Corporate social responsibility and financial performance : what is the role of innovation?

(1)

What is the role of innovation?

MASTER THESIS

University of Amsterdam

Faculty of Economics and Business

Author: Volkan Capkurt

Student number: 10615296

Supervisor: Dr. Dipl.-Wirt.-Ing. Sebastian Kortmann

(2)

Abstract

Corporate social responsibility (CSR) has become increasingly important as the concept that frames the business contribution to sustainable development. Over the past decade, there has been a dramatic increase in the implementation of CSR programs from firms and organizations of all sizes. The link between corporate social performance (CSP) and corporate financial performance (CFP) has been the subject of a large number of studies. Most of the studies tend to share the conclusion that the stronger the firm’s involvement in CSR activities is, the higher the economic and financial value firms will be able to obtain. However, there is a shortcoming of research on the role of innovation in this linkage. Hence, this study attempts to fill this gap and provide information about the role of innovation. Based on a considerable body of literature built around the CSP-CFP linkage, this research is set out to test the mediating effect of innovation on the relationship between CSP and CFP. The hypotheses were built under the assumption that CSP has a positive influence on CFP; CSP has a positive influence on innovation; and innovation mediates the relationship between CSP and CFP. However, the results were inconsistent with the hypotheses and no mediating effect was found for innovation. The obtained non-significant results in this study cannot provide substantial information for answering the research question about the role of innovation on the CSP-CFP link. Both the decreasing R&D expenditures and the influences of the financial crisis on the CSP-CFP link can be considered as reasons for the unsatisfactorily answer of the research question. However, the results provide fruitful insights for future research.

(3)

1.3 Measuring Innovation ... 34 2. Sample ... 35 3. Control variables ... 35 3.1 Size ... 36 3.2 Risk ... 36 3.3 Industry ... 37 4. Research method ... 37 IV Results ... 38 1. Descriptive statistics ... 38 2. Correlation analysis... 40 2.1 Hypothesis 1 ... 41 2.2 Hypothesis 2 ... 41 2.3 Hypothesis 3 ... 41 3. Regression analysis ... 42 3.1 Hypothesis 1 ... 42 3.2 Hypothesis 2 ... 43 3.3 Hypothesis 3 ... 45

V Discussion and conclusion... 46

1. Discussion ... 47

2. Conclusion ... 52

(5)

I

Introduction

- “There is one and only one social responsibility of business – to use its resources and engage in activities designed to increase its profits so long as it stays within the rules of the game, which is to say, engages in open and free competition without deception or fraud.”1 -

Nobel Prize winner Milton Friedman started a big debate in 1970 with the abovementioned quote in the New York Times. According to Friedman corporate social responsibility (CSR) programs are unnecessary expenditures, because they are outside the profitmaking scope of firms. Progressing deeper into his argument, you can also argue that Friedman supports the integration of CSR programs, but only if it has a positive effect on the profitability of a firm in the long-term. But still, Friedman labeled CSR programs as “unnecessary expenditures”.

Nowadays the perception of CSR is remarkably different. CSR has become increasingly important as the concept that frames the business contribution to sustainable development. Over the past decade, there has been a drastic increase in the implementation of CSR programs from firms and organizations of all sizes. The increase in expenditures to enhance the social responsibilities of firms suggests that managers’ perceptions of such policies have shifted from unnecessary expenditures to a critical business function.

The concept of CSR finds its origin somewhere in the 1950s at Bowen’s book “The social Responsibility of Business Man” (Bowen, 1953). In this book the author provides the first modern contribution to the CSR topic by referring to “the social

1

Milton Friedman, "The Social Responsibility of Business Is to Increase Its Profits." The New

York Times, September 13, 1970.

(6)

responsibility of the businessman”. The CSR movement finds its origin in the United States, where other social movements, including consumer rights and the environmental movement, spread out. In the following years, researchers gave many definitions of CSR and many CSR related concepts were introduced. The development of the definitions of CSR has centered around two themes: (1) corporate actions in the economic, societal and environmental dimension and (2) sustainability. Furthermore, a major attention began to be dedicated to the output of social initiatives and the concept of corporate social performance (CSP) was born. CSP can be defined as: “a business organization’s configuration of principles of social responsibility, processes of social responsiveness, and policies, programs, and observable outcomes as they relate to the firm’s societal relationships” (Wood, 1991). This definition focused on results put the basis for the research of the potential link between CSR and corporate financial performance (CFP), whereby CFP can be defined as “the financial outcome of business operations”.

The link between CSP and CFP has been the subject of a large number of studies. Most of the studies tend to share the conclusion that the stronger the firm’s involvement in CSR activities, the higher the economic and financial value firms will be able to obtain (Orlitzky et al., 2003). However, there is not that much research on the role of innovation in this linkage. Hence, this study attempts to fill this gap and provide information about the role of innovation. I propose to examine the relation between CSP and CFP and, most importantly, the role of innovation in this relationship. There have been a large number of studies focusing on the link between CSR and a company’s financial performance. But the role of innovation in this relationship was neglected until McWilliams and Siegel (2000) published their work. They showed a flaw in previous studies about the link between CSR and CFP since

(7)

these studies do not control for innovation. McWilliams and Siegel (2000) have empirically showed that the contradictory results obtained on the relation between CSP and CFP are due to the existence of the omitted variable, innovation that is positively related both to CSP and CFP. The reason for this positive relationship can be primarily traced back to the company strategy. CSR can be seen as part of a differentiation strategy that also implies an investment in innovation. According to McWilliams and Siegel (2000, p. 608), “many firms that actively engage in CSR are also pursuing a differentiation strategy, involving complementary strategic investments in research and development”.

This thesis aims to touch upon the neglected role of innovation on the linkage of CSP-CFP and investigates the role of innovation on the link between CSR and corporate financial performance. Based on the discussion above, the following research question is formulated:

What is the role of innovation in the relationship between corporate social performance (CSP) and corporate financial performance (CFP)?

This study has implications for the beneficiaries of CSR programs, for managers and future research and aims to analyze the effect that innovation has on the link between CSP and CFP providing a broader understanding of this effect and contributing to the literature in this area. Future research in the area of CSR may consider how innovation impacts the relation between corporate social performance (CSP) and CFP. This thesis is structured as follows. Firstly, a critical review of the existing literature is outlined. Secondly, a literature gap is revealed and a research question is formulated. Thirdly, the research design is drawn. Subsequently, the results are discussed. Finally, the main conclusions and contributions of the paper to both theory and practice are presented.

(8)

II

Literature review

Based on the research question, literature on the relationship between corporate social performance and corporate financial performance and the role of innovation in this link will be examined. Firstly, an overview is given on the concept of corporate social responsibility in section 2.1 and the concept of corporate financial performance in section 2.2. Secondly, an overview is given in section 2.3 on the different theories and result from empirical research into the theorized relation between CSP and CFP.

