The impact of a firm’s orientation on negative reciprocal behavior of primary stakeholders

(1)

The impact of a firm’s orientation on negative

reciprocal behavior of primary stakeholders

Master Thesis

MSc. Business Administration – Strategy

Faculty of Economics and Business

University of Amsterdam

Date: 17-8-2017

Student: Maarten van der Noordaa

Student nr: 11397748

(2)

2

Statement of Originality

This document is written by Student Maarten van der Noordaa, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

This thesis attempts to address a call from literature to integrate a more realistic assumption of human behavior in instrumental stakeholder theory. The effects of a firm’s orientation (stakeholder vs. profit) on negative reciprocity in case of a negative event are measured by making use of two vignette studies with 1 x 3 and 1 x 2 between-subjects experiments. The first vignette experiment manipulates the underlying management orientation of a firm, while the second study manipulates perceived cues of being watched. Possible mediating effects of anger-like emotions and fairness perceptions are being explored by using Hayes’ PROCESS plug-in for SPSS. The results of this research indicate that there are no significant differences between the vignettes in both studies. Despite the absence of significant results, this research presents circumstantial evidence for the difference in stakeholder punishment between firms projecting either a profit-, or a stakeholder-oriented image. The findings suggest that firms projecting a profit-oriented image elicit less stakeholder punishment compared to firms that project no specific orientation at all. Furthermore, the results indicate that projecting a stakeholder orientation does not increase punishment compared to firms projecting no specific orientation at all. Besides that, no significant results were found for the mediating effects of fairness perceptions and anger-like emotions. To conclude, the implications for the literature and managers are limited to the preliminary finding that stakeholder punishment may decrease when firms project a profit-oriented image.

Keywords: Instrumental stakeholder theory, firm’s orientation, reciprocal stakeholder behavior, self-interested stakeholder behavior, perceived cues of being watched

(4)

4

1. Introduction

"It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our

dinner, but from their regard to their own interest" (Smith, 1776: p.16).

This quote of Adam Smith is a rendition of classical economic theory at the end of the 18th century. More recent literature is based upon similar reasoning and assumes that people are “self-interested utility maximizers”, who “grab whatever they can for themselves” (Economist, 2007: 93). Widely accepted and fundamental theories in the field of strategic management such as transaction cost economics, agency theory and resource-based theory are based on this assumption (Bosse, et al., 2009). However, a recent stream of research on the behavioral micro-foundations of stakeholder theory argues that this depiction of typical human behavior is inaccurate (Bridoux & Stoelhorst, 2016; Bosse & Phillips, 2016; Bridoux & Stoelhorst, 2014; Hahn, 2015; Hayibor, 2012). Moral psychology scholars found that a majority of the people behave reciprocal instead.

At an elementary level reciprocity can be explained as the reward of kind actions and punishment of unkind ones (Gouldner, 1960), at a cost to the individual itself. The reward or punishment occurs to the extent of sacrificing self-interest for the sake of one’s own principles (Bosse et al., 2009), because the actors´ desire to maximize utility is bounded by norms of fairness (Jolls, Sunstein and Thaler, 1998). Reciprocal behavior can either be positive or negative. Positive reciprocal behavior can be explained as rewarding good behavior at a cost to yourself. Applied this to a business context it can be explained as stakeholders’ reward towards a firm, when the firm treats its stakeholders in a way that is being perceived as fair by the stakeholders. On the other hand, when a firm is being perceived as unfair towards their stakeholders, reciprocal stakeholders will punish the firm (Bosse et al., 2009).

(7)

7

The central proposition of instrumental stakeholder theory, that a stakeholder-oriented management approach positively affects long-term performance (Donaldson & Preston, 1995; Jones, 1995), is supported by this mechanism (Quintelier, Stoelhorst & Bridoux, n.p.; Bosse et al., 2009; Bridoux & Stoelhorst, 2014;).

This would imply that profit-oriented firms are likely to elicit more punishment, because they are being perceived as more unfair. Contrarily, are stakeholder-oriented firms more likely to elicit reward because they are perceived as more fair (Quintelier, Stoelhorst & Bridoux, n.p.). However in contrast to this reasoning, in practice is seen that stakeholder-oriented firms encounter both increased reward and increased punishment from their stakeholders (Quintelier, Stoelhorst & Bridoux, n.p.; Briscoe, Chin, & Hambrick, 2014; Janssen, Sen & Bhattacharya, 2015; Schrempf-Stirling, Bosse, & Harrison, 2013). Moreover, Schrempf-Stirling et al. (2013) argue that stakeholder-oriented firms sometimes seem to be more likely, instead of less likely, to become targets of stakeholders activism in comparison to profit-oriented firms (Quintelier, Stoelhorst & Bridoux, n.p.). So what could cause that comparable firms which behave apparently similar can face very different stakeholder reactions?

