• No results found

Exploring the link between CSR and competitive advantage : the possible influence of employee skepticism on employee performance and behavior

N/A
N/A
Protected

Academic year: 2021

Share "Exploring the link between CSR and competitive advantage : the possible influence of employee skepticism on employee performance and behavior"

Copied!
49
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Executive Programme in Management Studies – Strategy Track Amsterdam Business School, University of Amsterdam

Master Thesis

Exploring the link between CSR and competitive advantage:

The possible influence of employee skepticism on

employee performance and behavior

Author: Rudo Gischler

Student number: 10317236

Date: 31st of August, 2014 First Supervisor: E. Dirksen MSc Second Supervisor: dr. W. van Eerde

(2)

1

Preface

“Believe you can and you’re halfway there.” - Theodore Roosevelt.

The last letters of my Master Thesis have been written and as such there comes an end to the exhausting and at the same time (although contradicting) energizing quest in pursuit of my Master title. It was challenging sometimes, but I can honestly say I enjoyed to raise the bar every step of the way, improving my performance every semester in continuous search of my limits. Looking back I would not recommend anyone working over 40 hours a week and study simultaneously, but in hindsight things are always a lot more obvious than they are during execution....

Although I first decided not to, I have a strong desire to make use of the opportunity to express my gratitude to those that have accepted, supported and challenged me during the course of the study. There are two specific mentions that have to be made. I can state with certainty that the results of my study behavior would have never been so fruitful without the support of my project team members and friends Chahid, Victor and Ferdinand. Endless nights have we spent discussing theory and expanding our knowledge base during every course. 30% of the time spend was because we had to, the remaining 70% was because we wanted to. We challenged and complemented each other. Debated and confirmed each other’s statements. Supported and burned each other’s work. I cannot find the words that would reflect my feelings sufficiently, but I am sure you know how I truly feel when I state that it was mad fun with you guys and I will surely miss our frequent hang out sessions.

(3)

2 The second token of my appreciation can only go to one person. As always the most important person is to be thanked last and in this case it is no different. Fleur, I can honestly say that you have been a bigger support to me than I could have wished for at the beginning of the journey. You left me studying hard when I needed it most, managing all peripheral matters without hesitation. You accepted my dedication and fanatical habits towards both work and study and often gently reminded me not forget to eat, drink sleep and love before things would escalate. I know you are proud of me and my achievements during the last couple of years, both in work as well as study, but you deserve a big part of any credit I receive. Thank you. Thank you. Thank you. And a million times more. I am done studying now. Let’s enjoy life together again. We have some catching up to do.

Rudo Gischler

(4)

3

Table of Contents

Abstract ... 4

1. Introduction ... 5

2. Literature review and hypotheses ... 9

2.1 Competitive advantage ... 9

2.2 Corporate Social Responsibility as a source of competitive advantage ... 11

2.3 Employee skepticism as moderator on employee performance and behavior ... 14

2.4 Research model and hypotheses ... 16

3. Methods ... 19

3.1 Sample characteristics ... 19

3.2 Procedure ... 19

3.3 Scales ... 20

4. Procedures and measures ... 22

5. Data analysis and results ... 24

6. Discussion ... 33

6.1 Limitations ... 34

6.2 Managerial implications ... 36

7. Conclusion ... 38

Appendices ... 44

Appendix 1 – Professional Skepticism ... 45

Appendix 2 – Employee Job performance ... 46

Appendix 3 – Employee behavior ... 47

(5)

4

Abstract

Corporate Social Responsibility (in short: CSR) has been identified as a positive influence of a firm’s competitive advantage. Whereas most CSR related research focuses on the institutional or firm level, this study focuses on the individual employee level. Using 111 survey responses this study identifies and confirms the positive relationship between a firm’s CSR activities and both employee performance as well as employee behavior. The expected moderating effect of employee skepticism on both relationships has not been identified. This could be a result of limitations of the current research model or the limited amount of responses and as such there is a need for future research.

Keywords: CSR, Competitive advantage, Skepticism, Employee performance, Employee behavior, individual level, workforce

(6)

5

1. Introduction

Although some firms are still having difficulties to identify the specific benefits of their Corporate Social Responsibility (CSR) programs (Ambec & Lanoie, 2008), research has indicated there is a link between the level of CSR related investments and the level of competitive advantage a firm has. The key is found in individual stakeholder management, specifically employees (Mitchell, Agle & Wood, 1997). This study discusses the current linkages between CSR activities of the firm and the impact on its employees. Specifically this research addresses the gap present on individual employee level: the influence of employee skepticism on employee performance and employee behavior.

Competitive advantage of the firm has been a fundamental topic ever since the rise of profit driven organizations and the market environment they compete in (Bain, 1956; Kay, 1994; Porter, 1980; Chaharbaghi & Lynch, 1999). In earlier research, competitive advantage was found in an advantageous position one has in comparison to another (Porter, 1985). Later unique and valuable resources were described to lead to a competitive advantage (Barney, 1986; Peteraf, 1993). Widely discussed examples of specific resources such as fertile land were a source of competitive advantage because of its value, its rareness, its inimitability and its non-substitutability (Barney, 1986). These resources generated rents where others had none and thus created an increased financial performance over competitors.

Although this resource driven research is more than 20 years old, resources are currently still important in the market place. The types of resources that actually have an impact on a firm’s performance however are changing. Traditional sources of competitive advantage have become less significant (Pfeffer, 1994), whereas the availability of a qualitative workforce has become increasingly important (Pfeffer, 1994; Teece, 1998; Turban & Greening 1997; Greening & Turban, 2000). Corporate Social Responsibility (CSR) has been identified as one of the triggers for (prospective) employees to join a firm (Turban & Greening, 1997, Greening

(7)

6 & Turban, 2000) and as such CSR is a source of competitive advantage, specifically for job seekers with a high level of job choice (Albinger & Freeman, 2000).

CSR has been an influencing topic for some time now, and it is stated that most of the executed research focused on the institutional or firm level, rather than the individual level (Aguinis & Glavas, 2012). In the researches that did have the individual level as a primary focus, stakeholder involvement is identified as being an important factor within CSR (Mitchell, et.al, 1997). However, the specific relationship between CSR and individual employee performance has not received much attention in recent research. Managers are expected to raise awareness of their CSR activities in order to attract and retain talented employees (Bhattacharya, Sen & Korschun, 2008). They are expected to do this in such a way that the sincere motives of the firm to conduct in CSR related activities are debated by the public nor its employees, in order to avoid claims of ‘green washing’ (Du, Battacharaya & Sen, 2010). Ideally, a firm would not communicate its CSR related efforts in order to avoid skepticism of the public. The result is that the public might not be aware of the firm’s good deeds. The result is a ‘catch-22’ situation, as described by Morsing, Schultz and Nielsen (2008). Firms are expected to engage in CSR activities but are not expected to actively seek acknowledgement from its stakeholders. The solution is found in executing an inside-out approach in which employees perform a key role. Firms should focus the communication about their CSR activities on finding employee commitment, before external stakeholders are to be informed (Morsing et. al, 2008).

