Voluntary ‘overcompliance’ in terms of environmental and social policies is what then may be called corporate social responsibility.

9

generally provided more abstract definitions as “doing business in a way that maintains and improves both the customer’s and society’s wellbeing” (Kotler, 1991).

From all the definitions Carroll’s (1979) received most attention. This four-part definition of CSR states that “the social responsibility of businesses encompasses the economical, legal, ethical and discretionary (later on philanthropic) expectations that society has of organizations at a given point in time” (Carroll, 1979, p. 500). Twelve years later, Carroll (1991) presented his pyramid model of CSR (figure 1) which is in line with his definition as introduced in 1979.

Figure 1: Carroll’s pyramid of CSR (1991)

Corporate social responsibility as depicted in figure 1 consists of 4 components that are ranked according to importance from the bottom to the top of the pyramid; economic responsibilities, legal responsibilities, ethical responsibilities and philanthropic responsibilities. The economic responsibility of a company is to create wealth by selling goods and services against a profit. The motive of making profits is historically the main incentive of entrepreneurship, later on this motive changed into profit

10

ethical responsibilities relate to those practices that are expected or prohibited by society but are not codified by law. As explained earlier, these practices are not yet codified by law because of the

inefficiency of legislative bodies (bureaucracy) but they can definitely be a driving force for future rules and regulations. A firms’ Philanthropic responsibilities encompass corporate actions as a response to society’s expectation that firms should be good corporate citizens. In order to be this good corporate citizen, businesses should engage in activities which support and promote human welfare. For example, donations to charity, hiring weaker or handicapped people or contributions to education. Philanthropic responsibilities differ from ethical responsibilities in that it is not per se unethical to refuse philanthropic behavior where it would definitely be unacceptable to ignore ethical responsibilities. The concept of Carroll’s CSR pyramid has been empirically tested in multiple studies (Aupperle, Carroll & Hatfield, 1985; Pinkston & Carroll, 1994; Edmondson & Carroll, 1999; Burton, Farh & Hegarthy, 2000) and was largely supported by the findings.

The CSR definitions outlined above are only a minor part of the total number of definitions provided by the CSR literature. Although they differ in length, form and their degree of abstraction and

completeness, the bottom-line is the same; CSR refers to voluntarily taking responsibility for causes that extent beyond the direct economic interests of the firm. Corporate social performance differs from these definitions in that, CSR mainly refers to assuming and accepting business’s responsibilities to society and CSP emphasizes the actual performance and outcomes of social initiatives (Carroll, 1979; Wood, 1991). CSP must therefore be regarded as a snapshot of CSR (Barnett, 2007). When studying the effect of CSR on firm financial performance, CSP is a frequently used measure (Margolis & Walsh, 2003) and the most suitable in my opinion. In the context of this study I define CSP as “a firms’ voluntary engagement into activities that benefit the environment and/or society”.

2.2 A history of CSR

The first foundations of CSR are generally sought in the beginning of the post second world war period. It was indeed during this period that the social responsibilities of businesses were increasingly acknowledged by scholars and business men (Spector, 2008). However, the real roots of CSR can be traced back to far before the Second World War. Even before the great Adam Smith published his book “The wealth of nations” (1779), he wrote one of his more infamous works “The theory of moral

11

sympathy and ethics in an economic context. In here, he emphasized the necessity of maintaining ethical behavior in order to achieve long-term growth and he proposed the idea that “honesty is profitable”. The concept of CSR as we know it today, is based on thoughts like these. During the industrial revolution, Adam Smith’s ethical ideas were put to practice by industrialist and social reformer Robert Owen. Owen was concerned about the precarious health and working conditions of factory workers. He improved these conditions by reducing working days, setting a minimum age for workers and providing better housing for his employees (Carrasco, 2007). Although the term “social responsibility” was not yet used during these times and there were no attempts at all to theorize the concept, the notion that businesses should be concerned with the wellbeing of their employees and the society at large started here.

It took more than 150 years before academics and business executives picked up the social responsibility trend as initiated by Adam Smith and Robert Owen. It was the dean of the Harvard

Business School, Donald K. David , who spoke to the incoming MBA class in 1946 about the new role of business men in the post-war period. During the decades before the war there were certain events, under which the collapse of capitalism during the great depression, that gave rise to great suspicion of the motives and intentions of corporations. The reputation of big businesses and the men who directed them worsened; the idea that a few rich men and large corporations wielded an unhealthy amount of power, was widely shared among the American public. Into this mixture of corporate dominance and increasing suspicion, “Donald K. David urged businessmen to abandon their ivory tower and engage the larger world by acknowledging the public responsibilities of enterprise” (Spector, 2008, p 317/318). By doing this, companies could improve their reputations and they would be able to play a new role in the post-war world. At the same time, the broader responsibilities of business men and stable financial performance could serve as a roadblock against the danger of Soviet Communism (Spector, 2008). It could serve as proof for the soundness of capitalism and business leaders would be the agents of worldwide benefit.

In the following years, Donald K. David and other CSR advocates used the widely respected Harvard Business Review as a stage for promoting the new role of companies as corporate citizen. As a result, CSR started to grow in popularity. An increasing number of business men and scholars started to emphasize that companies had broader responsibilities than maximizing profits only. They argued that firms and its management had to think of their customers, employees and the public at large as well (Spector, 2008).

In the 1960’s the social environment changed in that society started to demand and expect higher ethical standards from corporations, especially civil right, woman’s rights, consumer rights and

12

(1958) stated that social concerns and the general welfare were the responsibilities of the government, not of business. He feared that attention to social causes would go at the cost of the profit motive, which is so essential for business success. As with all topics in science, skepticism remained but the majority of both the academic and professional world kept supporting the CSR trend.