A possible answer to this question is a theory by Quintelier, Stoelhorst & Bridoux (n.p.), which is founded on crucial insights from moral psychology. This theory proposes that when a firm projects a profit image it is likely to “increase the strength of strategic

evaluations of the firm’s behavior”. Put differently, communicating the pursuit of profit as

underlying management orientation incentivizes self-interested behavior among stakeholders. Stakeholders motivated by self-interest will only contribute to the value creating potential of the firm when this maximizes their own extrinsic payoff.

(8)

8

On the contrary, a firm that projects a stakeholder image is likely to “increase the

strength and frequency of moralistic evaluations of their behavior”. Thus, a management

orientation that aims for the collective well-being of all their stakeholders triggers moralistic evaluations. Stakeholders that engage in moralistic evaluations compare an act or actor with the salient moral norm, which functions to sustain within group cooperation and uphold moral norms. Compliance with this salient moral norm results in a contribution to the firm’s value creating potential over and above what personally pays off, which can be defined as; reward (Quintelier, Stoelhorst & Bridoux, n.p.). Transgression of the salient moral norm results in a detraction form firm’s value creating potential over and above what personally pays off, which can be defined as; punishment (Quintelier, Stoelhorst & Bridoux, n.p.). Put differently, a firm projecting a stakeholder-oriented image elicits moralistic evaluations and consequently this type of evaluations is expected to result in reciprocal behavior.

Depending on the cognitive frame that is being used, a different level of reciprocity is expected. Strategic evaluations are expected to elicit self-interested behavior and therefore not likely to result in punishment above what personally pays off. Contrarily, moralistic evaluations elicit reciprocal behavior and are therefore likely result in punishment above what personally pays off. Taking these different evaluations, or cognitive frames, in consideration could shed light on why stakeholder-oriented firms are more likely to elicit both positive and negative reciprocity compared to profit-oriented firms (Quintelier, Stoelhorst & Bridoux, n.p.).

The first independent variable in this research is therefore the orientation of a firm. This can be either a profit orientation, characterizing a management orientation that aims for profit maximization (Parmar et al., 2010; Freeman, 1984; Freeman, Wicks, & Parmar, 2004) or a stakeholder orientation. The latter epitomizes a management orientation that aims for the collective well-being of all their stakeholders or society at large (Donaldson & Preston, 1995;

(9)

9

Freeman, 1984; Friedman, 1970; Jones, 1995; Smith, 2003). This study will test the effects of the different orientations on the intentions to engage in negative reciprocal behavior, which is the dependent variable in this research. Negative reciprocal behavior will be measured in the form of the intentions to engage in negative word of mouth.

In order to test the arguments for the previously discussed mechanism, and whether this mechanism actually is causing the proposed differences in stakeholder punishment between stakeholder- and profit-oriented firms, a second study within this thesis was developed. The independent variable in this second study is the perceived cues of being watched. The arguments leading to this second study are based on a study of Bateson et al. (2006). This study argues that individuals engage in more cooperative behavior when they feel that their behavior is being watched because they experience a more social context. Therefore this second study developed a vignette with an image of a recognizable pair of eyes above the survey. This image provides perceived cues of being watched, which triggers social context. Since behavior or actions are morally evaluated in the case of a personal context, it is likely that moralistic evaluations are being triggered by these perceived cues of being watched. On the other hand, a profit vignette which could be described as impersonal, is likely to increase strategic evaluations, because a profit image elicits typical economic behavior, which prioritizes self-interest. Therefore it is expected that individuals, when being watched, are more likely to engage in moralistic evaluations rather than in strategic evaluations and thus expected to elicit reciprocal behavior.

By performing two studies, this research will address the gap in the current academic literature regarding the scarcity of empirical evidence on the topic of negative reciprocal behavior in a business context. The aim of this research is to contribute to the existing literature of stakeholder theory and reciprocal behavior theory by empirically testing the relationship between the projected image of a firm and its effect on stakeholder cooperation.

(10)

10

Secondly, this research contributes to the scientific literature by empirically testing the proposed theory of moralistic evaluations by Quintelier, Stoelhorst & Bridoux (n.p.). Finally, this research has also managerial relevance because it sheds light on how to incentivize primary stakeholders to go beyond their self-interest, by providing insights on how a projected profit-oriented image or a projected stakeholder-oriented image affects stakeholder reciprocity.

The remainder of this thesis starts with a chapter that reviews relevant literature and introduces the key constructs, followed by the research question. Subsequently, the chapter thereafter consists of the theoretical framework including development of the six hypotheses. The next section describes the methodology of this thesis. The fifth chapter comprises the results section and is followed by the last chapter including the discussion and conclusion.

(11)

11

2. Literature Review

2.1 Stakeholder Theory

According to the traditional, neo-classical economics and finance view, the ultimate objective of companies is to maximize risk-adjusted return on capital (Jensen & Meckling, 1976; Hahn, 2015). This implies that an adequate public policy framework ensures that profit-maximizing behavior of private firms is in line with over-arching societal objectives, leaving business with the imperative to maximize shareholder returns while respecting legal obligations (Friedman, 1970). This view, which is commonly referred to as shareholder view, will be considered as the profit-oriented view of a firm in this paper.