Managers have to overcome employee skepticism regarding the sincerity of the CSR activities of the firm, in order to truly win the hearts of the employees (Du et al., 2010). As such employee skepticism regarding CSR activities of the firm is a possible influencer of an employee’s mindset. Skepticism might influence an employee’s performance and behavior,

(8)

7 although the direct link between employee skepticism regarding a firm’s CSR activities and the performance and behavior of the employee has never been sufficiently researched.

Not many research papers focused on the individual level of analysis when researching CSR (Aguinis & Glavas, 2012). Several empirical papers however identified elements such as employee retention (Jones, 2010), employee identification with the organization (Carmeli, Gilat & Waldman, 2007), employee engagement with the firm (Glavas & Piderit, 2009) and improved employee relations (Glavas & Piderit, 2009) as individual level outcomes of CSR activities of the firm. In their summarizing research Aguinis and Glavas (2012) have categorized these outcomes of CSR on an individual level in three main categories, being ‘employee performance’, ‘employee behavior’ and ‘employee attitude’. This research will make use of these categories, with a small adjustment. The concept skepticism is perceived as being related to attitude, although both are conceptually not substitutable (Obermiller & Spangenberg, 1998). The concept attitude is larger than mere skepticism. However, given that the primary unit of analysis is skepticism as a possible moderating factor (as part of the broader concept attitude), the construct employee attitude from Aguinis and Glavas (2012) will specifically be ignored as output variable during this research to avoid ambiguous measurement results. The remaining two constructs, employee behavior and employee performance are to be researched as such.

This research will examine the impact of employee skepticism regarding a firm’s CSR activities on employee performance and employee behavior, as indicators of a firm’s competitive advantage. This research discusses the current linkages between CSR at one side and employee performance and employee behavior at another. It will research the gap currently present on individual employee level: the moderating effect of employee skepticism on the positive effect of firms CSR activities on employee performance and employee behavior. The research question is therefore:

(9)

8 In which way does employee skepticism have an influence on the effects of a firm’s CSR

related activities on employee behavior and employee performance?

In order to research the relationship between employee skepticism regarding a firm’s CSR activities at one side and employee performance and employee behavior on the other, this exploratory research will purely focus on the individual level perceptions. It will use deductive quantitative research. First theoretical research will be performed on each of the four constructs in order to provide a solid theoretical framework. Then the method and accompanied hypotheses will be described. This section is followed by an overview of items used within the questionnaires. Employees will be asked to self-administer their levels of skepticism, performance, and behavior and provide insight in their opinion about the firm and specifically the level of CSR activities of the firm. These individual measures will be obtained using surveys. The data received will be checked and altered where needed and then the relationship between the employee level properties and the influence of skepticism will be derived using statistical research. The results will be discussed in the discussion and conclusion section, including the limitations of current research.

In identifying and exploring the relationship between employee skepticism regarding CSR activities of the firm and employee performance and employee behavior, managers will be enabled to develop individual counteractive measures to signal or even decrease employee skepticism and thus increase employee performance and positively influence employee behavior. This will enable management to increase the added value of the specific ‘resource’ and thus positively influence its competitive advantage. From a theoretical stance this research will add value to the underrepresented individual level of CSR impact, and create new insight in the employee attitudes towards both the job performed as well as the outcomes of its own actions.

(10)

9

2. Literature review and hypotheses

To fully grasp the relationship between a firm’s CSR related activities, its competitive advantage and representation of this competitive advantage by specifically measurable properties of a firm’s employee, it is necessary to explain the existing views and the theoretical foundation of the individual concepts of the framework. The following paragraphs will discuss each individual element and provide a theoretical context to which the element is benchmarked.

2.1 Competitive advantage

Although in practice many organizations aim to achieve a sustained competitive advantage within the marketplace, the theoretical concept itself is highly ambiguous (Rumelt, Kunin & Kunin, 2003). From a positional perspective, competitive advantage is described as a relatively simple concept, namely a favorable position the focal firm has over its competitors, when delivering a good or a service in a competitive landscape (Porter, 1985). From a resource perspective competitive advantage is translated into above normal returns (Barney, 1986). Both stances imply there is a difference in performance of one firm when compared to another and thus competitive advantage as such is by definition a deviation from perfect market theory (Barney, 1986). In a (hypothetical) perfect competitive market there are large numbers of buyers and sellers which all act fully rational and have access to complete information. Additionally the demand for products or services is homogeneous and all transactions are performed without any costs. This results in all firms performing identical because every firm has the same access to information, resources and buyers (Conner, 1991). Consequently, there are no winners or losers in this market; there is zero economic profit to be made (Conner, 1991; Besanko et al., 2009).

The resource based school deviates from this perfect market competition by stating that factor market imperfections are the source of competitive advantages. Barney (1986)

(11)

10 translates competitive advantage into above normal returns, which are materialized while making use of the factor market. Barney introduced the ‘VRIN’ criteria of resources, stating resources need to be valuable, rare, inimitable and non-substitutable in order to enable a competitive advantage. Conner (1991) continued on the pursuit of above normal returns, stating that there are two ways of achieving such returns. The first is the possession of a low-cost position when a firm is selling an identical product as competitors do in the marketplace. The second is the existence of a buyer’s perception that the firm is selling a unique good or service, which is distinctive from substitutes in the market. In addition to both Barney and Conner, Peteraf (1993) made the distinction between the source of a competitive advantage and the sustainability of the competitive advantage. In order to have a competitive advantage there need to be heterogeneity and ex-ante barriers to competition. In order to be able to make it a sustainable competitive advantage there is a need for ex-post barriers to competition and imperfect mobility (Peteraf, 1993).

There are critics on the resource based school as well. For instance the adaptation of value with regard to competitive advantage is creating ambiguity, as described by Rumelt (2003). He states that it is never made clear when this value materializes, to whom and how it is measured. Rumelt (2003) also argues that there is confusion if competitive advantage “means winning the game or having enough distinctive resources to maintain a position in the game” (Rumelt, 2003, p.3).