In the decades after the 60’s, CSR kept on growing in popularity and lots of new streams within the literature emerged. CSR evolved primarily through academic contributions but during the 1970’s the business community followed. Businesses became progressively more concerned with topics like racial discrimination and pollution and there was a greater focus on charitable donations (Murphy, 1978) In the years that followed until today, both the academic and professional world continued developing and extending the concept of CSR, researchers became occupied with seeking all kinds of benefits of CSR (the business case for CSR) and a growing number of companies started to realize that socially

responsible was the new standard. Especially in the last 12 years the request for ethical behavior on the side of corporations accelerated. This was, among other things, caused by the notorious Enron accounting scandal (Carroll & Shabana, 2010) due to which 4500 people lost their jobs, investors lost about 60 billion dollars, thousands of pensions were obliterated and trust in the American economy was destroyed. At the same time, the media focused a greater deal of attention on the deteriorating environmental conditions of the world we live in which caused a strong environmental awareness trend. Especially the larger corporations were blamed for their negative contributions to the environment as a result of pollution, excessive CO2 emissions and forest clearing. As a response, society pressured these corporations to engage in more environmental friendly practices. The business community became fascinated with sustainability and environmental awareness which are an indispensible part of CSR nowadays.

Today, the majority of large multinationals (Kotler & Lee, 2004) and a growing number of SME’s (Grant Thornton LLP, 2009) are engaged in CSR initiatives and practices. Larger corporations even publish a yearly CSR report in which their socially responsible policies and activities are outlined and communicated to the outside world. CSR has become an inescapable priority for business leaders in almost every country (Porter & Kramer, 2006). It has become the new standard in business and must be considered as permanent rather than as a trend or hype. This is reflected in the increasing number of sustainability, business ethics and other CSR related courses given in MBA programs (Christensen et al., 2007).

2.3 Examples of CSR initiatives

13

over the world and come in many forms. Corporations have been very creative in coming up with all kinds of CSR initiatives. Especially companies operating in controversial industries, in which there is a high potential for damage to the environment and society, are likely to engage in CSR activities to compensate for the damage and negative publicity. A firms’ reputation is one of its most valuable assets and CSR can serve as an insurance against reputation risk (Minor & Morgan, 2011).

Many firms in the footwear and clothing industry for example, drew negative attention when it appeared that their clothing was manufactured by children working under terrible circumstances in the so called sweatshops. As a response, Nike has been very active during the last decade in all kinds of projects in third world countries. They support in developing several villages in Thailand; they assigned task forces to improve and control the working conditions in the factories of their contract manufacturers and they established the Nike Foundation to fight poverty in developing countries. Nike claims to have invested 100 million over the past 6 years in the Nike foundation only. Similarly, the oil & gas industry does not have the best reputation when it comes to environmental issues such as pollution and CO2 emissions. Therefore, oil & gas companies such as Shell and BP try to do everything they can to

compensate for the environmental damages made. Recently, Shell announced to have adopted initiatives for mitigating the negative effects of climate change and promoting learning and education in the Middle East. They have explored techniques to capture and utilize carbon dioxide in order to reduce the emission of greenhouse gasses. In line with this, they cooperate with local universities and communities to keep on researching new and environmental friendly technologies. In this way they contribute to both local knowledge and a better environment. British Petroleum (BP) is very much concerned with the conservation of fragile areas and the protection of endangered species. They founded the Leadership Conservation Program which funds young conservationists to protect and discover endangered species all over the world. To date, the program has resulted in the discovery of 123 new species. After the oil spill in the Gulf of Mexico, BP´s reputation slipped away in the midst of this tragedy, leading to a reduction in its market value of around 100 billion dollar (Associated press, 25th of June 2010). Since this day BP is trying to rebuild its reputation by heavily investing in CSR initiatives.

14

In addition, there are numerous small and medium sized enterprises involved in smaller and less costly forms of CSR. The German medium sized cleaning company “LR Gebäudereinigung”, among others, offers their employees with migration background free German classes in order to improve their language skills. Similarly, a small Austrian consultancy company named “Denkstatt GmbH” has launched an internal project promoting the health of its employees. The company provides its personnel with biological fruits and milk free of charge and a couple of the employees cook about three times a week a healthy meal for all during work time. Furthermore, the company encourages physical exercise by offering a sport program.

Although these examples constitute only a minor part of the wide variety of CSR initiatives out there, it gives a clear picture of the scale and scope in which CSR is practiced. It is evident that CSR is not limited to certain industries and countries or big corporations only. Furthermore, the examples reflect all the different forms companies can use to be good corporate citizens. Some CSR initiatives focus on improving the health and wellbeing of employees, some are aimed at conserving the environment and others contribute to society as a whole. Conclusively, CSR is a broad concept, both in its meaning and application.

2.4 The business case for CSR, what do we know?

As briefly touched upon in the introduction, the business case for CSR is concerned with the specific benefits to businesses in an economic and financial sense that flow from CSR initiatives (Carroll & Shabana, 2010). As it is firms’ their primary objective to make a profit, the business case for CSR refers to the business justification and rationale for engaging in CSR activities. For CSR to be ‘justified’ from a business perspective, there should be consistent evidence that engaging in socially responsible activities leads to a higher profit. Forty years of incomparable research to the CSP – CFP relationship provided us with inconsistent empirical evidence.

Margolis & Walsh (2003) reviewed 127 studies, published between 1972 and 2003, that all empirically examined the relationship between corporate social performance and corporate financial performance. Out of these studies, CSP has been treated as independent variable predicting financial performance, in 109 of the cases. Almost half (54) of the studies showed a positive association between CSP and CFP; only 7 studies reported a negative relationship; 28 studies found non-significant

15

Griffin & Mahon (1997) reviewed 51 studies that examined the CSP - CFP nexus and investigated whether methodological inconsistencies are the main cause of the conflicting findings. Their sample of 51 studies consists of 21 studies that found an evident positive relationship; 11 studies reported a clear negative association; 6 studies showed non-significant relationships and the remaining 13 studies reported both positive and negative results. They conclude that the variability in these findings can be attributed to the different measures used for CSP and CFP.