This profit-oriented view has been challenged by several management scholars, who argue that firms have a wider pro-social responsibility that goes beyond a neoclassical shareholder-orientation. These scholars recognize the strategic relevance of stakeholders in an increasingly complex world (e.g., Freeman 1994; Jones et al. 2007; Walsh 2005). A stakeholder can be defined as; “any group or individual who is affected by or can affect the

achievement of an organization’s objectives” (Freeman, 1984:46). Dunham, Freeman &

Liedtka (2006:25) add to this definition by describing stakeholders as “the group that the firm

needs in order to exist, specifically customers, suppliers, employees, financiers, and communities”.

This group, that is mentioned in the stakeholder definition, can be divided into primary and secondary stakeholders. Customers, suppliers, employees and financiers are primary stakeholders, because of their formal claim on the firm which is specified in contracts (Parmar et al., 2010). Secondary stakeholders can be defined as groups that influence or affect, or are influenced or affected by, a firm. An example of secondary stakeholders are communities. The main difference with primary stakeholders is that

(12)

12

secondary stakeholder do not engage in transactions with the firm and therefore are not essential for its survival (Clarkson, 1995). Secondary stakeholders are also more vulnerable due to the lack of a formal claim on the firm, which causes management to have no special duties pertaining to them, except for the moral duties, such as not harming them (Parmar et al., 2010; e.g. Carroll, 1993: 60; Gibson, 2000: 245). The recent problems regarding the Dakota Access oil pipeline in North-America is a case in point, because it shows the vulnerability of the Native American tribe as secondary stakeholder (Sylvester & Scheyder, 2017). This community is negatively affected by the actions of a firm, without having any formal claim on this firm.

2.2 Instrumental Stakeholder Theory

Donaldson and Preston (1995) have observed that the concepts of stakeholder, stakeholder management, and stakeholder theory have been used in distinctive ways and explained or criticized differently by a wide variety of authors. Their paper categorizes stakeholder theory in three main types; descriptive, instrumental and normative. The descriptive type of stakeholder theory reflects and explains the state of affairs of corporations and their stakeholders in the past, present, and future. The normative approach of stakeholder theory tries to clarify the function of, and provide guidance to the investor-owned corporation on the grounds of some underlying moral or philosophical principles. Lastly, instrumental stakeholder theory attempts to establish a relationship between stakeholder approaches and firm’s common objectives such as growth and profitability (Donaldson and Preston, 1995).

However, establishing a relationship between stakeholder approaches and firm’s objectives such as growth and profitability can be difficult due the complexities of today’s business. According to Parmar et al. (2010) the complexities of today’s business is: “the

(13)

13

and capitalism, and the problem of helping managers think about management such that the first two problems are addressed” (p.2). Instrumental stakeholder tries to explain these

challenges which are not confined to only management literature, but are in fact multidisciplinary and affect fields such as moral psychology, organizational behaviour and corporate social responsibility.

The general proposition of instrumental stakeholder theory based on the literature by Jones (1995) and Donaldson & Preston (1995) includes that the adherence to stakeholder principles and practices enable firms to develop a competitive advantage. Donaldson & Preston (1995) argue that: “corporations practicing stakeholder management will, other

things being equal, be relatively successful in conventional performance terms” (p.67). Put

differently, firms that contract with their stakeholders based on mutual trust and cooperation will achieve growth and profit objectives as well and possibly even outperform other approaches, such as the shareholder view. The study of Heskett and Kotter (1992) provides evidence for the general proposition of stakeholder theory. Heskett and Kotter found that highly successful companies such as; Wal-Mart, Hewlett-Packard and Dayton Hudson, although very diverse companies, share a stakeholder perspective.

2.3 Firm’s Orientation

The independent variable in this study is the underlying management orientation of a firm. This orientation is communicated through an image, which can be defined as the result of a thoughtfully coordinated attempt to communicate and convey the underlying management orientations of the firm (Quintelier, Stoelhorst, Bridoux, n.p.). The image of a firm depends on the impression that a group or an individual has of a firm. Regardless whether these impressions are true or false, real or imagined, images guide and shape behavior of individuals and groups (Barich and Kotler, 1991). Corporate communication is a critical link

(14)

14

between the underlying management orientation and the corporate image (Gray & Balmer, 1998), and includes images, logo’s, slogans, advertisements, as well as continuous interactions between the firm and its stakeholders, both internal and external (Quintelier, Stoelhorst, Bridoux, n.p.). The corporate image of a firm is a combination of what a firm communicates about itself and what other people think about the firm (Dowling, 1993). The more the communicated image of a firm corresponds with what people generally think about the firm, the stronger the corporate image of that firm (Dowling, 1993).