To avoid increasing complexity throughout this paper, competitive advantage will be defined as “the ability gained through attributes and resources to perform at a higher level than others in the same industry or market” (Christensen & Fahey 1984; Kay 1994; Porter 1980; cited by Chaharbaghi & Lynch 1999, p. 45). By using this definition the factor time as argued by Rumelt or the sustainability of the advantage as proposed by Peteraf are explicitly ignored, as is the value-focused approach of many previously mentioned authors. The factor

(12)

11 resources however is still significant. It was Pfeffer (1994) who firstly claimed that traditional sources of competitive advantage have become less significant, whereas the availability of a quality workforce has become increasingly important (Pfeffer, 1994; Teece, 1998; Turban & Greening 1997; Greening & Turban, 2000). It is this element, the quality workforce, that will receive specific attention during the course of this research.

2.2 Corporate Social Responsibility as a source of competitive advantage

There are many available definitions for Corporate Social Responsibility, often referred to as ‘CSR’, ‘Sustainability’, ‘Corporate citizenship’ or ‘Corporate Responsibility’ (Marrewijck, 2002). In the narrowest sense, CSR explains how managers should handle public policy and social issues (Windsor, 2006). As such CSR is merely a practical toolkit to be used by a firm’s management, rather than a mindset intertwined in a firm’s core. Wood (1991) takes an opposite stance, claiming CSR is ‘..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, p. 693). This stance emphasizes the mandatory interrelation throughout the firm, which is materialized by a set of rules and foundations that capture the spirit of the firm, rather than the willingness of the manager. One of the most commonly known definitions refers to ‘the triple bottom line’, or the three P’s, being People, Planet and Profit (Elkington, 1997). It refers to the quest of a firm’s economic prosperity combined with environmental and social responsive decision making. These actions and decisions should not only be considering a firm’s current needs, but enable future generations to fulfill their needs as well (WCED, 1987).

Although these views provide a general overview of CSR and its foundation, none of these views describe the voluntariness of CSR related activities. And it is this voluntariness that generates a large part of the debate. For years it has been the firms solely mission to gain financial benefit, serving the interests of whoever controls it, at the expense of who does not

(13)

12 (Dugger, 1989). Wealth creation for shareholders is the primary goal of any firm and such is a ‘zero-sum game’ during which one party wins at the expense of the other. This view is in support of Friedman’s view (1970), who states that the only goal of a firm is to make profit. CSR activities as such should only be executed if they add value to the firm. According to Devinney (2009) the corporate social firm as such is a contradiction: A firm as such will never be intrinsically responsible due to its purpose of existence and as such any form of CSR activities performed by the firm is the result of a trade-off. Without any sanctions (for instance a negative impact on the public view of the firm, or financial penalties) a firm would probably not focus on CSR activities at all (Devinney, 2009). And if sanctions do exist, it is to be expected that a firm executes just the mandatory amount of CSR activities to avoid sanctioning. As in any debate there are views supporting the opposite side, claiming that CSR activities need to be performed proactively and that CSR is a must for daily business of firms (Kramer & Porter, 2006; Marcus & Fremet, 2009).

In order to avoid ambiguity regarding the intended meaning of CSR, the following definition of CSR is used as proposed by Aguinis (2011, p. 855) and used by others (Aguinis & Glavas, 2012; Rupp, 2011; Rupp, Williams, & Aguilera, 2010): “context-specific organizational actions and policies that take into account stakeholders’ expectations and the triple bottom line of economic, social, and environmental performance.” In the use of this definition any discussion regarding the voluntariness of CSR (Devinney, 2009; Kramer & Porter, 2006) is avoided.

Within these ongoing discussions one specific area of interest is the relationship between CSR activities and firm performance. According to Devinney (2009) there is no clear relationship between ‘doing well’ and firm value. Devinney also states that the causal link between CSR activities and operational outcome has been highly unclear. This is refuted by Ambec and Lanoie (2008), who propose that the question regarding the positive financials

(14)

13 effects of ‘being green’ is not relevant. According to Ambec and Lanoie (2008), CSR is just another strategic activity performed by firms of which the outcome is ambiguous. CSR as such is no other than mergers & acquisitions, strategic planning and internationalization, all of which are commonly accepted as strategic activities (Ambec & Lanoie, 2008). This is in line with McWilliams and Siegel (2001), who claim that CSR is just another managerial decision which is to be treated as all other managerial decisions. A firm can calculate its ideal level of CSR using a cost-benefit analysis in which elements such as the level of R&D, diversification, advertising and consumer income are combined. As a result there is a neutral relationship between CSR and a firm’s financial performance: The execution of CSR related activities does not increase or decrease a firm’s financial performance. If there would be a positive relationship it would be immediately exploited by other competitors in the market and thus leveled (McWilliams & Siegel, 2001).

Recent research indicates that companies actively pursuing CSR are able to reap certain advantages that other companies cannot (Aguinis & Glavas, 2012). From an innovation perspective Porter and Van der Linden (1995) state that CSR activities could lead to a competitive advantage due to first mover advantage when legislation has not fully matured. They claim that the environment is an opportunity, rather than a threat or limitation. Other benefits of CSR include the ability to attract a higher amount of skilled people, due to a better access to these potential employees (Turban & Greening, 1997, Albinger & Freeman 2000). Additionally a link is made to social theory: Turban and Greening (1997) propose that an employee gains more positive self-concepts working for a firm which invests in CSR. As such CSR activities of the firm could lead to a possible competitive advantage of the firm due to better access to strategic resources combined with an increased attachment of these employees to the firm.

(15)

14 Competitive advantage as such is often measured on an industry or market level. The chosen definition of competitive advantage in this research also has explicit references to these focus areas. This research however focuses on the individual level. In order to have a representation of a firm’s competitive advantage on the individual level, specific elements of employees are used as measurable outcome variables. The constructs that will be measured are employee performance and employee behavior, as proposed by Aguinis and Glavas (2012).

2.3 Employee skepticism as moderator on employee performance and behavior As the attraction of a higher amount of skilled people could lead to a competitive advantage, employees are considered important stakeholders within CSR (Mitchell, Agle & Wood, 1997; Ambec & Lanoie, 2008). However, the subject employees or the individual level of analysis for that matter has received little attention in CSR research. Most of the research focuses on the institutional or firm level (Aguinis & Glavas, 2012).

From an informational or research perspective the subject employees should be of specific interest to any firm, given that employees have certain moral standards and values that influence their job decision (Judge & Bretz, 1992). These values act as a moral compass to the employee and provide direction of what is good and bad (Rokeach, 1973). Judge and Bretz (1992) claim that employees search to find an employer that suits their moral standards best and as such the perceived corporate culture of the firm and the messages sent by the firm (for example during the recruitment process) are weighed by the employee to determine a fit. Employees feel the need to be part of a team and are subconsciously searching to identification and confirmation of the team and the firm as such, in order to strengthen the level of loyalty and commitment to the firm (Ashforth & Mael, 1989). However, in order to achieve a positive influence on the employees’ decision, and thus trigger the positive side of the employees’ moral compass, it is necessary that (potential) employees are aware of positive

(16)

15 activities of the (potential) employer, such as the execution of CSR activities. The amount of highly skilled employees is finite, so in order to win the employees not only financially but also emotionally, firms are increasingly making CSR a part of their total value proposition (Bhattacharya et al., 2008). At the same time managers are having difficulties raising awareness of the employees regarding the company’s CSR activities (Bhattacharya et al., 2008), posing a challenge for a firm to reap the advantages of its CSR activities. Communication about the CSR activities a firm performs is needed, in order to inform all relevant stakeholders. But it is just this communication that could trigger a new challenge: Overcome stakeholder skepticism regarding the motives of the firm to perform CSR related activities (Du et al., 2010).