1. Corporate Social Responsibility

The way businesses run and operate has witnessed a remarkable change the past decades, due to the demand for excellence and all-round growth became the primary objective of firms. However, after the reputation for ethical practices by firms hit an all-time low in terms of consumers trust during the years 2001 and 2002, managers have been criticized for their single-minded view about value maximization. The turn of events, like the Enron scandal in the United States resulting in the loss of thousands of jobs, has pressured companies to put serious efforts in CSR activities. Primarily due to financial scandals and a serious drop in investors’ confidence CSR had become an important aspect in strategic decision making of a firm. CSR has emerged as an increasingly pervasive phenomenon on the global landscape that can increase the financial performance of a firm and suggest that corporate decision makers must take care of social and environmental affairs in order to maximize long-term financial returns.

Due to the many conflicting interests and goals of stakeholders, no consensus is reached on a universal definition of CSR. Research on CSR derives from various disciplines and literature is primarily analyzed based on different theoretical

(9)

frameworks across various levels (Aguinis & Glavas, 2012). Scholars have studied firms’ social responsibilities for many decades (e.g. Berle, 1931; Bowen, 1953; Frederick, 1960). However, only recently the interest in specifically CSR has become more widespread (Wagner, Lutz & Weitz, 2009). Bowen (1953) was one of the first to talk about business’ responsibilities to environment and to give a definition of CSR: “It refers to the obligation of businessmen to pursue those policies, to make those decisions, or to follow those lines of action which are desirable in terms of the objectives and values of our society”.

In the 1970s, CSR theories and definitions increased. An often cited definition of CSR is given by Carroll (1979):“The social responsibility of business encompasses the economic, legal, ethical and discretionary expectations that society has of organizations at a given point in time”. Carroll (1979:500) argues that “the social responsibility of business encompasses the economic, legal, ethical and discretionary expectations that society has or organizations at a given point in time". He was one of the first scholars who recognized that a definition of CSR should include an economic aspect and his definition is embedded in a conceptual model of CSP, well known as ‘Carroll’s pyramid’ (figure 1).

The economic responsibility forms the foundation of the pyramid, since businesses’ primary objective is to provide a return on investment to its shareholders. The second layer of the pyramid refers to the duty of a company to act within the given legal framework. The ethical dimension refers to the obligation to live by the moral values and to be fair. Philanthropic responsibilities, which deal with the way businesses contribute to the community, are placed on the top of Carroll’s pyramid.

(10)

Figure 1 The Pyramid of Corporate Social Responsibility.

In the 1980s and the 1990s the definition of CSR evolved into different aspects of CSR and alternative concepts now included: corporate citizenship, corporate sustainability, corporate accountability, sustainable business, corporate responsibility, etc. (Garriga & Mele; Zu, 2009). The concept that recently generated the most attention recently is corporate sustainability and was initially defined by the WCED (1987) as: “Sustainability development can be defined as development that meets the needs of the present without compromising the ability of future generations to meet their own needs”. Van Marrewijk and Were (2003) also defined corporate sustainability and argue that their definition also applies to CSR: “Corporate sustainability refers to a firm’s activities – voluntary by definition – demonstrating

(11)

the inclusion of social and environmental concerns in business operations and in interactions with stakeholders”.

Dahlsrud (2006) analyzed 37 definitions of CSR in the period between 1980 and 2003 and identified five dimension of CSR: environmental, economic, stakeholder, voluntariness and social. The environmental dimension refers to environmental issues in business operations. The economic dimension refers to the financial aspects of CSR that are contributing to firm’s economic development. The third dimension of CSR – stakeholders - is about the way that firms interact with their employees, suppliers and customers. The fourth dimension of voluntariness refers to CSR actions beyond legal obligations, which are based on ethical values. The social element implies the relationship between business and society and the way the firm integrates social concerns in their business operations.

The need for social responsibilities and ethical frameworks in business has become a main priority in our society. This is supported by the fact that a large number of firms integrated CSR programs into their business operations. Especially multinational enterprises (‘MNEs’) incorporated sustainability practices in their core strategy. Firms should not only focus on ethical and environmental issues, but take a broader perspective and also address economic and social issues (Closs et al., 2011). The growing importance of CSR initiatives today hints that the perception of such policies shifted from unnecessary expenditures to a critical business function. The critics believe that CSR programs will not provide any value for the business (Friedman, 1970). Proponents of CSR emphasize that sustainability practices might even lead to a competitive advantage for firms and provide value for business and society (Freeman, 1984; Porter & Kramer, 2006).

(12)

Aguinis (2011:855): “context-specific organizational actions and policies that take into account stakeholders’ expectations and the triple bottom line of economic, social and environmental performance.” Additionally, Aguinis & Glavas (2012) emphasize that although the definition of CSR refers to policies and actions by organizations, such policies and actions are influenced and implemented by actors at all levels of analysis (e.g. institutional, organizational and individual). The institutional level of analysis addresses at least one of Scott’s (1995) three pillars of institutions: normative, cultural-cognitive and regulative elements like laws and standards. On the other hand, the organizational level of analysis focuses more on the firm level and financial outcomes, reputation and firm risk. The individual level of analysis generally focuses on psychological theories and draws upon normative motives such as alignment to commitment and awareness of CSR (Aguinis & Glavas, 2012).

2. Corporate Social Performance

CSR is difficult to measure due to its broad scope, because it covers internal and external factors of the firm and also aspects of implementation and performance. Scholars use corporate social performance (‘CSP’) as a variable to measure CSR. Carrol (1999) states in his overview of the corporate social responsibility definitions that the 1980s mark the start of the CSP-CFP relationship studies. Ever since then, the definition of CSP has evolved significantly. CSP was defined by Wood (1991) as: “a business organization’s configuration of principles of social responsibility, processes of social responsiveness and policies, programs and observable outcomes as they relate to the firm’s societal relationships.” This definition was later adjusted and redefined as “internal stakeholder effects, external stakeholder effects, and external institutional effects”, because of the increasing importance of stakeholders (Wood &

(13)

Jones, 1995). In the studies on the relation between CSP and CFP, scholars have defined CSP in two different ways: (1) CSP is a multidimensional construction because it is comprised of both internal and external factors that must be considered when measuring CSP or (2) as a function of how stakeholders are treated by a company.