A distinction will be made in this research between two different images that a firm can communicate: a profit-oriented image and a stakeholder-oriented image. Projecting a profit-oriented image conveys the firm’s aim to maximize an abstract and extrinsic measure of utility, which is usually profit (Friedman, 1970; Rappaport, 1986; Smith, 2003). On the other hand, projecting a stakeholder-oriented image communicates that the firm is intrinsically motivated by norms regarding the firm’s impact on its stakeholders’ experiences, needs or desires (Quintelier, Stoelhorst, Bridoux, n.p.; Parmar et al., 2010; Freeman, 1984; Freeman, Wicks, & Parmar, 2004; Jones & Felps, 2013; Phillips, Freeman, & Wicks, 2003). In other words, profit-oriented firms are driven by profit maximization, while stakeholder-oriented firms aim for the collective well-being of all their stakeholders or society at large (Donaldson & Preston, 1995; Freeman, 1984; Friedman, 1970; Jones, 1995; Smith, 2003). Firms that project either a profit or stakeholder-oriented image communicate different goals and objectives to their stakeholders, resulting in different stakeholder behaviors towards the firm. The next paragraph will elaborate upon the different stakeholder behaviors.

2.4 Reciprocity

Agency theory and resource-based theory are widely accepted theories in the strategic management literature. Although their assumption that people are self-interested utility

(15)

15

maximizers (Bosse et al., 2009), and solely motivated by extrinsic gains and losses such as money or prestige, has been criticized by scholars in various disciplines due to lacking realism (Heath, 1999; Ghoshal, 2005). Extensive research in disciplines such as philosophy (Rawls, 2001; Becker, 2004), sociology (Cropanzano and Mitchell, 2005), psychology (Rabin, 1998), social psychology (Cialdini, 1984), and economics (Fehr and Gächter, 2000) shows that people are not solely motivated by self-interest (Bosse et al., 2009). This is being supported by literature reviews of Rabin (1998) and Fehr and Gächter (2000), which conclude that a growing number of studies in the economic literature find that a vast majority deviates from this assumption (Bosse et al., 2009).

Recently, a more realistic assumption of human behavior, that of reciprocity, is successfully incorporated into economic theory (Quintelier, Stoelhorst & Bridoux, n.p.; Camerer & Thaler, 1995; Kahneman, Knetsch, & Thaler, 1986). Reciprocity can best be explained as a response to others behavior’ or actions. Behavior or an action that is being perceived as fair will be rewarded by reciprocal individuals to a larger extent than predicted by self-interested models. On the contrary, when the actions or behavior of other parties is perceived as unfair, reciprocal individuals seek to punish the other party to a larger extent than predicted by self-interested models (Quintelier, Stoelhorst & Bridoux, n.p.; Bosse & Phillips, 2014; Bosse et al., 2009; Bridoux, Coeurderoy, & Durand, 2011; Fehr & Gächter, 2000; Hahn, 2015; van Dijke et al., 2012). The responses to both fair and unfair behavior or actions should come at a cost to the individual itself in order to be reciprocal.

Negative reciprocal behavior of primary stakeholders is the dependent variable in this study. More specifically, negative reciprocal behavior is measured in terms of negative word of mouth intentions, which can be regarded as costly punishment (Quintelier, Stoelhorst & Bridoux, n.p.). It is important to note that this study will conceive reciprocity in terms of stakeholder behavior rather than the individual personality trait.

(16)

16

2.5 Moralistic & Strategic Evaluations

Recent studies suggest that firms projecting a stakeholder image will increase the likelihood of both stakeholder reward (Du, Bhattacharya, & Sen, 2007; Sen & Bhattacharya, 2001; Zhang, Wang, & Fung, 2014) as well as stakeholder punishment (Ashforth & Gibbs, 1990; Briscoe et al., 2014). Based on moral psychology literature, Quintelier, Stoelhorst & Bridoux (n.p.) propose a theory that offers a possible explanation, which argues that stakeholders evaluate a firm on the basis of two distinctive cognitive frames: moralistic and strategic evaluations. Depending on the cognitive frame, individuals will elicit a certain degree of cooperation (reward) or noncooperation (punishment) towards the firm.

Individuals that engage in strategic evaluations determine whether an act is good or bad, or if an actor is competent or incompetent. The act or actor is evaluated in relation to the extrinsic payoff of the evaluator. Thus, when the extrinsic payoff of the evaluator increases it is evaluated as strategically good, however when it decreases, it is evaluated as strategically bad (Declerck et al., 2013). So strategic evaluations motivate behavior that relates to maximization of one’s own extrinsic payoff (Lyons & Mehta, 1997). Firms projecting a profit-oriented image are therefore likely to elicit these strategic evaluations.

Conversely, firms projecting a stakeholder-oriented image are likely to elicit moralistic evaluations. Moralistic evaluations determine whether an act or actor is right and complies with the salient moral norm, or if the act(or) is wrong and transgresses the salient moral norm. The salient moral norm is based on for example: rules, laws and consensus (Treviño et al., 2006). Compliance with the salient moral norm increases reward, while transgression of the moral norm increases punishment. Stakeholders that evaluate an event or action as morally wrong, instead of strategically bad, will increase punishment towards a firm. Wyatz et al. (2010) found that when firms are more anthropomorphized, thus triggering

(17)

17

more human like traits such as emotions, they were held more accountable for their behavior compared to less anthropomorphized firms. Stakeholder-oriented firms thus trigger a more social context compared to profit-oriented firms, which causes stakeholders to assess firm on the basis of an emotional response instead of a deliberate rational reasoning process.