Skepticism often occurs when a received message is not in line with the expected content of the message, as exemplified by Ditto and Lopez (1992). In their research they demonstrated that people who would receive a negative message (in this example a negative message regarding one’s health, received in a hospital) are more motivated to find a second or even third opinion that is in line with their initial believes, than people that received positive news. When it comes to CSR, skepticism occurs when the decisions to perform CSR activities are not perceived to be triggered by sincere concerns regarding the wellbeing of the world as such, but merely as a reason to raise the firm’s financial performance (Du et al., 2010; Forehand & Grier, 2003). And as such these motives do not connect to the initial message the stakeholders expected, being that CSR activities are performed due to sincere motives. However, stakeholders (including employees) are increasingly willing to accept a certain trade-off. As long as the CSR activities are performed and stakeholders are informed accordingly, these stakeholders are not necessarily negatively affected by the financial benefits that a firm may endure through these activities (Du et al., 2010). Again the key lies in open and consistent communication about the advantages for both the firm as well as the

(17)

16 environment, in order to avoid the occurrence of skepticism regarding the primary drivers of CSR activities (Forehand & Grier, 2003).

It is the construct of CSR related skepticism that has received very little attention within the field of research. Specifically the consequences of this skepticism at an employee level are unclear. Little is known about the relationship between employee beliefs and opinions regarding the CSR activities of the firm and the impact of these beliefs on personal performance and behavior. This exploratory research will provide a starting point of which future research can foster.

2.4 Research model and hypotheses

This research combines four specific constructs, being CSR, employee skepticism, employee performance and employee behavior. These last two constructs are selected as a representation of a firm’s competitive advantage. The research is set up to explore the influence of employee skepticism regarding CSR related activities of the firm on the positive influence of these activities on employee performance and employee behavior. A visual representation of this research set-up is stated below (refer to Figure 1), where all constructs are measured on an individual (employee) level:

Figure 1. Conceptual research model

+ Employee behavior Level of CSR investments of the firm Employee skepticism Employee performance - + - Competitive advantage

(18)

17 Given the findings of Du, Bhattacharya and Sen (2010), employee skepticism regarding CSR might exist, but these findings do not explain the possible moderating effect of employee skepticism on the appraisal of CSR investments. When an employee does not believe in the goals of the CSR activities, one could expect the employee to be increasingly skeptical with higher levels of CSR investments. Consequently one could expect that a higher level of skepticism regarding CSR activities will therefore not lead to an increase in employee performance or employee behavior. Simultaneously it would be expected that the absence or lower levels of skepticism regarding CSR activities would result in higher measures of both items. The primary foundation of this assumption is that a high level of CSR always leads to higher employee performance and an improvement of employee behavior, motivated by the statement that high levels of CSR lead to a certain form of competitive advantage (Albinger & Freeman, 2000).

In order to build a construct, first the two basic assumptions are translated into hypotheses. Higher levels of CSR lead to a higher level of competitive advantage on the firm level, which translates into an improved employee performance or behavior on the individual level. This leads to the following two hypotheses:

Hypothesis 1: The total level of CSR activities of the firm is positively related to the level of employee performance.

Hypothesis 2: The total level of CSR activities of the firm is positively related to the level of employee behavior.

One might expect specific properties of CSR to have an influence on each of both items on the employee level. Specifically one might expect that employee related CSR investments have a positive influence on the employee and enable the employee to give their optimal performance within the organization. This leads to the following hypothesis:

(19)

18 Hypothesis 3: Employee related CSR investments are positively related to employee performance

Additionally one might expect that an employee feels strongly attached to a company that ‘does the right thing’ beyond its legal responsibilities and would therefore do his/her best in order to support the organization for this good cause. In short: An employee would be on his best behavior within a firm that is on his best behavior. This leads to fourth hypothesis:

Hypothesis 4: Social and nonsocial CSR investments are positively related to employee behavior

As previously stated, one would expect that an employee with a higher level of skepticism regarding CSR activities of the firm will not feel as strongly attached to a firm as an employee with a lower level of skepticism, and therefore one might not put as many efforts in one’s daily work as a less skeptical employee might do. This leads to the fifth hypothesis:

Hypothesis 5: The relation between CSR and employee performance is moderated by skepticism, such that the relation is weaker when skepticism is high

Additionally, one would expect that an employee with a higher level of skepticism regarding CSR activities of the firm will have a less positive mindset against the employer, and therefore one might behave less favorably when it comes to the firm or its values. The sixth hypothesis is therefore:

Hypothesis 6: The relation between CSR and employee behavior is moderated by skepticism, such that the relation is weaker when skepticism is high

These six hypotheses will form the foundation of the research. By finding proof for all hypotheses more insight on the individual level attitudes regarding CSR will be gained.

(20)

19

3. Methods

A self-administered survey will be used, which will be shared with the respondents via the Internet. Discussed below are the sample characteristics, the procedure and used measure scales used.

3.1 Sample characteristics

Given the fact that the target audience exists of all employees worldwide, the population is quite extensive and the generalizability of the data is fairly limited. The sample is gained from multiple firms and from people with different backgrounds, education levels and ages. The majority will have the Dutch nationality. The only true common denominator will be the fact that all of the respondents will be employed. The desired error margin rate is 5% and the desired confidence level is set at 95%. With such a large population size the difference in calculation between 20.000 and 20.000.0000 employees is hardly significant. Expected is a response of at least 100 respondents.

3.2 Procedure

Respondents will be reached using direct e-mail and social media, such as LinkedIn and Twitter. Using these methods a link to a Qualtrics survey is shared. This website has clear logos from The Amsterdam Business School and the introductory text clearly states the purpose and estimated duration of the survey, ensuring each respondent will know the research is executed to enable a master thesis research. In the introductory text anonymity and the length of the survey are stressed in order to lower the dropout rate during the execution of the survey. All grouped items are represented together on one page, in order to ensure full understanding of the linkages between those questions and to manage switches between item scales. Additional control variables such as age, gender, level of education and nationality are also asked. Also a GICS business class indicator is used to indicate the working environment.