Margolis et al. (2007) argue that previous studies on the relationship of CSP and CFP can hardly be generalized, because they vary in how they measure CSP. Orlitzky et al. (2003) presented four strategies in their analysis on the best means to measure CSP. In that analysis they refer to different measurement tools such as reputation ratings, CSP disclosures, social audits and managerial CSP principles and values. Additionally, Margolis et al. (2007) recognize that previous studies use self-reports of CSP and third party audits. Third-party audits refer to the systematic assessment of data by researchers who evaluate a firm along a set of criteria (Margolis et al., 2007). Third-party audits are commonly used in studies on the CSP-CFP relationship. The most used third-party audits can be considered Kinder, Lydenberg and Domini (KLD) and Dow Jones Sustainability Index (DSJI).

For the purpose of this study, CSP refers to the observable and measurable outcomes of the corporate social actions, in line with the initially definition of Wood (1991). For measuring the CSP and the innovation data, I will use the KLD database.

3. Corporate financial performance

Corporate performance reflects to the outcomes of the goals that were set by the management and is the ability of the firm to use its resources effectively and efficiently in order to reach those goals. There are various ways of measuring the concept of corporate performance. For example, buyers will rate corporate

(14)

performance in operational terms like product quality, service and affordability. But in the financial markets, it is more likely to consider financial indicators, like stock performance (Fauzi et al., 2010). Corporate performance can be divided into two categories: operational performance and corporate financial performance (CFP). Measurements of operational performance are the quality of the product, market share and the effectiveness of marketing. Same as the definition of CSP, the definition of CFP varies in its scope and measurement means in the previous studies on the CSP-CFP link. Most of the studies use data from databases such as AMADEUS, COMPUSTAT or information provided by stock exchanges (Margolis et al., 2007). Griffin and Mahon (1997) conducted an extensive analysis of CFP indicators and identified more than 80 CFP indicators. Orlitzky et al.(2003) divided the CFP measures in 3 measures: accounting-based, market-based and perceptual measures. However, the most commonly used CFP measures are accounting-based and market-based. The first is for example referring to profit and returns on assets (ROA), the latter includes stock performance and price per share. Market-based measures are used to evaluate future performance, whereas accounting-based measures are used to

evaluate past performance.

This research focuses on corporate financial performance and the link on CSP-CFP and the role of innovation in this link will be defined in financial terms, whereby CFP refers to the financial outcomes of business operations. The study will use accounting-based measures, which will be explained more specifically in the research design.

(15)

4. Friedman vs. Freeman

Researchers have presented opposite opinions on whether CSR activities add any financial value to firms. There are two major approaches linking CSR to CFP. Ironically, one of the most influential written texts about CSR and is effects on CFP cannot be categorized as a research paper. Milton Friedman harshly criticizes the increasing popularity by academics and practitioners towards CSR, in his article “The Social Responsibility of Business is to increase Profits’ (Friedman, 1970), and argues that firms do not have any obligations other than obedience to current legislation and maximize shareholder value. Anything beyond this is a misallocation of valuable resources and should not be considered as corporate policy. Firms incur costs to be socially responsible, which reduces profits as well as the shareholder value thereby leading to a competitive disadvantage (Waddock and Graves, 1997). Friedman (1970) proclaims that employing principles of CSR in the business activities is unfair to the firms´ shareholders and to the general public. The two-fold basis of his critique is basically the misappropriation and misallocation of corporate funds (Margolis & Walsh, 2003).

On the other hand, the stakeholder theory of Freeman (1984) argues that firms need to meet the expectations of shareholders and other stakeholders because the management’s main goal is to align competing stakeholder demands, whereby stakeholders are defined as “any group or individual who can affect or is affected by the achievement of the firm´s objectives” (Freeman, 1984, p. 46). The stakeholder theory of the firm is about creating value for stakeholders through the integration of business and societal considerations. In the stakeholder’s view socially responsible behavior may lead to continuous profitable operations and maximizing shareholder wealth (Clarkson, 1995 and Donaldson and Preston, 1995).

(16)

The theory is based on the following assumptions. First, firms are open systems that interact with the external environment on a continuative basis, whereby the external environment is constructed in constituent groups – stakeholders – that affect and are affected by firms´ actions. Stakeholders have been classified in different ways; either as resource providers or as risk bearers and residual claimants for the value created by the firm (Jones, 1995). Second, the interests of all legitimate stakeholders have intrinsic value and there is no set of interests that dominates the others (Clarkson, 1995; Donaldson & Preston, 1995). This is because the stakeholder theory is driven by taking ethics, morals and values explicitly into consideration as a central feature of the firm, contrary to other theories. Lastly, the stakeholder theory assumes that CSR activities increase the trustworthiness of the firm, which strengthens relationships with stakeholders and decreases transaction costs and so leads to financial gain (Freeman, 1984).

5. The stakeholder theory

Since the publication of Freeman’s (1984) work, the idea that corporations have power over stakeholders has manifested in academic literature. An often-stated argument by CSR commentators is that in order for CSR to be meaningful and effective, the interests of a range of stakeholders other than shareholders need to be taken into account by firms. Therefore is it important to take a closer look at the stakeholder theory.

Hillman et al. (2001) listed four main statements about the stakeholder theory: “(a) the corporation has relationships with constituent (stakeholder) groups, (b) the processes and outcomes associated with these relationships are of interest, (c) the interests of all legitimate stakeholders have value, (d) the focus of the stakeholder

(17)

theory is on managerial decision-making”.

It is often argued that the stakeholder theory is not viable, because it implies the sacrifice of the business objective, which is to make profit, to morally acceptable social goals (Vinten, 2000). However, the stakeholder approach does not reject making profit as a corporate purpose, but rather widens the shareholder approach. The stakeholder approach challenges the idea that shareholders should be the only claimants over the interests of other legitimate claimants (Emiliani, 2001). Clarkson (1995) argues that firms, which do not include their primary stakeholders’ concerns within their strategy, challenge their long-term survival. Therefore, organizations must try to achieve their own objectives (i.e. profitability) meanwhile satisfying the legitimate claims of their stakeholders.

The stakeholder theory can be divided up in two different dimensions. The instrumental stakeholder theory considers stakeholder management as an instrument to achieve expected outcomes. The normative stakeholder theory acknowledges the ethical legitimacy of the stakeholders’ claim on the organizational purpose as primary (Jones et al., 2002).

Johnson and Scholes (2002) defined stakeholders as “those individuals or groups who depend on the organization to fulfill their own goals and on whom, in turn, the organization depends”. A submission to the Australian Parliamentary Joint Committee on CSR from the Australian Conservation Foundation (ACF, 2005) identified the possible stakeholders in a firm’s activities as follows:

(18)

List of Corporate Stakeholders

Group Contributions Relationship Corporate

obligation Shareholders - Financial capital

- Assumption of top risk band - Ultimate management Primarily legal; may also be contractual Dividends and/or increase in capital value consistent with other obligations Financial investors - Financial capital - Assumption of risk - Expertise Primarily contractual Repayment of interest and capital Directors - Management oversight Legal and contractual Compensation

Employees - Intellectual and

physical labor - Experience, initiative, commitment, continuity. Contractual (individual or collectively) Fair compensation and conditions; respect for human rights; safety; employment security consistent with other obligations Customers and end consumers - Intermediate and ultimate demand for

products and services May be direct and contractual, or mediated through retailers; also subject to legal regulation

Duty of care; fair competition and trade practices.