Orlitzky et al. (2003) were the first to conduct a systematic review of all the present CSP-CFP studies. By performing a meta-analysis integrating 30 years of research and 52 different studies (including the above mentioned study of Griffin & Mahon, 1997), they found that the majority of the empirical evidence points to a positive correlation between CSP and CFP. “The meta-analytic findings suggest that corporate virtue in the form of social responsibility and, to a lesser extent, environmental responsibility is likely to pay off, although the operationalizations of CSP and CFP also moderate the positive association” (Orlitzky et al., 2003, p 403). The authors discuss the possibility of an availability bias, which means that studies finding negative or non-significant relations are unlikely to be published. However, a file drawer analysis revealed that 1000 such studies would be needed to invalidate the results (Orlitzky et al., 2003). This emphasizes the robustness of their results.

Although prior results regarding the CSP-CFP relationship are fractured, it is clear that the majority of the evidence is in favor of a positive association between CSP and CFP. However, prior studies suggest the direction of causality to go two ways, from CSP to CFP (57out of 109 studies) and from CFP to CSP (16 out of 22 studies). In other words, does social performance lead to financial performance or is it exactly the other way around? Waddock & Graves (1997) developed and tested two different theories, one in support of CFP as an outcome variable and one in support of CSP as an outcome variable.

Slack resources theory argues that businesses with strong financial performance, have resource available (slack resources) which are invested in CSR initiatives. If these slack resources are available and management decides to allocate them into the social domain, financial performance will positively affect social performance and not the other way around. Alternatively, good management theory states that strong financial performance is the consequence of good management practices. This because attention to CSR will lead to the creation and maintenance of good relationships with key stakeholder groups. As will be discussed later, strong stakeholder relationships result in superior (financial) performance.

16

in an initial attention to the social performance dimensions, there may be a simultaneous and interactive impact, possibly forming what we earlier termed a virtuous circle” (Waddock & Graves, 1997, p. 315). This implies that neither good management theory nor slack resources theory seems to be incorrect. Regardless of where the circle begins, socially responsible firms benefit from stronger financial performance which results in the availability of slack resources. These resources are allocated to new CSR investments which again results in stronger social, and thereafter, financial performance etc. etc.

In conclusion, the causality issue in the CSP-CFP relationship remains to be resolved. Despite theories and empirical evidence in support of CFP as a predictor of CSP, I firmly believe CSP to predict CFP because of two reasons. First, stakeholder theory together with the four modes of value creation from Kurucz et al. (2008), which will be discussed in chapter 3, provide managers with several incentives to engage in CSR initiatives. Second, to my knowledge, the only theory in support of CFP as predictor of CSP is slack resources theory. Slack resources theory is based on an, in my opinion, unlikely assumption. It assumes that managers will naturally allocate the ‘overabundance’ of resources, or at least parts of it, to CSR initiatives. There is no guarantee whatsoever that this will actually happen.

The fact that most of the empirical studies to date support a positive impact of CSP on CFP, does not mean that the business case for CSR is now established (Barnett, 2007). It appears that the strength of the CSP-CFP relation is dependent upon methodological issues (Orlitzky et al., 2003; Margolis & Walsh, 1997) and many researchers have studied the relationship under wrong assumptions. Most scholars have assumed that CSP has a direct effect on CFP and that the returns to CSR are homogeneous across firms and industries, neither of which seems to be correct. Understanding of the complex relationship between social responsibility and performance could be improved by identifying mediating variables and

situational contingencies (Carroll & Shabana, 2010).

Luo & Bhattacharya (2006) and Pivato et al. (2008) found that CSP positively influences a firms’ market value through intermediate performance measures such as customer satisfaction, brand loyalty and consumer trust. Others (Lai et al., 2010; Orlitzky et al., 2003) showed that a firms’ degree of social investments positively impacts its reputation, which again leads to improved financial performance. These results imply that CSP is indirectly rather than directly related to CFP and that a simplistic view towards the relationship does not suffice.

Others argue that the effect of CSR on financial performance varies from one firm to another. Such variation can be attributed to contextual factors which are specific for each situation. Some firms will therefore be rewarded for their corporate social performance while others will not. According to Barnett (2007) returns to CSR are dependent upon a firms’ stakeholder influence capacity, a self-created construct which measures the ability of a firm to identify, act on and profit from opportunities to improve

17

from CSR through enhanced stakeholder relationships (Luo & Bhattacharya, 2006; Pivato et al., 2008) and that they will only capture the full benefits when they have the capacity to positively influence their stakeholders. In addition, Hoepner et al. (2010) found that the CSP - CFP relationship is heterogeneous across industries, in other words, the relationship is moderated by several industry characteristics. Also Luo & Bhattacharya (2006) found empirical evidence in favor of a contingency perspective, they showed that social investments only pay for firms with a certain standard in product innovativeness and product quality. The relationship between CSP and CFP is thus dependent upon corporate abilities.

The contingencies mentioned above, very well explain why the effects of CSR on financial

18

3. Theory and hypotheses

Besides a considerable amount of empirical evidence in favor of a positive association between CSP and CFP, the business case for CSR finds great support from several scientific theories as well. Although classic economic theory suggests a negative, or at least neutral relationship between corporate social responsibility and financial performance, more recent theories predict the opposite. The idea of firm size, proximity to the end consumer and damage potential as situational contingencies affecting the CSP-CFP nexus, is also embedded in theory.

3.1 The CSP-CFP relationship

Considering the fact that CSR initiatives always come at a cost and have no direct and tangible return, classic economic theory suggests that CSR investments go at the expense of financial

performance. This because there is a cash outflow which does not yield a direct return. Even when the costs of CSR initiatives would be fully borne by an external stakeholder, socially responsible companies would theoretically seen not outperform their ‘irresponsible’ competitors (Dam, 2008). Economic theory thus advocates a trade-off between CSR and financial performance because it neglects the possible returns of CSR investments.