(21)

20 Overall it results in quite a long questionnaire (84 items) and as such it is a possible burden for employees to complete. In order to ensure complete and correctly answered questionnaires, correct information regarding the duration and content has been supplied up front to potential participants in order to manage expectations.

3.3 Scales

To measure the level of skepticism of the employee, a model to measure professional skepticism as proposed by Hurtt (2010) will be used (refer to Appendix 1). This model consists of 30 items and will be used to benchmark the initial level of skepticism present at one person. The default level is a 6 point Likert scale, however, in order to ensure continuity for the respondents between this construct and the constructs following in the questionnaire a 5 point Likert scale is used ranging from ‘Strongly disagree’ to ‘Strongly agree’. Pretests performed do not identify a loss in reliability for this switch.

In order to measure employee performance, a model of Welbourne, Johnson and Erez (1998) will be used, which proposes the role-based performance scale (RBPS) in order to measure employee performance (refer to Appendix 2). This model includes 20 items divided over five categories, being job, career, innovator, team and organization. Answers will be judged on a 5 point Likert, ranging from ‘Needs much improvement’ to ‘Excellent’. Dhammika, Ahmad and Sam (2012) have tested this model (although specifically in a Sri Lankan context) and confirmed the validity of this model. Their model included 19 items (in contrast with the original 20 items Welbourne et al. suggested). In this research all 20 items of Welbourne et al. will be used.

In order to define the employee behavior, the model of Organizational Commitment (OC) will be used, as proposed by O’Reilly and Chatman (1986) and as applied by Dhammika, Ahmad and Sam (2012) (refer to Appendix 3). This model includes 10 items which will be judged on a 5 point Likert scale, ranging from ‘Strongly disagree’ tot ‘Strongly agree’. This

(22)

21 model arouses the positive influences of CSR best, represented by the fact that employees often select firms that closely matches their beliefs (Chatman, 1989).

To finalize the research model, questions regarding the perceived participation of the firm in CSR activities will be included, according to a model as proposed by Turker (2009). This model identifies four specific elements of CSR, being ‘social & non-social’, ‘employees’ ‘customers’ and ‘government’. This last element is highly debated (Turker, 2009), and

therefore the two questions regarding governmental CSR are excluded. Given the weakness of this component exclusion of these elements does not impact the overall validity of the item list. The model thus includes 15 items which will be judged on a 7 point Likert scale, ranging from ‘Strongly disagree’ to ‘Strongly Agree’ (refer to Appendix 4).

(23)

22

4. Procedures and measures

In total 143 surveys have been submitted. A quick analysis on the level of completion demonstrated that 32 surveys were incomplete. 23 of these had zero answers and were thus excluded from further analysis. Nine surveys were partially completed, but the average level of completion was less than two full completed constructs (of four constructs). Given this level of missing data these surveys were excluded from further analysis. The remaining 111 surveys are complete and used for further analysis (n = 111).

In order to measure the mean and reliability per construct reversed items are recoded into new variables. Specifically items 1, 10, 11, 16, 17, 19, 25 and 26 are recoded into new variables using the following transformations: 1=5, 2=4, 3=3, 4=2 and 5=1. In order to obtain the appropriate data to test all hypotheses, a combination of four constructs is used; one independent variable, two dependent variables and one moderating variable. Refer to Table 1 for all constructs and descriptive statistics per construct.

Table 1. Constructs and statistics per construct.

N = 111 Scale Minimum Maximum Mean Std. Deviation Crohnbach’s Alpha # Items

Skepticism 1-5 3,17 4,50 3,777 ,251 ,737 30

Performance 1-5 2,40 4,50 3,691 ,433 ,877 20

Behavior 1-5 1,80 4,10 3,190 ,516 ,706 10

CSR 1-7 1,27 6,73 4,733 1,075 ,919 15

The perceived participation of the firm in CSR activities is identified as the independent variable, using a 15 item construct. Specifically items 61 to 75 of the survey are used, which are scored on a 7 point Likert scale. These 15 items are combined to form one mean per respondent of perceived level of CSR participation called ‘CSR’ (Mean = 4.73, α =.919), but are also grouped to three separate totals, being ‘Social & nonsocial’ (α =.924), ‘Employees’ (α =.876), and ‘Customers’ (α =.778). All reliability checks were conducted with positive results.

Two dependent variables are identified: employee performance and employee behavior. Employee performance is measured by items 31 – 50 of the survey. Answers were given

(24)

23 using a 5 point Likert scale. A mean of these items is created called ‘Performance’ (Mean =

3.69). This variable was then tested on reliability and approved (α =.877).

In order to measure employee behavior items 51 to 60 of the survey are used, which are scored on a 5 point Likert scale. The answers on these 10 questions are combined into one variable called ‘Behavior’ to form the mean per respondent of employee behavior (Mean = 3.19). This variable was then tested on reliability and approved (α =.706).

The level of skepticism of the employee is identified as a potential moderating variable. To measure the level of skepticism, items 1 to 30 of the survey are used. These 30 items are scored using a 5 point Likert scale. These answers are combined to form a mean per respondent on the level of skepticism called ‘Skepticism’ (Mean = 3.77). This variable was then tested on reliability and approved (α =.737).

(25)

24

5. Data analysis and results

In order to test the identified hypotheses, several parametric tests were executed. Before parametric tests could be performed however, first four assumptions had to be confirmed in order to ensure that the data is indeed parametric. These assumptions are: 1) interval level data, 2) normality of distribution, 3) homogeneity of variance and 4) independence (Field, 2009). The first assumption refers to the type of data: In order to execute parametric testing the “..data should at least be measured at the interval level..” (Field, 2009, p.133). This assumption was tested and accepted: All constructs consisted of continuous interval data, with equal sized intervals (Field, 2009). The fourth assumption, independence, was to be tested through several tests and touches upon several of the identified hypotheses. Therefore the current focus was set on finding confirmation for assumption two and three.

The second assumption refers to the distribution of the data collected and specifically the normality of it. This could be checked in several ways, however the first check was to identify the levels of skewness and kurtosis for each individual construct. These give a first indication of a normal distribution (Field, 2009). After calculating the Z-value for both kurtosis as well as skewness for each construct (using the following formula: kurtosis or skewness divided by the standard error of kurtosis or skewness) it was concluded that there was no significant level of kurtosis at all four constructs. Kurtosis and skewness are identified as being significant if the calculated Z-value is larger than 1.96 or smaller than -1.96 (p > .05 or p < .05) (Field, 2009). There was however a significant level of skewness present at three of the four constructs, being ‘Performance’ (z = -3.23), ‘Behavior’ (z = -2.35) and ‘CSR’ (z = -2.14) (refer to Table 2). This significant level of negative skewness indicated there was no normal distribution and as such the data was not parametric in its current form.