Suppliers - Business inputs Primarily contractual Payment for inputs; fair competition and trade practices Local communities in which the company operates - local security - conducive business - environment

- social, cultural and environmental

amenities - environmental

carrying capacity (biodiversity, land, renewable and non-renewable resources, ecosystem services) Primarily informal and implicit; some local regulation Compliance with laws, taxation, responsible use of environmental carrying capacity and support for community

(19)

5.1 CSR and the stakeholder theory

Carrol (1991) expresses that there is a natural fit between the idea of CSR and an organization’s stakeholders, because the concept of stakeholders personalizes societal responsibilities by delineating the specific groups or persons business should consider in its CSR orientation. Within this context, Maclagan (1998: 147) argues that all stakeholders are claimants on the moral direction of the company: “Corporate social responsibility may be viewed as a process in which managers take responsibility for identifying the interests of those affected by the organization’s actions.”

Wood (1991) points out that creating a balanced outcome, which is acceptable to the majority of the stakeholders, notwithstanding the possible differences in the value systems and ideological positions of stakeholders, is a harsh but necessary task for corporate managers. Furthermore, he suggests that stakeholders are likely to develop a different understanding of what CSR means, what they expect from the corporation in relation to CSR and how they assess CSR.

- subsidies and other support - physical infrastructure State/national communities in which the company operates - As above, plus: - national security - regulation - license to operate - assumption of residual risk in insolvency Implicit in license to operate; legal regulation Compliance with laws, taxation, responsible use of environmental carrying capacity and support for community Global community - International trade - Environmental carrying capacity Almost wholly implicit; mediated through national governments Responsible use of greenhouse and other global environmental carrying capacity.

(20)

5.2 Stakeholder theory in the CSP-CFP literature

The stakeholder theory is often being taken as the conceptual basis of the literature arguing in favor of the relationship in between CSP and CFP. The theory of Freeman (1984), based primarily on the stakeholder theory, argues that superior CSR can lead to improved relationships with key firm constituents, which creates sustainable comparative advantages and increases the long-term financial well being of the corporation. Jensen (2002) developed the “enlightened value maximization theory”, which basically supports that in order for a firm to become and remain successful, the corporate objective has to remain shareholder value maximization and should at the same not mistreat the several constraints with regards to key stakeholders of the firm.. Jensen concludes that the long term survival and successful operation of a firm is to some extent dependent on the relationships with its primary stakeholders. Views like this have spurred an extremely growing body of literature that has tried to answer the question of whether CSR can increase, decrease or not materially affect the financial performance of the firm. With the stakeholder theory being the conceptual basis of the CSP-CFP literature, it seems necessary to take a closer look at some of the work in this field.

Jones (1995) uses various notions and frameworks coming from transaction cost economics, agency theory and business ethics to show that in the case of repeated interactions between a firm and its stakeholders, mutual collaboration based on trust will lead to competitive sustainable advantage. This paper has important implications for the CSP-CFP literature, because the main conclusion can be broken down into more detailed paths that link social attributes of the firm to an improved financial performance.

(21)

financial effect of CSR and show that there is a mismatch between CSP manifestations and CFP measures that are not theoretically linked in any way. They reviewed several studies dealing with carious dimensions and aspects of CSR and argue that market-based CFP measures seem to provide more consistent results among studies, because there is sufficient theory linking CSR dimensions with market oriented stakeholders.

Another paper about the stakeholder theory, which has played an important role in the empirical CSP-CFP literature, comes from Clarkson (1995). He argues that firms are managing their relationships with their stakeholders and not with the society as a whole. Therefore it is important that firms make a distinction between social issues and stakeholder management. Clarkson uses the three levels of analysis of Wood (institutional, organizational and individual), to categorize the stakeholders. Hillman and Keim (2001) go beyond Clarkson (1995) and show that stakeholder management does have an effect on the firm’s CFP while social involvement does not. Hillman and Keim argue that only the CSR dimensions that lead to improved relationships between the firm and their primary stakeholders are important while other dimensions do not have a positive effect on CFP.

6. Relationship between CSP and CFP

The first attempt to analyze the relationship between CSP and CFP has been performed by Bragdon & Marlin (1972). After this, an increasing number of empirical studies have been undertaken to investigate the financial and economic impact of CSR actions. These studies examined the direction, the strength and causality of the CSP and CFP relationship in various industries. Researchers have theorized in many different directions and thought of many mechanisms that could explain such a

(22)

relationship. Despite these effects, there is little consensus on whether CSP may positively impact CFP. Wu (2006) argues that the reason for the contradictory findings in the previous studies might be due to the different CSP indicators used. Margolis et al. (2007) propose that the CSP measures and indicators hardly capture the same underlying concept, because previous studies differ in CFP measurements and the scope that those measurements cover. Those differences enable Fauzi (2009b) to argue that there is a need for a clearer universal definition of CSP and what CFP aspects can be related to it.

In this section, some of the more influential studies are reviewed. Roman et al. (1999) argue that the results produced by CSP-CFP studies can be divided into (1) studies that show no link, (2) studies that show a negative link and (3) studies that show a positive link. Empirical studies that have the most comparable methodology for measuring the relationships between CSP and CFP will be discussed for each of the three linkages.

6.1 No link

The most recent empirical study finding no relationship between CSP and CFP is the study conducted by Aupperle et al. (1985). The components of CSR programs in this study were legal, economic, ethical, and philanthropic responsibilities. In order to avoid some methodological problems of measuring CSP, they created their own measurement for CSP and, through empirical testing, concluded that their methodology was reliable. Their measure included a survey of 241 CEOs. To measure financial performance, they used return on assets (ROA) and “employed both short-term (one year) and long-short-term ROA (five years)” (Aupperle et al., 1985). They concluded that no measure of financial performance is significantly related to social responsibility: “it did not matter whether long-term or short-term ROA were used, nor

(23)

did it matter if the indicator were adjusted or unadjusted for risk”(Aupperle et al., 1985).

Ullmann (1985) also concludes that there is no association between CSP and CFP. He argues that there are so many intervening variables between social and financial performance that there is no reason to expect a relationship to exist, except possibly by chance. Uhlmann (1985) reviewed 13 studies on the relationships between CSR and financial performance. The theoretical study provides an overview of the literature, explaining that the conclusions concerning the relationships of CSP and CFP are conflicting because of the lack of theoretical frameworks, inappropriate definitions of key notions, the use of different CSR databases and differences in the use of control variables. Furthermore, Ullmann indicated that there are many measurement problems that exist to measure the intangible impacts of CSR.