The most fundamental and probably most cited theory in the context of the business case for CSR, is stakeholder theory (Freeman, 1984). In the traditional view of the firm, shareholders are the owners of the company and the highest priority stakeholders. Shareholders provide money which is regarded as input in the firm. With help of suppliers and employees, this input is converted into usable output which is then bought by the customer. According to this view, a firm only has four relevant stakeholders;

investors/shareholder, suppliers, employees and customers. Stakeholder theory argues that there are many more relevant stakeholders involved, including trade unions, labor unions, governmental bodies,

19

later. However, solely paying attention to the interest of financiers will lead to dissatisfaction of other stakeholder groups. Hence, firms should focus on all relevant stakeholders and not only on financiers and CSR can be a strong tool to do so. In order to work in the interest of communities, employees,

environmental groups, human right activists, customers and governments, firms have to behave and operate in a socially responsible manner. By doing this, firms prevent themselves from getting into trouble (consumer boycotts, lawsuits etc.) and they could benefit from rewards such as improved brand reputation and customer loyalty. The positive effect of CSP on CFP is thus an apparent product of

stakeholder theory in the sense that managers should engage in CSR practices to create and maintain long lasting relationships with all of their relevant stakeholders from which the firm will eventually benefit.

Another theory providing support for a positive association between CSP and CFP is the resource based view of the firm. The resource based view argues that the basis of competitive advantage lies within the (application of the) valuable resources a firm possesses (Wernerfelt, 1984). If these valuable resources are rare, imperfectly imitable and not substitutable without great effort, a firms’ bundle of resources can assist them in achieving long term above average performance (Barney, 1991). These resources can come in all forms but several scholar have explicitly recognized intangible assets such as corporate culture (Barney, 1986), know-how (Teece, 1980) and corporate reputation (Hall, 1992) as main sources of competitive advantage. In this context, CSR initiatives must be regarded as valuable intangible resources leading to competitive advantages such as a strong corporate culture or a good corporate reputation. Russo & Fouts (1997) examined the relationship between environmental performance and profitability from a resource based view perspective and found that it “pays to be green”. They relate this finding to the resource based view by stating that strong environmental performance constitutes a source of competitive advantage which results in superior profitability.

These theories form the basic foundation for the business case for CSR. More concrete theories describing the mechanisms through which CSP positively affects CFP are derived from both stakeholder theory and the resource based view. To my knowledge there is one theory which schematically outlines these mechanisms. Kurucz et al. (2008) describe four modes of value creation through corporate social performance: cost and risk reduction, gaining competitive advantage, developing reputation and legitimacy and synergetic value creation.

1. Cost and risk reduction

20

equal opportunity policies for example, reduce costs and risks as a result of improved morale and low absenteeism (Smith, 2005). It appears that a lack of diversity may cause higher turnover and absenteeism rates among employees (Berman et al., 1999). Similarly, cost savings can be

achieved through strong environmental performance. Environmental rules and regulations change from time to time, firms that are proactive on environmental issues are one step ahead of future regulations and save the costs of complying with new rules. At the same time, environmentally responsible companies reduce the risk and associated costs of being sued because of an

environmental scandal or inadequate standards. Furthermore, CSR activities directed at social causes such as the community, may contribute to cost and risk reduction (Berman et al., 1999). As a good corporate citizen, governments and institutions can decrease the amount of regulation and taxes imposed on the firm. These examples are in line with stakeholder theory which states that maintaining good stakeholder relationships is vital to a firms’ success. A survey of

PricewaterhouseCoopers revealed that 73% of the business executives perceive cost and risk reduction as one of the top three incentives to become more socially responsible (Carroll & Shabana, 2010).

2. Gaining competitive advantage

According to Kurucz et al. (2008) firms create a competitive advantage by engaging in CSR activities. In other words, firms can distinguish themselves from competitors by being socially responsible. There is strong empirical evidence that CSR practices improve brand loyalty (Pivato et al., 2008) and corporate reputation (Lai et al., 2010), two intangible assets which may

definitely be regarded as a competitive advantage. Similarly, companies that have a strong focus on the wellbeing of their employees can benefit from an extraordinary motivated workforce which acts in the interest of the company. Additionally, these improved employee relations can result in a strong and positive corporate culture and a significant reduction in conflicts.

Furthermore, it appears that socially responsible firms are not only appreciated by employees and customers, but by investors as well (Smith, 2005). Especially institutional investors prefer to avoid companies that violate the ethical standards of society. In this sense, CSR contributes to attracting capital and investment as well.

Conclusively, by satisfying the demands of multiple stakeholders, firms create a competitive advantage over its competitors. Stakeholders who consider CSR as important will prefer

21

creation is funded within the resource based view of the firm (Barney, 1991). Firms can build competitive advantages by exploiting their ethical resources and capabilities, according to the resource based view these competitive advantages will lead to superior performance over their rivals. In line with both stakeholder theory (Freeman, 1984) and the resource based view of the firm (Barney, 1991), CSR is justified from an economic point of view in that firms build competitive advantages over their rivals.