(26)

25

Table 2. Indicators of normality per construct

Skepticism Performance Behavior CSR

Skewness -,123 -,742 -,540 -,492

Std. error of skewness ,229 ,229 ,229 ,229

Calculated z-value skewness -,536 -3,236 -2,354 -2,142

Kurtosis ,334 ,332 ,005 ,168

Std. error of kurtosis ,455 ,455 ,455 ,455

Calculated z-value kurtosis ,735 ,729 ,011 ,369

To continue the use of parametric testing, data transformation was performed. By transforming the data in a fixed manner a normal distribution is created, without the current significant levels of skewness and kurtosis. In order to maintain current linkages and context between the individual constructs, all transformations need to be executed similarly (Field, 2009). However, Field (2009) states that “if you are looking at relationships between variables (e.g. regression) it is alright just to transform the problematic variable…’. (Field, 2009, p.154). This indicates that not all but three out of four constructs are transformed, being Performance, Behavior and CSR. The construct Skepticism will be excluded from the transformation.

Before the actual data transformation could be executed, first the assumptions of data transformations were tested. If a square root or log transformation is considered, the data must be checked against specific data properties: Negative data or zeroes cannot be used (Field, 2009). All three to be transformed constructs were tested and given the results in Table 1 for these constructs, no zero or negative values for these construct were found. This led to the conclusion that there were no limitations to execute the transformation. Given the negative direction of the significant skewness, all three items were transformed in an inversed (or ‘reflected’) way. This impacts the current hypotheses and linkages (Field, 2009). When continuing analysis on this data, one must keep in mind that all data is reversed, and any indicators coming from the data should be reversed as well. In this transformation log10 was

(27)

26 selected after careful analysis of the results of both square root transformation as well as log10 transformation. The transformation using square root did not appear to be sufficient in order to provide a normalized result set. The transformation using log10 did provide a normal distribution for the selected constructs and therefore the log10 method was used.

To execute the inversed transformation, for each construct the maximum value was selected (refer to the maximum values per construct in Table 1) and increased by a fixed value, in this case the value ‘1’. Using this fixed value it was ensured that no zero or negative values appear as a result of the transformation. The actual value per construct was then deducted from this new value and the log10 function was applied to this result. For instance the construct Performance: The maximum value was 4.50. The value ‘1’ was added to avoid zero or negative values, resulting in 5.50. The actual value of Performance per row was then deducted from this 5.50 and the log10 function was applied to the result. The result was then saved in the variable ‘Performance_Log’. The total transformation for this construct results in the following formula: Performance_Log = log10(5.50 – Performance). For each individual construct the described steps were performed, resulting in three new variables for the constructs, being ‘Performance_Log’, ‘Behavior_Log’ and ‘CSR_Log’. A new test of

skewness and kurtosis was executed to demonstrate a normal distribution for all three constructs. The results are showed in Table 3.

Table 3. Indicators of normality per construct after transformation

Performance_Log Behavior_Log CSR_Log

Skewness ,127 -,140 -,416

Std. error of skewness ,229 ,229 ,229

Calculated z-value skewness ,554 -,610 -1,813

Kurtosis -,107 -,239 -,191

Std. error of kurtosis ,455 ,455 ,455

(28)

27 The Z-values for both skewness and kurtosis were no longer larger than 1.96 or smaller than -1.96. It was therefore concluded that no significant skewness or kurtosis was present for Performance_Log (skewness = .55, kurtosis = -.23), Behavior_Log (skewness = -.61, kurtosis = -.53) and CSR_Log (skewness = -1.81, kurtosis = -.42). The normality was confirmed by the execution of a Shapiro-Wilk normality test and the results are shown in Table 4. The significance for each construct was higher than p = .05 and thus normality was established (Field, 2009) for all four constructs. This concluded the tests for the second assumption of parametric testing.

Table 4. Results of the Shapiro-Wilk test of normality.

Statistic df Sig.

Skepticism ,985 111 ,237

Performance_Log ,986 111 ,290

Behavior_Log ,986 111 ,298

CSR_Log ,979 111 ,078

In order to test the third assumption of parametric testing, homogeneity of variance, a Levene’s test was performed for all four constructs, of which three are transformed, where ‘Gender’ is used as an independent grouping variable. The item Gender was selected because

of its assumed independence with all other items. According to this test it was concluded that all variances were considered homogeneous, whereas all significant vales were above the critical level of p = .05 (refer to Table 5). This concluded the tests related to the four assumptions of parametric testing. All assumptions were met and thus the execution of parametric testing was allowed.

Table 5. Levene’s statistics to indicate homogeneity of variance for all four constructs

Levene Statistic df1 df2 Sig.

Skepticism ,029 1 109 ,865

Performance_Log ,022 1 109 ,882

Behavior_Log ,226 1 109 ,636

(29)

28 Now all data was normally distributed and variances were considered homogeneous, a Pearson correlation was executed in order to determine the strength and direction of association between the individual constructs. For this Pearson correlation the original Skepticism construct was used, as well as the three transformed constructs Performance_Log, Behavior_Log and CSR_Log. This analysis provided an association between multiple variables (refer to Table 6). A value of 0.1 is considered a weak association, 0.3 is considered a medium association and 0.5 is considered a strong association (Field, 2009).

Table 6. Pearson correlation after transformation

M SD Skepticism Performance _Log Behavior _Log CSR _Log Skepticism 3,78 ,25 - Performance_Log ,25 ,10 -,308** - Behavior_Log ,27 ,12 -,137 ,289** - CSR_Log ,45 ,16 -,247** ,425** ,419** -

**. Correlation is significant at the 0.01 level (2-tailed).

One must take into consideration that the means and standard deviation of the constructs Performance_Log, Behavior_Log and CSR_Log as represented in Table 6 were transformed and as such deviate strongly from the mean and standard deviation of the Skepticism construct. Additionally Table 6 shows several significant associations. The direction however was not always clear, due to the previous transformation of data. Except for the Skepticism construct all variables were reversed, so any direction of relationship between Skepticism and the other constructs needed to be reversed as well. Individual relationships between Performance_Log, Behavior_Log and CSR_Log were already in the correct direction, given that these items were all transformed in an equal way. Any positive relationship before transformation would also be positive after transformation. Taken this direction in consideration, there were significant medium to strong positive associations between CSR_Log and Behavior_Log (r = .419, p = .000) and CSR _Log and Performance_Log, (r = .425, p = .000) and a significant medium association between Behavior_Log and Performance_Log (r = .289, p = .002). There were weak to medium significant positive

(30)

29 associations (although represented negatively in the table) between Skepticism and Performance_Log (r = .308, p = .001) and Skepticism and CSR_Log (r = .247, p = .009). Based on these findings it can be concluded that there is a significant positive relationship between the perceived level of CSR activities of the firm and the perceived performance of the employee. Hypothesis 1 is thus supported. Simultaneously there is evidence that supports hypothesis 2: there is a significant positive relationship between the perceived level of CSR activities of the firm and the perceived behavior of the employee.