This explains why firms with a stakeholder orientation elicit more punishment compared to firms with a profit orientation. Collective engagement in moralistic evaluations therefore can bring about the ability of sustaining within group cooperation, and hence, increasing norm-compliant behavior within this group (Boyd & Richerson, 1992; DeScioli & Kurzban, 2009; Haidt et al., 2009; Koleva et al., 2012).

2.6 Emotions

Moralistic evaluations can be either positive or negative. Positive moralistic evaluations include for example the assessment of certain behavior and whether this is in compliance with the salient moral norm (Boone et al., 2008; Kurzban & Aktipis, 2007). Positive moralistic evaluations are related to emotional mental processes and increase feelings and emotions such as warmth, trust and moral elevation (Aquino et al., 2011). By triggering emotions and morality, the willingness to benefit to a social cause at cost to themselves increases. For example the study of Podsakoff et al. (2000) shows that in a business context positive moralistic evaluations increase organizational citizenship behavior (OCB), which can be regarded as positive reciprocity.

Whereas assessing if behavior is wrong, or an actor blameworthy and whether this transgresses a salient moral norm, describes the process of negative moralistic evaluations (Boone et al., 2008; Kurzban & Aktipis, 2007). Emotions and sentiments that correspond with negative moralistic evaluations are for example; threat (Declerck et al., 2013), disgust (Wheatley & Haidt, 2005), anger and contempt (Hutcherson & Gross, 2011; Rozin, Lowery,

(18)

18

Imada, & Haidt, 1999). Similarly, Bisckei et al. (2016) argue that emotions that correspond to negative reciprocity are anger-like emotions and include anger, irritation and contempt. Thus similar to positive moralistic evaluations, when individuals are triggered in moralistic terms when assessing behavior, it increases the likelihood of punishment (Skitka, 2002; Skitka et al., 2005).

2.7 Research Gap and Research Question

The current scientific literature on the reciprocal behavior of stakeholders mainly consists of theory building, but lacks empirical testing. This thesis therefore aims to contribute to the recent stream of literature on reciprocal stakeholder behavior by empirically testing the effects of a firm’s orientation on stakeholder punishment. More specifically, this thesis investigates whether firms projecting a stakeholder image elicit more stakeholder punishment than firms projecting a profit-oriented image in the case of a human resource policy violation. Resulting in the following research question:

“How does the projected image of a firm affect negative reciprocal behavior of primary stakeholders in the case of a negative event?”

The next chapter elaborates on the theoretical framework and includes the development of the hypotheses of the two studies in order to answer the main research question.

(19)

19

3. Theoretical Framework

The theoretical framework of this thesis is divided into two sections. The first section describes the first study including the first three hypotheses, and the second section describes the second study including the last three hypotheses. Both sections are structured similarly and consist of three paragraphs, starting with a conceptual model including a visual representation of the relationship between all the variables. The second paragraph embodies the development of the hypotheses regarding the direct effect between the independent and dependent variable. The last paragraph of both sections include the development of the hypotheses regarding the mediating variables.

3.1 Study One: Profit Orientation versus Stakeholder Orientation

3.1.1 Conceptual Model

Figure 1 shows a conceptual representation of the relationship between the multi-categorical independent variable (IV) and the dependent variable (DV). The first hypothesis (H1a, H1b, H1c) represents the direct effect between a firm’s orientation and negative word of mouth. The second and third hypotheses (H2 & H3) include mediation and propose the mediating effect of fairness perceptions (M1) and anger-like emotions (M2).

(20)

20

3.1.2 Firm’s Orientation

The image a firm projects conveys information about the management orientation and tells something about the firm’s goals and objectives. This can either be a profit-oriented image or a stakeholder-oriented image. Both images have different effects on stakeholder behavior, that can be distinguished between self-interested behavior and reciprocal behavior. Profit-oriented firms are driven by profit maximization (Milton Friedman, 1970; Rappaport, 1986; Smith, 2003) and are therefore expected to motivate strategic evaluations among its stakeholders. Strategic evaluations are the result of deliberate reasoning processes, meaning that individuals actively determine whether an act is good or bad, or if an actor is competent or effective. The act or actor are evaluated in relation to the extrinsic payoff of the evaluator. An increase in the extrinsic payoff is evaluated as strategically good and a decrease in the extrinsic payoff is evaluated as strategically bad (Declerck et al., 2013). Thus, strategic evaluations motivate behavior that relates to maximization of one’s own extrinsic payoff (Lyons & Mehta, 1997).