6.2 Negative link

The argument for a negative relationship follows the thinking of Friedman (1970) and other neoclassical economists. According to their view, firms that engage in CSR activities have a competitive disadvantage (Aupperle et al., 1985), because they incur costs that fall directly upon the bottom line and reduce profits, while these costs could be avoided. According to this line of thinking leads CSP to significant additional costs and thus it cannot lead to a better financial performance.

An empirical study that supports the ideas of Friedman is the study of Lopez et al. (2007). They analyzed CSP and CFP between the period 2002 and 2004 and found a negative association. They found that the effect of corporate sustainability on CFP indicators is negative during the first years in which it is applied. Additionally, they found that the expenses that firms incur as a result of CSR activities, can place them at an economic disadvantage with respect to other firms. To measure CFP, they

(24)

used the accounting-based instrument of profit/loss before taxes. To measure CSP, the study used the Dow Jones Sustainability Index (DJSI) and the sample consisted of 110 European firms. The study controlled for industry, size, and risk. They concluded that the companies that engage in CSR programs are at a disadvantage because they are making unnecessary costs. The limitation of this study is that it only measures the short-term relation between CSP and CFP.

Similar to the Lopez et. al. (2007) study, Makni, Francoeur & Bellevance (2009) found a negative relationship between CSP and CFP. They studied a sample of Canadian firms and did not find significant relationship between social performance and financial performance. They conclude that CSP leads to limited benefits for the firm and significant additional costs, and thus cannot lead to an increased CFP.

6.3 Positive link

The third view proposes that there is a positive linkage between CSP and CFP, because the actual costs of CSR are covered by the benefits. A firm that attempts to decrease its implicit costs by socially irresponsible behavior will eventually incur higher explicit costs. Socially responsible firms have less risk of negative events. The positive CSP-CFP relationship is explained by stakeholder theory, as mentioned earlier, developed by Freeman (1984). According to this theory, a company can actually financially perform better by bringing value not only to its shareholders, but to several stakeholder groups such as client, employees, suppliers and the society. It argues that social responsible behavior can find reward in for example improved worker productivity or reputation (Moskowitz, 1972).

The most comprehensive study about the positive link between CSP and CFP was the study conducted by Orlitzky et. al. (2003). They examined 52 studies with a

(25)

sample of 33,878 over 30 years and concluded that CSP has a positive influence on CFP. Additionally, they found very little evidence of a negative association. Their findings support the position that CSR programs are associated with financial benefits in the long term. Orlitzky et. al. (2003: 427) described the relationship as “bidirectional and simultaneous”.

Margolis and Walsh (2003) count 127 studies on the relationship between CSP and CFP in the period of 1972-2002. Even though the majority of the studies found a significant positive relationship, Margolis and Walsh (2003) concluded that the relationship between CSP and CFP should be recognized as ambiguous, complex and nuanced, not allowing for much theoretical generalizations on the strategic implications of CSR activities.

Waddock and Graves (1997) also found a bidirectional relationship between CSP and CFP. Their theory states that good financial performance makes funds available and resources necessary to spend on CSP activities that improve social performance. They argue that CSP is a predictor of CFP and a result of CFP. This is based on the slack resources theory, arguing that firms with slack resources may have greater freedom to invest in CSP, due to strong financial performance. The study of Waddock and Graves (1997) analyzes CSP and financial data of 469 companies from the S&P 500 from 1989-1990.

Surroca et al. (2010) also find a bidirectional relationship between CSP and CFP and linked this with the resource-based view. They found a moderating role for intangible assets on the relationship between CSP and CFP. Their results showed that an increase in CSP or CFP will always result in an increase of the others, when new intangible assets are developed.

(26)

7. Innovation

McWilliams and Siegel (2000) revealed a flaw in previous studies about the relationship between CSP and CFP, because these researches did not control for innovation. Their work started a big debate on the key role of innovation in the CFP-CSP link. McWilliams and Siegel (2000) show that innovation is a key driver of CFP. Furthermore, they suggest that the significance of the relationship between CSP and CFP could disappear, when innovation is included as an independent variable. Researchers have different explanations for the influence of innovation on the CSP-CFP link. Hull and Rothenberg (2008) argue that innovation besides being a strong predictor of financial performance, may also moderate other variables that affect CFP. Wagner (2010) argues that research & development efforts do not have a moderating role on the link between CSP and CFP. This section will give an overview of the studies on the role of innovation in the relationship of CSP and CFP.

7.1 McWilliams and Siegel (2000)

McWilliams and Siegel (2000) have empirically demonstrated that a significant part of the literature on the relationship between CSP and CFP is based on misspecified models, since innovation is not included as a control variable. They show that the contradictory results obtained on the CSP-CFP literature are due to the existence of an omitted variable, innovation that is positively related to both CSP and CFP. Innovation is found to be an important determinant of CFP, and is argued to also be positively correlated with CSP. This mediating role of innovation implies the existence of a positive effect between CSP and the innovation performance of the firm. According to the authors this positive relationship can be traced back primarily to the firm strategy. As they note in McWilliams and Siegel (2001: 608), “many firms

(27)

that actively engage in CSR are also pursuing a differentiation strategy, involving complementary strategic investments in R&D”. They suggest that a firm may improve its CSP by investments in innovation. Therefore, McWilliams and Siegel (2000) suggest that innovation, represented by R&D costs, should be included as individual control variables in studies on the relationship between CSP and CFP.

7.2 Hull and Rothenberg (2008)

Hull and Rothenberg (2008) studied the interaction of CFP with firm innovation and they showed that CSP enhances CFP by allowing the firm to differentiate, and that his effect may be moderated by innovation. Their study measured innovation using a 3-year average of R&D expenditures. They found that CSP has a higher positive impact on CFP when a firm is low on innovation and they argue that firms, which are not able to develop innovative products might obtain a competitive advantage by engaging in CSR activities. Therefore the authors argue that firms that are not able to differentiate themselves with innovative products, may differentiate through engagement in CSR activities, since firms with innovative products will have less incentive to engage in CSR activities to differentiate themselves even further. The results of Hull and Rothenberg (2008) show innovation as a moderator for the positive relationship between CSP and CFP, and suggest that innovation should be included as a moderator between CFP and CSP.

7.3 Padgett and Galan (2010)

Padgett and Galan (2010) examined the impact that R&D has on CSR of manufacturing and non-manufacturing industries and found that the positive effect of R&D intensity on CSR is not the same across industries. The study provided evidence that R&D intensity in manufacturing industry firms are positively associated with

(28)

CSP and that this relationship is significant. They measured R&D intensity specifically and its influence on CSP through a multi-variants regression analysis and showed that there is a significantly positive relation between R&D intensity on the CSP. Padgett and Galan (2010) show that the effect of innovation on CSR is more intense in manufacturing industries than in non-manufacturing ones.