To analyze the associations between ‘Social_nonsocial’ and ‘Employees’ (both part of the CSR construct) and Performance_Log and Behavior_Log, an additional Pearson correlation needed to be executed. However, first these items needed to be checked on normality. Given the skewness of the overall CSR construct, it was expected that the sub constructs would be significantly skewed as well. Table 7 shows that, as expected, these sub-CSR constructs were significantly skewed. Social_nonsocial had a negative skewness of z = -1.98 and Employees had a negative skewness of z = -3.23. This data received similar transformation rules as did the previous skewed constructs, including all checks on negative values and empty fields. The results of the transformation are also found in Table 7. After transformation Social_nonsocial_Log had a negative skewness of z = -1.76 and Employees_Log had a negative skewness of z = -.90. Both were less than -1.96 and thus not significant.

Table 7. Indicators of normality per sub-CSR construct

Before transformation After transformation

Mean Social_nonsocial (Sub CSR) Mean of Employees (Sub CSR) Mean Social_ nonsocial_Log (Sub CSR) Mean of Employees_Log (Sub CSR) Mean 4,5367 4,7586 ,4988 ,4805 Skewness -,454 -,741 -,404 -,205 Std. error of skewness ,229 ,229 ,229 ,229

Calculated z-value skewness -1,979 -3,229 -1,761 -,895

Kurtosis -,646 ,316 -,275 -,128

Std. error of kurtosis ,455 ,455 ,455 ,455

(31)

30 Now that the data is transformed a Pearson correlation was executed using these specific items. Table 8 shows these results. One must take into consideration that the mean and standard deviation of each construct as represented in Table 8 were transformed and as such deviate strongly from the original mean and standard deviation of the untransformed constructs. As all constructs were transformed similarly, all relationships were in the right order; a relationship that was positive before transformation would still be positive after transformation. Given this direction there was a significant strong positive associations between Social_nonsocial_Log and Employees_Log, which is to be expected given that they were both part of the same CSR construct. Therefore this association was ignored. More interestingly, there was a strong significant association between Employees_Log and Behavior_Log (r = .513, p = .000) and a medium to strong significant association between Employees_Log and Performance_Log (r = .422, p = .000). There was a medium significant association between Social_nonsocial_Log and Performance_Log (r = .291, p = .002) and a weak to medium significant association between Social_nonsocial_Log and Behavior_Log (r = .243, p = .010). The medium to weak significant association between Performance_Log and Behavior_Log was identified earlier during the analysis of the high level constructs.

These findings support two of the identified hypotheses. Hypothesis 3 states that there is a positive relationship between the activities of the firm that stimulate the well-being of the employee and the performance of the employee. Given these findings there is a significant relationship between these two items and thus support for hypothesis 3. The significant relationship between social CSR activities and employee behavior is also confirmed, although it is a weak to medium association. This outcome is in support of hypothesis 4.

(32)

31

Table 8. Pearson correlation of sub-CSR constructs

M SD Social_nonsocial_Log Employees _Log Performance _Log Behavior _Log Social_nonsocial_Log ,50 ,20 - Employees_Log ,48 ,17 ,513** - Performance_Log ,25 ,10 ,291** ,422** - Behavior_Log ,27 ,12 ,243* ,513** ,289** -

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

In order to test the possible moderating effect of Skepticism on the relationship between both CSR and Performance as well as CSR and Behavior, two moderated regressions were performed. For these regressions the transformed item CSR_Log was used as an independent item. The first regression used Performance_Log as a dependent item, the second made use of the Behavior_Log item. Skepticism is used as the expected moderating item in both regressions.

Before the actual regression could be performed first the independent and moderating items were ‘centered’; The mean value per item was subtracted from each item in order to

facilitate correct interpretation of the results. Simultaneously this lowers the occurrence of muticollinearity (Voeten & Bercken, 2004). To center the independent item CSR_Log a new item was created, named CSR_Centered. The mean of CSR_Log (Mean = .447) was subtracted from each individual value per respondent, and then the result was copied into CSR_Centered. Subsequently the same action was performed for the expected moderating item Skepticism: the mean of Skepticism (Mean = 3.77) was subtracted from each individual value per respondent and then the result was copied into the newly created item Skepticism_Centered. Lastly a new item was created to enable the measurement of the actual moderating effect in the regression. This new interaction predictor is calculated as the product of CSR_Centered and Skepticism_Centered. This result is named CSR_Skepticism_Centered_Product. A constant (control) variable was added to the regression model to ensure unbiased results by making the mean of the residuals equal to zero.

(33)

32 The first regression that was performed concerned the possible moderating effect of Skepticism on the relationship between CSR and Performance. The results are shown in Table 9. Although the results show significant results for the individual constant and centered items, they do not show a significant result for the product term CSR_Skepticism_Centered_Product (r = .061, p = .485). As such there is no significant interaction effect and thus no moderation effect to further investigate. This gives no evidence in support of Hypothesis 5. Hypothesis 5 is therefore rejected.

Table 9. Regression results CSR, Skepticism and Performance

Beta Significance

Constant ,000

CSR_Centered ,377 ,000

Skepticism_Centered -,204 ,024

CSR_Skepticism_Centered_Product ,061 ,485

The second regression that was performed concerned the possible moderating effect of Skepticism on the relationship between CSR and Behavior. The results are shown in Table 10. Although the results show significant results for some of the individual items, they do not present a significant result for the product term CSR_Skepticism_Centered_Product (r = .091, p = .303). As such there is no significant interaction effect and thus no evidence in support of Hypothesis 6.Hypothesis 6 is therefore rejected.

Table 10. Regression results CSR, Skepticism and Behavior

Beta Significance

Constant ,000

CSR_Centered -,421 ,000

Skepticism_Centered ,060 ,510

(34)

33

6. Discussion

As the impact of CSR investments is not always clearly identified, most of the research performed focused on the organizational or institutional level, rather than the individual level (Ambec & Lanoie, 2008). The goal of this study was to identify CSR related linkages which occur on an individual (employee) level, in order to expand the knowledge base of CSR effects. Specifically the impact of CSR related investments on a firm’s competitive advantage was researched, where employees were the representation of one firm’s advantageous resources over another (competitive) firm. The research started by identifying the relationships between three specific constructs. For each construct specific questions were used to identify the effects of CSR related activities of the firm on employee performance and employee behavior. It was expected that CSR activities of the firm would positively influence both employee performance and employee behavior. Additionally a fourth construct was used, employee skepticism. It was expected that skepticism of the employee would moderate the positive effect of CSR activities of the firm on both employee performance as well as employee behavior.