On the other hand are stakeholder-oriented firms characterized by the aim for the collective well-being of all their stakeholders or society at large (Donaldson & Preston, 1995; Freeman, 1984; Friedman, 1970; Jones, 1995; Smith, 2003). Stakeholder-oriented firms are more likely to trigger moralistic evaluations instead of strategic evaluations. Moralistic evaluations include the assessment of an act or behavior as being right or wrong and if an actor is morally responsible for this act or behavior (Quintelier, Stoelhorst & Bridoux, n.p.). Moral psychologists argue that the strength of these moralistic evaluations depend on two conditions; whether an act is caused by an intentional agent, and secondly, if an act impacts a sentient being (Gray et al., 2012). Intentional agency is “ the perceived capacity to intend and

to act (e.g., self-control, judgment, communication, thought, and memory)” (Gray et al.,

(21)

21

happiness (Gray et al., 2012: p.103). The stronger the projected stakeholder orientation, the more likely stakeholders will perceive sentience and intentional agency.

The projected image of a firm communicates the underlying management orientation. In the case of a stakeholder orientation conveys this image the aim and objective of a firm to care about its stakeholders and the relating stakeholder issues (Maignan & Ferrell, 2004). Intentional agency should be included in the corporate image, as well as the importance of caring about the stakeholders and their relating issues because this acknowledges that stakeholders are sentient beings. Subsequently, firms that project a stakeholder orientation portray their behavior as an expression of their core values (Maignan & Ralston, 2002) and the presence of core values evoke perceptions of intentional agency (Waytz et al., 2010). Thus when firms project a stronger stakeholder orientation it is likely that stakeholders perceive more intentional agency and sentience. Consequently, the more intentional agency and sentience the stronger an act or behavior is evaluated in terms of morality (Gray et al., 2012). A firm acting in the best interest of all its stakeholders, can expect reciprocal stakeholder behavior. On the other hand, the less intentional agency and sentience, the less reciprocal behavior occurs. A firm that is acting solely in the best interest of only the shareholders is likely to expect self-interested behavior among their stakeholders. Following this argumentation the following hypothesis emerges:

Hypothesis 1a: Profit-oriented firms elicit less punishment than firms without a specified orientation

Hypothesis 1b: Stakeholder-oriented firms elicit more stakeholder punishment than firms without a specified orientation

Hypothesis 1c: Stakeholder-oriented firms elicit more stakeholder punishment than profit-oriented firms

(22)

22

3.1.3 Mediation

Instrumental stakeholder theory argues that treating stakeholders well, meaning fairly, contributes to firm performance (Donaldson and Preston, 1995; Freeman, 1984; Harrison et al., 2010; Jones, 1995). Consequently, Bosse et al. (2009) argue that the rent created by a firm is directly associated with the perceptions of fairness among its stakeholders. These perceptions of fairness are based on distributional justice, procedural justice and interactional justice (Bosse et al., 2009).

Distributional justice refers to the perceptions of fairness regarding the allocation of material gains (Bosse et al., 2009). Studies of Rabin (1993) and Nelson (2001) explain this phenomenon and argue that when actors receive a material outcome, which they perceive as fair (unfair), they subsequently are prone to exhibit positive (negative) reciprocity toward the other party by exerting more (less) effort. In addition to distributional justice, employees also value procedural justice. Procedural justice can be explained as the need of employees to be involved in the decision-making processes of their employers (Leventhal et al., 1980). Employees desire consistent decision-making processes that are free from personal biases and based on accurate information, also these processes need to allow for correction of bad decisions that comply with rampant standards of ethics and morality (Bosse et al., 2009). Employees will perceive this as procedural justice when employers comply with these decision-making process criteria (Colquitt et al., 2001). Finally, interactional justice involves the manner in which actors treat each other. For example, if employees feel that managers treat them respectfully by actively listening to their concerns and empathize with their points of view, they are being considered as interactionally just (Cropanzano et al., 2007).

A firm that is being perceived as fair according to the above mentioned types of justice will elicit a higher incentive for stakeholders to contribute positive effort to the firm, compared to a firm that is only perceived as fair on one of the dimensions of justice (Bosse et

(23)

23

al., 2009). Hence, a dearth in one dimension of justice may cancel out positive traits stemming from other dimensions of justice. Conversely, stakeholder punishment as result of perceptions of unfairness in one type of justice may be alleviated by stakeholder reward stemming from perceived fairness in other types of justice (Bosse et al., 2009).