3. Developing reputation and legitimacy

By engaging in CSR practices, firms are able to create value through strengthening their legitimacy and corporate reputation. The arguments underlying this mode of value creation are again based on a stakeholder approach but with a strong focus on the consumer. As mentioned, the improved reputation of a firm increases its ability to attract employees, investors and

consumers (Smith, 2003). Several researchers found that there is a substantial group of consumers who are influenced in their buying behavior by the CSR reputation of firms (Smith, 2003; Mohr et al., 2001) A market survey by Cone Communications (1999) showed that two-third of the sample was willing to switch to socially responsible brands under the condition of parity in price. Among socially and politically active consumers this amount increased towards 77%. Similar surveys were also conducted in the academic world, Smith and Alcorn (1991) found that almost half of their sample was willing to switch brands to companies supporting social causes and charity. Creyer & Ross (1997) studied consumer responses to both ethical and unethical behavior of firms and reported that a significant amount of the respondents considered CSR as an

important issue, some were even willing to pay a higher price for ‘responsible products’. Survey results show that consumers appreciate socially responsible behavior on the side of firms and that they are willing to reward them. In other words, firms build strong reputations by engaging in CSR activities and this reputation positively influences the purchasing decisions of consumers. This positive effect of CSR on reputation and legitimacy could be amplified when firms choose to disclose information concerning their social and environmental activities (Gelb & Strawser, 2001). Corporate social reporting enables firms to show the outside world that their operations are in line with the ethical expectations of stakeholders. If a consumer is not aware of the degree of social performance of a certain firm, he/she has no incentive to change his/her buying behavior in favor of the responsible firm.

4. Synergistic value creation

22

CSR to create a virtuous cycle of mutually reinforcing benefits for both stakeholders and the company itself. By realizing high corporate social performance, firms positively shape the environment in a way that is beneficial for themselves as well (Porter & Kramer, 2002). A good example of such a virtuous cycle is the following: when a firm donates charitable giving to a local university, the quality and knowledge of available human resources in the region increases. On the long term, the firm will have a larger pool of high quality professionals to recruit from. Similarly, a furniture manufacturer could seek a win-win outcome by investing in replanting pieces of rainforest; the environment benefits and the firm itself is guaranteed of a more stable wood supply in the future. So the synergistic value creation perspective suggests that businesses can use CSR practices in order to satisfy stakeholder demands while simultaneously pursuing its profitability interests.

Conclusively, the business case for CSR finds great support from a number of well established theories, a clear and straightforward line of reasoning and a great amount of empirical evidence. In order to investigate the central focus of this study, contingencies in the CSP – CFP nexus, I first examine the business case for CSR. Based on the arguments provided above, I hypothesize:

H1: Corporate social performance positively influences corporate financial performance

3.2 The CSP-CFP relationship and firm size

Several researchers have studied the effect of firm size on corporate social performance (Orlitzky, 2001; Udayasankar, 2007; Stanwick & Stanwick, 1998). However, none of them examined the CSP – CFP relationship in the context of firm size. This while there are convincing arguments for firm size to have a positive moderating effect on the relationship between CSP and CFP and would be valuable to pursue (Orlitzky, 2001). Udayasankar (2007) argues that firm size represents visibility, that is, bigger firms tend to be more visible to the large public. Because of their higher visibility, firms are pressured to respond to the socially responsible requirements as set by society. Given their comparative lower

23

likely to be recognized and rewarded by the consumer. This effect does not only apply to customers but to all relevant stakeholders. Also environmental activists, governmental bodies and institutions tend to be more aware of the CSR practices of larger firms in comparison to smaller firms. In general, stakeholder activism tends to focus more on large and publicly watched companies (Clark et al., 2008).

Another reason for large firms to benefit more from CSR engagement relates to economies of scale. (Keinert, 2008). Bigger firms generally have more resources to their disposal than relatively smaller firms. This allows them to take a more structured and efficient approach towards social engagement. Besides financial resources, large firms have all sorts of other resources which could be allocated to CSR activities. For example, a company with a big management team could create a task force that focuses on developing CSR programs. Due to such advantages, larger firms are able to manage their CSR activities more efficiently which results in a higher return on social investment.

In the last decades, an increasing number of rating agencies and magazines started issuing corporate social responsibility rankings aimed at both investors and consumers. These rankings rate a number of companies on a specified amount of social performance criteria in order for consumers and investors to base their purchasing and investing decisions on the social performance of a firm. These rankings generally focus on large and listed firms, leaving smaller firms unnoticed. The CSR activities of larger firms are therefore better promoted and receive more attention in comparison to smaller firms. Moreover, corporate social reports are either issued as an appendix of the annual report or simultaneously with the annual report. As mentioned, corporate social reporting can be a powerful tool to promote and

communicate the social performance of a firm. The majority of smaller firms does not have an annual report and when they do, it is not as intensively read as that of larger firms. The CSR initiatives of smaller firms are therefore not as visible as those of larger firms.

In conclusion, larger firms are more visible to the outside world and are therefore more likely to capitalize on CSR than smaller, less publicly watched companies. This implies that the effect of social performance on financial performance is stronger for larger than for smaller companies. Based on this line of reasoning I hypothesize:

H2: The relationship between corporate social performance and corporate financial performance is positively moderated by firm size

3.3 The CSP - CFP relationship across industries

24

Afterwards, I elaborate on the theoretical arguments with respect to a firms’ damage potential and proximity to the end consumer as moderating factors of the CSP – CFP relation.

Until recently, researchers have assumed that the effect of CSP on CFP is homogeneous across industries. The majority of the studies do include industry as a dummy variable to control for the effect of industry drivers on firm performance (Margolis & Walsh, 2008). However, there is hardly any study focusing on the moderating effect of industry characteristics on the relationship between social and financial performance. There are a handful of researchers who studied the business case for CSR using a sample of firms operating in one single industry (e.g. Ogden and Watson, 1999; Simpson and Kohers, 2002) and a few studies which examine the moderating effect of one specific industry characteristic (e.g. Hull & Rothenberg, 2008; Semenova & Hassel, 2008). It is quite odd that this factor has been neglected during 40 years of research to the business case for CSR. Considering the wide range of differing industry specific characteristics, it would be implausible to assume that the impact of CSP on CFP is consistent across industries.