Findings from 111 respondents indicated that there indeed is a positive relationship between CSR activities of the firm and employee performance. There is also a positive relationship between CSR activities of the firm and employee behavior. Both findings indicate that higher levels of CSR activities of the firm would lead to an increase in employee performance and employee behavior within the firm. As such CSR activities positively influence a firm’s resources and thus a firm’s competitive position.

It was also expected that the level of skepticism present at an employee would moderate the positive effect of CSR activities of the firm on employee performance and behavior. This study found no evidence in support of those hypotheses. No significant interaction effect was identified during this study, meaning that skepticism has no influence on the positive influence of CSR on employee performance and employee behavior within the firm. These

(35)

34 results are counter intuitive, however these findings might be caused by the limitations of the model. This will be discussed in the limitations of this research.

6.1 Limitations

Although this study finds evidence for four of six hypotheses, there are limitations in the research setup. The findings as presented in this study demonstrate there is a relationship between the total level of CSR activities and the level of employee performance. There is a significant medium to strong positive relationship, which could indicate that an employee works harder because of the CSR activities of the firm. Likewise there is a significant medium to strong positive relationship between employee behavior and CSR activities of the firm, indicating that an employee ‘does good’ because the firm ‘does good’.

Firstly it is only proven that there is a relationship between CSR activities of the firm and both employee performance and behavior. As such this is a limitation. The causality and direction of both relationships should be further investigated during future research. Currently it is unclear if the firm ‘does good’ because the employees demand a firm to do so, or that employees specifically select firms that match their intrinsic beliefs, as proposed by Judge and Bretz (1992). Future research could investigate the causality of the items.

Secondly it is very difficult to have a large and diverse enough sample set in order to draw generalizable conclusions from it for the whole population. The current sample size is not large enough to support any generalizable conclusions. Additionally, the current responses are mostly dominated by Dutch employees, with a few exceptions. Therefore the results could be flawed, due to specific cultural differences. Future research which includes a higher level of diversity could overcome these potential barriers.

(36)

35 Thirdly the list of questions is quite extensive, which might have resulted in a large dropout rate or flawed responses. In total there were 173 surveys started of which 111 were completed, resulting in a drop-out rate of almost 36%. However the total impact of the long questionnaire could be higher, as the duration and length of the questionnaire were clearly stated in the invitational e-mail which could have demotivated respondents to open the survey link. Although the current data set gives no clear indications for it, employees might have rushed answering questions due to the duration of the questionnaire.

Fourthly the timing of the questionnaire can have impacted the overall outcome. Currently the world is recovering from a financial crisis, which has negatively impacted the global financial system, including the Dutch market. The crisis led to a national recession, where jobs and wages were under high pressure. This resulted in many firms actively trying to survive, rather than trying to achieve a competitive advantage. Given this setting one could expect the results being influenced by a (temporarily) change in employees’ mindsets: An employee could be happy just to have a job, rather than trying to find an employer which meets their beliefs best. As such one could (temporarily) be less skeptical, or willing to (temporarily) accept a more negative working environment in order to be guaranteed a monthly income. Future research which includes an analysis of the local marketplace and economic performance could provide sufficient insight.

Fifthly this research had been executed in a quite homogeneous group, being employees. Although the firms involved quite differ, it could be that employees as such are less skeptical than, for instance, entrepreneurs. It could also be that the potential employees that were most skeptical were filtered from the application process due to the current interview-mindsets, assessments or judgments. The performance of a broader research with more variety in its target audience could answer these hypotheses.

(37)

36 Lastly the model used to measure skepticism is originally set up to measure professional skepticism for auditors. This model as such has no specific relationship with CSR and this could have influenced the findings regarding the presumed moderating effect of skepticism. Future research is needed in order to find a measure of CSR related skepticism, which then can be used to improve the research model used in this research.

6.2 Managerial implications

As resources are per definition scarce and the availability of highly skilled employees is finite (Barney, 1986; Bhattacharya et al, 2008), increasing the value of resources can be vital in the long term survival of the firm (Barney 1986, Peteraf, 1993). The findings of this research indicate that engaging in CSR activities has just that effect. This research confirmed that CSR related activities of the firm are positively related to the performance and behavior of the firm’s employees, thus increasing the value of these specific resources of the firm. As such active participation in CSR activities could bring two great benefits to a firm: not only can a firm ‘do good’ in the eye of the public by engaging in CSR activities, it can also benefit from doing good by improving its own competitive position.

When engaging in CSR activities however, managers have the challenge to inform potential employees about the CSR activities of the firm in order to attract employees with matching beliefs. While doing this, managers should prevent using CSR as a marketing mechanism, in order to avoid claims of ‘green washing’ (Du, Battacharaya & Sen, 2010) and thus raising the level of skepticism of stakeholders, including (potential) employees. Communication of CSR activities is not necessarily bad, as long as the communication is perceived as being honest and consistent, in order to avoid skepticism regarding the motives of the firm (Forehand & Grier, 2003). This is a balance for which current research has not yet found the best approach. This study confirms the need for CSR activities and indirectly the need for CSR communication to raise awareness of these activities, both for current as well as

(38)

37 potential employees. Although the moderating effect of skepticism is not confirmed during this study, managers should reevaluate their communication plan regarding CSR activities. Firms are advised to be honest about their motives to engage in CSR activities, especially because these activities are expected not only to benefit the environment but also the firm itself. Stakeholders are increasingly willing to accept this trade-off as an acceptable outcome (Du et al., 2010) and an open communication initiated by the firm could prevent stakeholder skepticism against these motives.

Referenties

GERELATEERDE DOCUMENTEN

omzetting van het gereedschapmateriaal 1-n een andere chemische verbin- ding wordt bepaald door de reaktiesnelehid van de omzetting. Deze is sterk afhankelijk van

 The obtained velocity fields resolved under structured and unstructured mesh conditions show minor dependence on the used mesh in the mean velocity compared to the

This study employed a critical approach towards the discourse of advertising in order to ascertain the linguistic and visual features of the persuasive language

In the pilot, we evaluate the four services mentioned: social interaction, social activities, medication intake and compliance, and health monitoring.. Before the pilot,

The tool framework is used to answer the questions of the deployment question set and the textual representation of the architectural model is produced by the tool given in

6.1 Framework Agent-Principal problem Develop a new PMS New cues Network implements Establish a network Prinicpal Network Performance Measurement System Scripts and

While we know from previous research that contingent reward leadership in general is positively related to employee job performance (Podsakoff, Todor &amp; Skov, 1982; Bass,

Therefore, by means of this explanation, we expect that job satisfaction can explain why extraverted employees in general have better employee job performance than those