7.4 Surroca et al. (2010)

Surroca et al. (2010) examined the effects of a firm’s intangible resources (Human & Resources, innovation, organizational change and reputation) on the relationship between CSR and CFP by conducting a study among 599 industrial firms between 2002 and 2004. They expected that intangible assets mediate the relationship between CSP and CFP. Their results indicate that there is no direct relationship between CSR and CFP – merely an indirect relationship that relies on the mediating effect of a firm’s intangible resources. The study underlines the importance of intangible assets as determinants of performance. Surroca et al. (2010) argue that profitable and social responsible firms are more capable than socially irresponsible firms in generating intangible assets such as innovation. The study finds that CSR stimulates the development of intangible assets and that these intangible assets in return lead to an improved CFP.

7.5 Wagner (2010)

The study of Wagner (2010) analyses the link between sustainability management and economic performance and possible moderation effects. The authors did not find evidence for the moderation effect of R&D efforts on the link between corporate sustainability and economic performance. Their analysis reveals that innovation

(29)

efforts do not per se improve the effect of CSP and innovation is also not found to be significant as a factor to associate with CFP.

8. Literature gap and research question

The abovementioned CSR literature suggests the need for further research on the role of innovation on CSR-CFP linkage. Researchers have reported different effects of innovation on the relationship of CSP and CFP. This inconsistency may be due to flawed empirical analysis. This study examines empirically the role of innovation on the relationship between CSP and CFP. Consequently, the following research question is proposed:

What is the role of innovation on the relationship between corporate social performance (CSP) and corporate financial performance (CFP)?

9. Conceptual framework

To analyze the relationship between CSP, CFP and innovation, a system of hypotheses is constructed. The hypotheses of this study take the dyadic nature of the CSP and CFP relationship into account and investigates the role of innovation on this link and assume that each of these elements influences each other. The following hypotheses are proposed:

Hypothesis 1: Corporate Social Performance (CSP) has a positive influence on Corporate Financial Performance (CFP).

Hypothesis 2: Corporate Social Performance (CSP) has a positive influence on innovation.

(30)

Hypothesis 3: Innovation mediates the relationship between Corporate Financial Performance (CFP) and Corporate Social Performance (CSP).

The proposed hypotheses suggest that CSP has an effect on CFP, directly and through innovation. Therefore, this study investigates the mediating role of innovation concerning the relationship between CSP and CFP. A mediation effect occurs when a third variable plays an important role in governing the relationship between two other variables (Baron, 1986). The first variable affects the mediator and the mediator affects the second variable. In this case is proposed that innovation is mediating the

(31)

III

Data and method

In this chapter, the research design of the thesis will be introduced, the concepts of CSP, CFP and innovation are operationalized and a description of the data selection process is given.

1. Data sources

In terms of research design, the nature of my topic and research question suggests a quantitative research design. More specifically, a regression analysis will be used for investigating the relationship between CSP and CFP. I will use the linkage of two databases to be able to associate CSR, CFP and innovation. The Wharton Research Database Service (WRDS) is used to access the specific databases. For determining the CSP data, the KLD database will be used, which evaluates the CSP of firms based on various sustainability criteria. For determining the CFP and innovation data, I will use the COMPUSTAT database.

This study will undertake a panel data analysis of firms listed at the S&P 500 Stock Exchange Index (‘S&P 500 Index’), from 2008 till 2012. This specific sample is chosen, because the abovementioned databases allow me to collect data for all the firms that are listed at the S&P 500 Index. Furthermore, I chose the specific time schedule in order to get a recent sample. These databases enable a search for companies listed on the S&P 500 Index by composing an index constituent for the period between 2008 and 2012.

1.1 Measuring Corporate Financial Performance

Although measuring CFP is considered as a simple task, it has its specific complications. There is little consensus about which measurement instrument to

(32)

apply. Some researchers use market-based measures (Alexander & Buchholz, 1978), others use account-based measures (Cochran & Wood, 1984) and some adopt both of these measures (McGuire, Sundgren & Schneeweis, 1988). However, most previous literature and empirical research concerning CSP and CFP used accounting-based data to measure CFP, as opposed to market-based measurements (Waddock & Graves, 1997; Simpson & Kohres, 2002). According to the existing literature, the two measures have different theoretical implications and each is subject to particular biases (McGuire, Sundgren & Schneeweis, 1988; Hillman & Keim, 2001). For this reason the use of different measures can complicate the comparison of the results of the study. Based on the most recent previous study, this study will use accounting-based measurements of CFP.

The three most common accounting-based measurements for CFP are:

• Return on assets (ROA): the ROA will be calculated as the net operating income divided by the total assets.

• Return on equity (ROE): the ROE will be calculated as the net operating income divided by the total equity.

• Return on sales (ROS): the ROS will be calculated as the net operating income divided by the total revenues.

According to Hagel and Brown (2010) ROA is the best way to measure financial performance, because ROA explicitly takes into account the assets used to support the business activities and determines whether the company is able to generate an adequate return on these assets instead of simply showing robust returns on sales. This study will focus on the accounting-based measurements of CFP, whereby CFP will be defined in terms of ROA. This measure seems appropriate, because it yields

(33)

the most direct information about the results in the allocation of resources by a firm as it seeks competitive advantage (Hull and Rothenberg, 2008). The source of the data is the COMPUSTAT database (CompScore) and the used variables to calculate the ROA are: NI (Net Income) divided by AT (Assets Total). The survey covers the firms included in the S&P 500 Index for the years 2008-2012.

1.2 Measuring Corporate Social Performance

Previous studies made use of a wide variety of methods to measure CSP: self-made surveys (Aupperle, 1985), the Dow Jones Sustainability Index (DJSI) (Lopez et al., 2007) and the Kinder, Lydenberg and Domini Index (KLD) (Waddock & Graves, 1997). There are internal (employees, governance, etc.) and external (environmental, etc.) factors that must be taken into account when measuring CSP. The KLD Index takes all those factors the best into account, because it is compromised out of the previous MSCI ESG (Environmental, Social and Governance) Index2, which takes both the internal and external factors into account.

KLD is a rating agency, which assesses hundreds of firms with regard to their strengths and concerns in a series of dimensions. The KLD index is one of the earliest tools for evaluating corporate social performance and one of the most used and accepted tools for measuring CSP. KLD rates companies on nine dimensions of CSP – community relations, diversity issues, employee relations, governance, environmental issues, product issues, military contracting, nuclear power and involvement in areas with human right violations – with a binary score of 1 or 0 being assigned to each of the many indicators of these dimensions. If a company meets the

2_{MSCI ESG Index changed in 2011 in KLD Index after a change of data ownership}

from MSCI Inc. to KLD.