The hypothetical case that is being described in the vignettes of this research is a negative event which transgresses a salient moral norm. In line with the aforementioned theory it is expected that stakeholder-oriented firms are morally evaluated and thus in this case are being perceived as more unfair compared to profit-oriented firms. A transgression of the salient moral norm can cause stakeholders to engage in negative moralistic evaluations, because they assess this certain behavior as morally wrong (Boone et al., 2008; Kurzban & Aktipis, 2007). Stakeholders who feel that they have been treated unfairly have according to Loewestein (2000) increased preferences to engage in stakeholder punishment. Profit-oriented firms are evaluated strategically and thus perceived as less unfair compared to stakeholder-oriented firms. Fairness perceptions are therefore expected to mediate the relationship between the orientation of a firm and stakeholder punishment, resulting in the following hypothesis:

Hypothesis 2: Fairness perceptions mediate the relationship between the orientation of a firm and stakeholder punishment

In addition to fairness perceptions are emotions also considered as proximal causes of behavior (Elster, 1998; Martinez et al., 2011; Zeelenberg & Pieters, 2006). Behavior that is being assessed as morally wrong results in negative reciprocal behavior. The emotions that correspond with this kind of behavior can be described as anger-like emotions, including; anger, irritation, contempt (Bicskei et al., 2016), as well as moral disgust (Hutcherson and

(24)

24

Gross, 2011). A transgression of the salient moral norm is likely to increase anger-like emotions. Stakeholder-oriented firms are evaluated morally and profit-oriented firms are evaluated strategically, therefore is a stakeholder orientation expected to elicit more anger-like emotions than a profit orientation. Anger-anger-like emotions are thus expected to mediate the relationship between the orientation of a firm and stakeholder punishment, because an increase in anger-like emotions results in an increase in stakeholder punishment. Resulting in the following hypothesis:

Hypothesis 3: Anger-like emotions mediate the relationship between the orientation of a firm and stakeholder punishment

Both mediators are expected to mediate the main relationship parallel, meaning that fairness perceptions and anger-like emotions simultaneously mediate the relationship between the orientation of a firm and stakeholder punishment. Although these variables may correlate with each other, they are expected to separately mediate the main relationship between the independent and dependent variable. Put differently, the variables fairness perceptions and anger-like emotions individually explain the relationship between the orientation of a firm and stakeholder punishment.

(25)

25

3.2 Study Two: Perceived Cues of Being Watched

3.2.1 Conceptual Model

Figure 2 shows a conceptual representation of the relationships between the variables in the second study. The independent variable (IV) in this study is dichotomous, including two conditions with either perceived cues of being watched in the vignette or without these perceived cues of being watched. Similar to the first study, is the dependent variable (DV), negative word of mouth. This main relationship between the IV and DV is represented by hypothesis 4. Furthermore, hypothesis 5 and 6 represent the mediating effects of both fairness (M1) perceptions and anger-like emotions (M2).

Figure 2 – Conceptual Model Study Two

3.2.2 Perceived Cues of Being Watched

The underlying mechanism that is likely to explain increased stakeholder reciprocity in the case of a projected shareholder orientation, namely moralistic evaluations, is expected to be triggered by a social context. Moralistic evaluations as opposed to strategic evaluations are not generated by a deliberate reasoning process, but are generated by faster and more emotional processes which are less controlled and more intuitive (Boone et al., 2008; Evans 2008; Evans & Stankovich, 2013; Kahneman, 2003). An example would be a social context

(26)

26

in which an individual bumps into someone in a crowded street, this unexpected event happens so quickly that there is no time to respond through a deliberate reasoning process, but rather through a less controlled emotional process.

Milinski et al. (2002a,b), Wedekind & Braithwaite (2002) and Barclay (2004) have performed laboratory experiments which tried to artificially replicate a social context and found that individuals increase their levels of cooperation when they know that their behavior is being watched by others. Additionally, in deciding how to interact with others, reputational information is used (Milinski et al. 2002a,b; Wedekind & Braithwaite 2002; Barclay 2004). According to the more recent studies of Haley & Fessler (2005) and Burnham & Hare (2007), subjects in experiments responded to subtle cues of being watched, being the presence of eye-like spots on the background of the computer on which they completed a task. This was even the case when the participants were told they were anonymous (Burnham & Hare, 2007). However, these studies only demonstrate this effect in artificial laboratory scenarios.

The study of Bateson, Nettle and Roberts (2006) investigates whether subtle cues of being watched also affect individuals in a real-world setting, using their own money when donating to a honesty box. Their studies show that, in a real-world context indeed an image of a pair of eyes induces the perception of being watched and therefore increases cooperative behavior, in this case by a significant increase in the contributions of participants. A possible explanation is the human perceptual system, which contains neurons that respond selectively to stimuli involving faces and eyes (Emery 2000; Haxby et al. 2000), “it is therefore possible

that the images exerted an automatic and unconscious effect on the participants’ perception that they were being watched” (Bateson, Nettle & Roberts, 2006, p.413).