Recent empirical evidence shows that the CSP - CFP relationship is indeed more pronounced for firms in particular industries. Hull & Rothenberg (2008) report that the level of differentiation in the firms’ industry partly determines the return on CSR investments. They reason that CSR can be used as a differentiation strategy and that the added differentiation provided by CSR initiatives, has a lower impact on firm financial performance in high-differentiation industries than in industries characterized by low differentiation. To date, only one study examined the business case for CSR in the situational context of several industries. Hoepner et al. (2010) demonstrated that CSP only positively influences CFP in 2 out of 10 industries, the results suggested a neutral relationship in 5 industries and a negative association in the remaining 3 industries. In line with the contingency perspective of the present study, the results imply that the CSP - CFP relationship is heterogeneous across industries. Hoepner et al. (2010) argue that the main reason for this heterogeneity stems from the different business models on which industries operate. They distinguish the different business models upon four industry specific traits: dependence upon individual stakeholder groups; proximity to the end consumer; potential for social and/or environmental damage; level of product/service differentiation. In the context of the business case for CSR, the second and third are the most relevant and interesting in my opinion.

25

level of social responsibility of the products they buy. Therefore, it is most likely that firms operating on a B2B model will be less rewarded for their social performance than firms operating on a B2C model. This assumption is supported by empirical evidence of Baron et al. (2009), who found a positive association between CSP and CFP for businesses who serve end consumers and a negative association for businesses serving other businesses. These findings are supported by Lev et al. (2010) who report that contributions to charity lead to a significant sales growth, but only for firms operating in consumer industries.

Therefore, I hypothesize:

H3a: The relationship between corporate social performance and corporate financial performance is stronger for firms with a high proximity to the end consumer

The potential of a certain industry to cause damage to the environment and society has been identified as a cause to the heterogeneous CSP – CFP relationships across industries as well. Arguments in favor of this view mainly focus on the cost side of the business case for CSR. A net profit flowing from CSR participation only exists when the returns are higher than the investment itself. For firms with a high potential for social and/or environmental damage, e.g. firms operating in the oil & gas industry, the costs of preventing latent damages are of course higher than for firms with lower damage potential. For some firms it is thus more expensive to attain a certain degree of social performance than for others, this leads to heterogeneous returns on CSR investment. Accordingly, Semenova & Hassel (2008) find that the relationship between environmental performance and market value is indeed better depicted in “low environmental risk industries” than in “high environmental risk industries”. Furthermore, the reputation of firms that structurally harm the environment and/or society, is generally not strong. Heavily engaging in CSR initiatives can therefore be perceived as hypocrite and will have a weaker, or even a negative, effect on a firms’ stakeholders. Based on these arguments I propose that:

H3b: The relationship between corporate social performance and corporate financial performance is weaker for firms with a high potential for damage to the environment and society.

27

4. Data & Method

Where the previous chapters were concerned with the theoretical foundations of the present study, this chapter focuses on the dataset, variable construction and methodological issues. First, I discuss the secondary data that I have collected in order to test my hypotheses. Second, I explain how I constructed the variables present in my conceptual model. Finally, I discuss the methods according to which I test my hypotheses and analyze the dataset. Descriptive statistics regarding the data are presented throughout the chapter.

4.1 Data

28 Table 1: Descriptives and correlations

Mean S.D. 1. 2. 3. 4. 5. 6. 7. 1. CSP 0.7400 2.2170 1 2. ROA 4.5465 10.260 0,067** 1 3. ROE 14.732 28.112 0,092** 0,301** 1 1 4. Assets/1000 1,966.1 63,021 0,057** -0,060** 0,024 5. Employees 31,834 65,947 -0,045** 0,028 0,048** 0,281** 1 6. CSP*Assets 1,454.9 7.5924 0,172** -0,027 0,000 0,361** -0,062** 1 7. CSP*Employees 23,557 48,801 0,205** -0,007 -0,013 -0,064** -0,405** 0,262** 1 8. R&D Exp./100 2,934.1 8,070.9 0,008 -0,006 0,021 0,072** 0,007 -0,010 0,011 Notes: ** p<0.01, * p<0.05. ROA and ROE reported in percentages.

4.1.1 Dependent variable

Corporate financial performance

Among the most frequently used measures for corporate financial performance in the context of the CSP - CFP link, are return on assets (ROA) and return on equity (ROE) (Griffin & Mahon, 1997). Also in constructing this variable I follow previous studies (Wokutch & McKinney, 1991). The initial analysis was conducted using ROA as a measure for CFP. Afterwards, I replaced ROA for ROE to see whether I obtain comparable results when using a different financial performance indicator. I collected annual ROA data for 90,4% of the observations and annual ROE data for 91.4% of the total number of observations. Data on both CFP indicators was gathered through ‘data stream’. ROA represents net income over the total assets of a particular firm and is expressed in percentages. For industrial and manufacturing companies the amount of total assets is represented by the sum of total current assets, long term receivables, investments in unconsolidated subsidiaries, other investments, net property, plant and equipment and other assets. For banks, insurance companies and other financial companies a slightly different approach is used, this can be found in appendix A1.1. Return on equity stands for net income over the total amount of shares outstanding and is expressed in percentage terms as well.

4.1.2 Key independents

Corporate social performance

29

scholars as well. Waddock & Graves (1997, p. 508) acknowledge that “KLD’s rating scheme makes several advances beyond those used in earlier research”. Similarly, Mock & Hoy (1995) argue that because of its multiple and objective social performance screening criteria, the KLD ratings are superior in comparison to the largely perceptual data from the Fortune 500 survey.