(34)

criteria, it will score a “1” and otherwise a “0”. The KLD Index uses both internal (e.g. annual reports) and external sources (e.g. articles) to conduct year-by-year assessments of hundreds of firms. Based on this information, KLD constructed the Domini 400 Social Index (DSI 400), the functional equivalent of the S&P 500 Index for socially responsible firms. In order to get in the DSI 400, a firm must derive less than 2% of its gross revenue from the production of military weapons, it must have no involvement in nuclear power, gambling, alcohol, and tobacco, and have a positive record in each of the remaining categories. For example, a firm that actively promotes minorities and women to top managerial positions and membership on the board of directors will receive a similar positive score along the diversity dimension.

There are several advantages that come from using KLD to measure CSP. First, KLD investigates the CSP in various dimensions, and is in these ways more useful than using one-dimensional CSP measures that are appropriate only for companies in specific sectors. Second, KLD offers a significantly large data sample of hundreds of firms from multiple industries. Third, it allows for more sophisticated empirical research, because it quantifies many qualitative aspects of CSP and allows for an aggregation amongst several CSP dimensions. Lastly, KLD can be considered as highly reliable, because of the interdependence of research and the use of both internal and external information sources. There has also been critique on the use of the KLD database. According to Wood and Jones (1995: 164), “KLD attempts to quantify the nearly unquantifiable”. Another critique is the use of numerically crude scales. All the above considered together, this study will use the KLD index to measure CSP. The use of KLD ratings tends to constitute the norm when it comes to doing empirical research with CSP as a measure. I will use KLD Statistical Tool for Analyzing Trends in Social and Environmental Performance (KLD STATS) for this

(35)

study by focusing on the companies listed in the S&P 500, starting in 2008 and ending in 2012. The analysis will be conducted at the firm level. As originally suggested by Ullmann (1985), Waddock and Graves (1997) and Hull and Rothenberg (2008) I will construct an aggregate measure of the eight categories of CSP as rated by KLD. According to Hull and Rothenberg (2008) this approach has the advantage of providing a numerical score, rather than a dummy variable as was used by McWilliams and Siegel (2000).

1.3 Measuring Innovation

According to Hurley and Hult (1998) the measurement for innovation should reflect the implementation of new ideas, products or processes. There are three common secondary sources for measuring innovation in business research (Gonzalez-Padron, 2008). The first one relates to the quantity and quality of patents, using the Citation Impact Index as a measure of the importance of a firm’s innovation in relation to other firms. The second source for innovation that is often used for CSR research comes from Fortune’s Most Admired database (i.e. Luo & Bhattacharya, 2006). The third source for measuring innovation is calculating the R&D intensity. The R&D intensity is typically a proxy for innovation as an input in empirical studies (McWilliams and Siegel, 2000; Padgett, 2009).

This study will use the R&D intensity as a proxy for innovation. The R&D intensity is calculated by dividing total expenditure on R&D by total revenues, based on the measures used by McWilliams and Siegel (2000) and Prior, Surroca and Tribo (2008) who showed that R&D is positively correlated with CSR. The source of the data is the COMPUSTAT database and the used variables to calculate the R&D intensity are: XRD (Research and Development Expense) divided by REVT (Revenue Total). The period researched has a span of 4 years, from 2008 to 2012.

(36)

2. Sample

There were different samples used in the previous studies on the CSP-CFP relationship. Scholars often used the Fortune Ratings or the S&P 500 Index, which implies that the researches are often conducted on large companies. This is also the focus of this study. The sample of this study consists of all the companies listed in the S&P 500 Index and rated by KLD on their corporate social performance during the period between 2008 and 2012. While there is an abundance of research on the relationship between CSP and CFP, no study has examined the most recent years and looked specifically on the role of innovation in this context. Appendix 1 (in progress) shows the ticker (TIC), name and industry of each of the firms in the sample. Not all the companies of the S&P 500 will be included in the study due to missing CSP, innovation or CFP data for the years 2008, 2009, 2010, 2011, or 2012. Table 1 (in progress) outlines the sample selection procedure by looking to the availability of the CSP data and the availability of the CFP data (Revenue > 0).

3. Control variables

Van Beurden & Gössling (2008) referred to the importance of factors that might influence the relationship of CSR on economic, environmental and social dimensions and pointed out to control those factors. In order to isolate impact of CSP on CFP this study identifies several factors that can influence CFP and vice versa. Control variables are variables that affect the dependent variable, but which are not part of the relationship. By introducing control variables in this testing, it is possible to determine the influence of CSP on CFP and the other way around without the effects of other variables. These control variables are held constant. Previous studies suggested risk, size and industry as factors that affect both a firm´s performance and

(37)

CSP (Ullman, 1985; McWilliams & Siegel, 2000). Each of these characteristics will be used as a control variable. These variables, and the reasons for controlling them, are discusses in more detail below.

3.1 Size

Size is often called an important control variable of CSP, since larger firms seem to adopt the CSR principles more often, because larger firms have more resources to devote on CSR activities (Waddock & Graves, 1997). According to Artiach et al. (2010) larger firms tend to perform better in terms of CSP, because larger firms have a more thorough assessment of their activities, larger firms attract more attention from stakeholders and larger firms have more visibility with the public because they have larger marketing and advertising budgets. In this study, size is controlled by using the net sales and total assets, as proposed by Callen and Thomas (2009). The source of the data is the COMPUSTAT database and the used variable to calculate the size is net sales, SALE (Sales/Turnover Net).

3.2 Risk

Risk is the second control variable in this study. There are several reasons to control for firm’s attitude towards risk when studying the relationship between CSP and CFP. According to Margolis et al. (2007) is the firm’s risk profile is strongly related to its financial returns. Furthermore, the authors argue that firms that are less risky are more likely to engage in CSR activities, because firms facing high debt/assets ratios will have fewer resources to spend on CSR.

In order to measure risk the ratio debt to assets will be used. This ratio gives an indication of the firm’s financial risk and addresses the ratio of a firm’s long-term debt to its total assets and is also used in previous studies (Waddock & Graves, 1997;

(38)

Hull & Rothenberg, 2008). The source of the data is the COMPUSTAT database and the used variables to calculate the ratio debt to assets are: DT (Total Debt Including Current) divided by AT (Assets Total).

3.3 Industry

The third control variable is industry, since there are clear differences on corporate social performance among industries (Waddock & Graves, 1997). Due to the different internal and external environments of firms, every industry is faced with unique challenges in its economic, social and environmental domains. These differences may relate differently to CFP and, consequently, to different relations between CSP and CFP for different industries (Griffin & Mahon, 1997). For this reasons, the type of industry is controlled for studying the relationship between CSP and CFP.

This study will use the sector classification of the Standard Industrial Classification (SIC) codes, as developed in the United States in 1937. The SIC makes a distinction in the following industries: agriculture; mining; construction; manufacturing; transportation, communication and gas; wholesale; finance, insurance, real estate; services; or public administration. The source of the data is the COMPUSTAT database and the used variables to get the SIC code is: SIC (Standard Industry Classification Code).

4. Research method

This study will apply a “Pearson correlation analysis” and multiple regression analysis (Field, 2009). The Pearson correlation coefficients technique will examine the direction and the strength of the relationship between the dependent and independent variables. A regression analysis will be used to test the hypotheses. This research will use the statistical software SPSS 21.0 (IBM Corp., 2011).

(39)

IV

Results

In this chapter, the results of the analysis will be provided. Firstly, the descriptive statistics of the independent and dependent variables will be presented. Secondly, a correlation matrix will be presented. Finally, the results of the regression analysis will be provided, which also includes the mediating role of innovation on the CSP-CFP link.

1. Descriptive statistics

Table 1 presents the descriptive statistics of variables that are specified in the conceptual framework of the study. It provides the definition, mean, median, standard deviation, minimum and maximum of each of the variables of the entire sample (n = 2588). The sample represents firms from the S&P 500 from ten industries.

According to the statistics provided in Table 1, the mean value for the CSP variable is quite small, ranging between the highest value of 66 and the lowest value of 0. It should also be noted that the CSP value could only take specific discrete values within the [0,1] range. For example, Corporate Governance strength can take a value of 0, or 1 depending on how many of the respective indicators are present for a

(40)

particular firm in a specific year according to the KLD database. For the purpose of this study, the KLD measures are not standardized. This is in line with academic empirical studies, which also did not standardize the CSP values (Artiach et al., 2010; Hillman & Keim, 2001; Waddock and Graves, 2007). The standard deviation of the CSP score indicates an acceptable spread. Therefore, the current analysis can be considered as reliable to work with. However, the standard deviation of the variable size indicates an undesirable high spread and needs additional analysis to make this value more reliable to work with.

The minimum and maximum values for ROA can be considered as representative for the investigated firms and are in line with the expectations. The standard deviation of the ROA indicates an acceptable spread and can be considered as a reliable value to work with. Although, the current ROA value seems to be reliable, the standardization of the variable is preferred to facilitate an appropriate data analysis. Using non-standardized ROA data can be difficult to work with. The data can, for example, be both negative and positive, or be very divergent, since some variables can have very high values, whereas other have very low values. Therefore the standardization of the CFP variable leads to a more reliable assessment of the data

(41)

for the upcoming correlation, regression and mediation analysis. Table 2 presents the descriptive statistics, whereby ROA is standardized. This table provides the most relevant information for the correlation, regression and mediation analysis.

2. Correlation analysis

The Pearson Correlation Coefficients were used to check the strength of the association between the variables CFP, CSP and innovation. Table 3 presents the correlation matrix between the various variables.

As stated in the conceptual framework, for investigating the role of innovation on the relationship between CSP and CFP, the following hypotheses are defined:

Hypothesis 1: Corporate Social Performance (CSP) has a positive influence on Corporate Financial Performance (CFP).

Hypothesis 2: Corporate Social Performance (CSP) has a positive influence on innovation.

Hypothesis 3: Innovation mediates the relationship between Corporate Financial Performance (CFP) and Corporate Social Performance (CSP).

The following sections will describe the correlation results between the various relevant variables independently for each the hypotheses.

(42)

2.1 Hypothesis 1

The correlation between the dependent variable, CFP, and the independent variable, CSP, is relatively low positively related (p = 022). Additionally, the correlation between CSP and CFP is not significant. At this moment, there is evidence for the positive correlation of CSP and CFP. However, the evidence is not strong, due to the missing signification level. Therefore, there is a need to undertake a regression analysis before drawing a conclusion with regard to hypothesis 1. Furthermore, the dependent variable seems significantly negatively correlated to the control variables, and the independent variable is low positively correlated to the control variables. Accordingly, the addition of these control variables is justified. Therefore, these control variables will be included in the multivariate analyses.

The correlation result indicates that there is a low but positive relationship between CSP and innovation. However, the correlation is not significant. As with the previous hypothesis, there is a need to undertake a regression analysis before taking strong conclusions with regard to the second hypothesis. Innovation seems significant negatively related to the control variable size.

The third hypothesis investigates the expected mediating role of innovation in the relationship of CSP and CFP. The correlation between the mediating variable, innovation, and CSP and CFP is positive, whereby the correlation between innovation and CFP is significant. However, the correlation is relatively low in magnitude, with Pearson correlation coefficients < 0,043 in all cases. As with the previous hypotheses,

(43)

in order to understand the mediating role of innovation on the CSP-CFP relationship better, a multivariate regression analysis will be undertaken in the next section.

3. Regression analysis

In order to test the hypotheses a regression analysis has been performed. The following section will present the results for each of the hypotheses. Hypothesis 3 investigates the proposed mediating role of innovation in the CSP-CFP relationship.

Table 4 presents the relevant values to interpret the regression analysis between the dependent variable CFP, the independent variable CSP and the control variables risk, size and industries. The regression results in Model 1 contain only the control variables, after which Model 2 adds the independent variable CSP to Model 1.

The regression matrix presents some interesting observations. Table 4 indicates that the control variables risk, size and industry are mainly negatively

(44)

related to CFP, whereby only the risk effects are significant (p < 0,01). There are also some industry differentials that stand out, given that only the industry effect of the construction industry is significant (p < 0,05), while the other industries have a non-significant effect on CFP. From the eight industries, there are five negatively related to CFP, and there are three with a positive influence on CFP. Model 2 includes the dependent variable CSP in the regression analysis and shows that the R square stays equal. This is an interesting observation, which shows that the addition of CSP does not influence the CFP more or less than the control variables do in Model 1.

Concerning hypothesis 1, even though the effect of CSP on CFP is positive, the results appear to be not significant. The results are showing the importance of the control variable risk and industry. The control variable risk shows a negative significant relation to CFP. The results of the industry effects are indicating industry differences, whereby only construction shows a significant relationship (p > 0,05). The most important observation of the regression analysis is that adding the independent variable CSP does not deliver a significant relationship or increases/decreases the explanatory power of the regression model. The regression model does not provide information about the relationship between CSP and CFP, because there is no significant relationship. Therefore, there is no empirical evidence found to support or reject hypothesis 1.

Table 5 presents the relevant values to interpret the regression analysis of hypothesis 2. As with the previous hypothesis, Model 1 contains only the control variables, after which Model 2 adds the independent variable CSP to the model.