The KLD database provides numerical assessments on eight different dimensions of social performance: community, corporate governance, diversity, employee relations, environment, human rights, product and the participation in controversial business practices such as tobacco, gambling and nuclear power. Every dimension contains a number of rating criteria which can be either strength or a concern. The firms in the KLD database are assigned a 1 or a 0 for every criterion, where 1 means that the firm satisfies the criterion and 0 represents a dissatisfaction of the particular criterion. For example, one of the strengths within the community dimension is the degree to which firms maintain good relationships with indigenous people in areas of their proposed or current operations. If a particular firm has notably strong relations with indigenous people in their area of operations, they are assigned a 1 on this criterion. Similarly, one of the concerns within the community dimension is the degree to which company’s actions have resulted into a negative economic impact on the community. If the company is guilty of such

practices, they will be assigned a 1 on this criterion as well. In total there are 39 strengths and 34 concerns to be judged on.

For this study I have assessed every single strength and concern on its relevance to corporate social responsibility. I have deleted the criteria which I considered to be irrelevant with respect to CSR. Accordingly, I removed the entire corporate governance dimension out of my dataset because corporate governance practices are aimed at wealth maximization of shareholders instead of being beneficial for the environment or society. Furthermore, I chose to not include the controversial business dimension in my research. This because practices such as alcohol, tobacco and gambling are sufficiently integrated in, and accepted by, society to my opinion. The inclusion of these criteria may negatively bias the results. After removal of all irrelevant criteria, a total of 28 strength and 25 concerns remain. A final score per company is calculated by subtracting the total amount of concerns from the total number of strengths. This leads to a scale ranging from a minimum CSP score of -25 (extremely low social performance) till a maximum CSP score of +28 (extremely high social performance).

Firm size

30

because of their small reliance on human resources, the number of employees is generally low. For labor intensive firms the opposite applies; as they rely primarily on human resources and to a lesser extent on capital, the value of their assets will be lower and their number of employees higher. Therefore, I ran the analysis with both constructs. In the initial analysis firm size was measured by the amount of total assets. Afterwards, I performed a robustness check by using the number of employees as an indicator of firm size. The data for both indicators was collected for respectively 90,5% and 89,4% of the total firm years by using ‘Datastream’. The total number of employees is calculated by taking the sum of all full and part-time employees, seasonal and emergency workers are excluded. For measuring total assets, I used the exact same data as for calculating return on assets. Details on this measure are outlined above and in the appendix (A1.1).

Proximity and damage potential

In order to assess the moderating effect of a firms’ proximity to the end consumer and potential for damage to the environment and/or society, I structured my sample into 10 different industry classes. Based on these classes I determine which industries are characterized by a high proximity to the end consumer and a high potential for damage to the environment and society. The firms belonging to these classes will then be considered as firms with either a high/low proximity to the end consumer or firms with a high/low potential for damage to the environment and/or society.

In structuring the dataset, I followed SEC level 2 industry classifications which recognizes the following industries: (1) Financials, (2) Technology, (3) Industrials, (4) Healthcare, (5) Consumer Services, (6) Utilities, (7) Telecommunications, (8) Basic Materials, (9) Consumer Goods, (10) Oil & Gas. I consider this classification as sufficiently detailed to draw conclusions concerning the proximity to the end consumer and the potential for social and environmental damage per industry. Additionally, by taking a relatively broad classification, the sample sizes of the 10 industry based sub-samples remain large which contributes to the reliability of my results.

In preparation of the statistical analysis I created two different industry clusters, one which comprises the industries with a high proximity to the end consumer and one comprising the industries with a high potential for damage to the environment and society. Based on an industries’ proximity to the end consumer, I expect the CSP – CFP relationship to be significantly stronger in 2 out of 10 industries; the consumer goods and consumer services industry. Companies operating in these industries generally sell directly (or via an intermediate such as a store or supermarket) to the (socially concerned) consumer. Thus, it is most likely that firms in these industries profit more from CSR than firms in industries

31

society, there is one industry particularly interesting; the oil & gas industry. It is common knowledge that oil refineries are among the most polluting firms of the world. Their potential for damage to the

environment, and therefore to society, is illustrated by a number of scandals like the 2010 BP oil leak in the Gulf of Mexico. Fighting and preventing such potential damages is so costly that it is harder to profit from CSR. Therefore, the oil & gas industry represents the industry class ‘damage potential’. Hereafter, I converted both the original and self-created industry classes into codes and dummy variables, one per industry class/cluster.

4.1.3 Control variables

In order to test whether the observed effects on firm performance are solely caused by the

independent variable and interaction terms, I extend my model with a number of control variables. For testing H1 and H2 , I controlled for the effect of firm size and industry. Furthermore, I re-ran all

regressions in addition of R&D expenses as a control variable in order the see whether the original results remain robust.

Previous studies have suggested firm size, industry and R&D expenses to be important determinants of CFP. (Waddock & Graves, 1997). Firm size is relevant because basic economic theory suggest large firms to be more profitable as a result of economies of scale and scope. This may lead to economic advantages in manufacturing operations (Thompson, 1967); increased control over external stakeholders and resources (Aldrich & Pfeffer, 1976) etc. In addition, firm size appears to be an exogenous

determinant of CSP as well (Stanwick & Stanwick, 1998). These external influences on both my independent and dependent variable, may seriously confound the relationship between CSP and CFP. Similarly, research has shown that there are significant differences in financial performance levels across industries (Waddock & Graves, 1994). Therefore, I took the effect of industry on financial performance into account as well. Finally, R&D expenses represent a firms’ level of innovativeness and may thus have a positive effect on CFP. McWilliams & Siegel (2000) show that neglecting R&D expenses as a control variable leads to an upwardly biased estimate of the CSP-CFP relationship.

R&D expenses are measured by all direct and indirect costs related to the creation and development of new processes, techniques, applications and products with commercial possibilities. The construct excludes government and/or customer sponsored research. A more detailed description of this construct is provided in the appendix (A1.2). The data was gathered by using ‘Datastream’ for a total of 3007 firm year observations. For operationalizing firm size and industry as control variables, I used the same data and approach as previously explained.

32

and it is most likely that time-specific circumstances, such as a financial crisis or a period of extreme growth affect a firms’ financial performance as well. Such an event may bias the results, therefore, I included year dummies in all regression models.

4.2 Empirical model

The model and its corresponding hypotheses were tested with SPSS statistical software. I estimated my proposed model by using several multiple regression analyses. Multiple regression analysis is used to determine the correlation between multiple independent variables and a single outcome variable. In order to determine the moderating effect of industry and firm size, I mean-centered the two indicators of firm size and the independent variable CSP by subtracting the average of all values from every single observation. Mean-centering is aimed at reducing multicollinearity between the interaction term and original variables. Afterwards, I generated the actual interaction terms by multiplying the mean-centered firm size by the mean-centered CSP and the self-created industry clusters, by the mean-centered CSP. Since the industry clusters are dummy variables, they were not mean-centered. During the regression analysis these interaction terms were treated as independent variables. This leads to the following model:

Y = ß

+ ß

X

+ ß

X

+ ß

X

+ ß

Y= corporate financial performance (ROA or ROE) ß0= CFP value where all other variables are zero

ß1= regression coefficient of corporate social performance (KLD scores)

ß2= regression coefficient of the interaction term firm size*CSP

ß3= regression coefficient of the interaction term damage potential*CSP

ß4= regression coefficient of the interaction term high proximity*CSP

33

on assets was substituted for return on equity and total assets was substituted for a firms’ number of employees. In addition, I performed a robustness check by including R&D expenses as control variable. As a final robustness test, I ran the original regressions with aggregated values per company. This means that for every company and every variable the average value of firm year observations was taken. This leads to one single observation per company and reduces the sample from 4,900 firm year observations to 1,288 observations. By taking average values per company over the time period of 1995 to 2003, I correct for company-specific circumstances which could lead to extraordinary values. The dataset allows me to conduct a time series analysis as well. However, since the focus of this study is on examining

contingencies in the CSP – CFP relationship and not on determining the direction of causality, I consider a longitudinal analysis as irrelevant.

All models were tested for all the firm year observations taken together and including the year dummies to control for time-specific influences. For example, if a certain company or even an entire industry experienced a financial crisis in between 2000 and 2003, it will only affect the results to a certain extent because the years 1995 till 1999 are included as well. Therefore, this approach works as a sort of control variable in the sense that all kinds of situational circumstances are filtered out of the data.

34

5. Empirical Results

Table 2 reports the results of the basic regression aimed at testing hypothesis 1 (model 1-4) and 2 (model 5). By performing the analysis according to an enter method, 5 different models are generated where every model is extended with an additional factor. Year dummies are included in model 4 and 5 but not reported because of space limitations. The first regression examines the prediction that CSP positively influences CFP, as can be noted, the results are statistically significant and in favor of this prediction. The regression coefficients for CSP are significant at a 99% confidence level in all of the 4 models, so before and after the addition of each variable. Table 2 presents the unstandardized beta coefficients, which is the actual amount by which the dependent variable increases as a result of a 1 point increase of the

independent variable. Considering the complete model, model 3, this means that a 1 point increase on the CSP scale leads to a 0.307 percentage point rise in ROA for companies in the consumer goods industry. Taking the mean ROA of 4.546 as reported in table 3, a 0.307 increase will result in a ROA of 4.853 which is an increase of 6.753%. Although 0.307 seems to be a small amount, it is quite substantial relative to the average ROA and sufficient to accept hypothesis 1.

Looking at the adjusted R2 in table 2, we can see that it increases from 0.008 to 0.038 when adding the industry dummies and to another 0.062 after addition of the year dummies. This means that the predictor variables explain 6.2 % instead of 0.8% of the variance of the outcome variable when the industry and year dummies are added to the regression equation. To assess the effect of industry on ROA, I entered 9 out of 10 dummy variables into the estimation model leaving one as a reference category. In this model the reference category is “consumer goods” and the regression coefficient for every dummy represents the amount by which ROA differs from the reference category. As can be noted in model 3, the majority of industries (Financials, Technology, Industrials, Utilities, Basic Materials and

35

in total assets, ROA decreases with 10.49 percentage points. Therefore, I conclude that firm size is negatively related with corporate financial performance in the form of ROA.

Table 2: The relation between X and Y and the moderating effect of firm size

Model 1 Model 2 Model 3 Model 4 Model 5

Intercept 4.317 ** (0.154) 4.511** (0.160) 6.711** (0.393) 6.157** (0.544) 6.138** (0.545) CSP 0.312** (0.066) 0.329** (0.066) 0.397** (0.067) 0.307** (0.067) 0.317** (0.068) Assets/billion = -10.49** (2.320) -6.321 (2.469) -6.982** (2.448) -6.269* (2.593) CSP*Assets/billion = = = = -0.559 (0.670) Industry dummies Financials = = -4.109** (0.579) -3.574** (0.574) -3.558** (0.574) Technology = = -5.124** (0.581) -5.124** (0.581) -5.110** (0.581) Industrials = = -1.733** (0.514) -1.735** (0.508) -1.724** (0.508) Healthcare = = 0.030 (0.645) 0.348 (0.638) 0.367 (0.639) Consumer services = = -1.270* (0.538) -1.076* (0.531) -1.060* (0.532) Utilities = = -4.128** (0.676) -4.076** (0.668) -4.063** (0.668) Telecommunications = = -4.841** (1.040) -4.784** (1.027) -4.767** (1.027) Basic materials = = -2.398** (0.659) -2.624** (0.652) -2.595** (0.653)

Oil & Gas = = -2.229**

(0.744) -2.160** (0.735) -2.144** (0.735) Year dummies included No No No Yes Yes Adj. R2 N 0.004 4,900 0.008 4,900 0.038 4,900 0.062 4,900 0.062 4,900

Notes: Standard errors in parentheses. ** p<0.01, * p<0.05. Reference categories: Consumer good and year 2003. Dependent variable: ROA