Individuals that perceive cues of being watched experience a situation or context as more social compared to a context without these cues of being watched. A social context is

(27)

27

likely to increase the strength of moralistic evaluations, consequently increasing the level of reciprocal behavior. This would suggest that individuals in a social context are more likely to elicit reciprocal behavior as opposed to individuals in a non-social context. Resulting in the following hypothesis:

Hypothesis 4: Perceived cues of being watched increases stakeholder punishment compared to the absence of these cues

3.2.3 Mediation

The independent variable in this second study is different from the first study, however the dependent variable remained the same. Therefore, similarly to the first study, are fairness perceptions and anger-like emotions both expected to mediate the main relationship between the perceived cues of being watched and stakeholder punishment. Leading to the following hypotheses:

Hypothesis 5: Fairness perceptions mediate the relationship between the perceived cues of being watched and stakeholder punishment

Hypothesis 6: Anger-like emotions mediate the relationship between the orientation of a firm and stakeholder punishment

(28)

28

4. Research Design

4.1 Research Method

This quantitative research was carried out using a vignette study. Vignette studies consist of two elements; (i) a vignette experiment, and (ii) a traditional survey. Atzmüller and Steiner (2010: p.1) describe a vignette as “a short, carefully constructed description of a person,

object, or situation, representing a systematic combination of characteristics”. Compared to

the traditional survey method, this special vignette design allows for simultaneous representation of contextual elements resulting in more realistic scenarios for the respondents (Atzmüller & Steiner, 2010). This combination of a classical experiment and a survey methodology counterbalances the weaknesses of both techniques and is therefore a promising research method for investigating mental processes and behavioral intentions (Atzmüller & Steiner, 2010; Aguinis & Bradley, 2014; Finch, 1987).

The design of vignette studies is comparable to that of experimental designs and can be distinguished between three general categories: (i) within-subjects design, (ii) between-subjects, and (iii) mixed designs (Atzmüller & Steiner, 2010). This thesis used a between subject design, because each respondent randomly received one of the four treatments (Atzmüller & Steiner, 2010). Drawing conclusions from this experiment required at least 160-200 respondents, because in order to measure effects a minimum of 40-50 respondents per treatment is needed.

4.2 Sampling Technique

All vignettes in this thesis include a description of a fictive company named Alpha. Alpha is depicted as a grocery retailer that sells products that you buy on a weekly basis which provides good value for money and satisfying service. Therefore the population of interest for this study is anyone that occasionally buys groceries at a retailer. Students are thus perfectly

(29)

29

suited as main target audience. The sampling method that was used conducting research was a non-probability sampling, more specifically a convenience sampling method was used since the population is large and the sampling frame was not available (Saunders et al., 2009). Respondents were approached online, using e-mail and social media. The response rate was difficult to predict, but the goal was to collect as many respondents as possible, with an absolute minimum of 160 respondents.

Some limitations arise regarding this sample frame. First of all, the sample group is relatively homogenous in terms of age, nationality and level of education. The homogeneity of the sample reduces the generalizability to the entire population. Secondly, the majority of the respondents is native Dutch, whereas the vignettes and survey were written in English. This could have caused a language barrier which might have led to difficulties and misunderstandings regarding the content of the study.

4.3 Vignettes

As mentioned before, Alpha is a grocery retailer that sells products that you buy on a weekly basis which provides good value for money and satisfying service. Alpha’s human resource policy includes the following rule: “If an employee has worked 20 hours' overtime in one month, its floor manager should give the employee a raise.” The vignette continues and describes that according to a reputable independent research organization Alpha often fails to implement this policy. The vignettes are further elaborated on in the next paragraph. See Appendix 8.1 for the full vignettes and survey.

(30)

30

4.4 Measurements

4.4.1 Independent Variables

The independent variable in the first study is the orientation of a firm. In the second study is the orientation of a firm replaced by the perceived cues of being watched as independent variable. The independent variables were manipulated in order to test for its effect on the dependent variable, stakeholder punishment. There are four different vignettes used in this research to manipulate the independent variable (See Table 1 & 2). Three different descriptions regarding the underlying management orientation were presented in the vignette. The fourth vignette however did not include a textual manipulation, but has an experimental setting including an image with a recognizable pair of eyes in the background of the survey (See Appendix 8.1). This fourth vignette was used in the second study and compared to the control group.

Table 1 – Vignette Description Study One

Table 2 – Vignette Description Study Two

# Vignette Manipulation

1 Control group Baseline description

2 Stakeholder-oriented Stakeholder-oriented characteristics in the company description 3 Profit-oriented Profit-oriented characteristics in the company description

# Vignette Manipulation

(31)

31

4.4.2 Dependent Variable

Stakeholder punishment as a form of negative reciprocity can be measured by the likelihood of a stakeholder to engage in negative word of mouth (WoM). The scale that measures WoM includes five items (See Table 3) on a 7-point Likert-Scale and is adopted from Skarmeas & Leonidou (2013). The participants first receive randomly one of the four vignettes and accordingly are asked to rate their intentions to engage in negative word of mouth on a scale ranging from (1) completely disagree, to (7) completely agree.

Table 3 – Word of Mouth Items

# Item

1 I would like to say negative things about Alpha to others 2 I would complain about Alpha to friends and neighbors

3 If others would ask me about Alpha, I would discourage them to go there

4 I would be tempted to give Alpha a negative review (for instance on social media) 5 I would actively boycott Alpha

4.4.3 Mediating Variables

One question is asked to measure the fairness perceptions: ‘How fair or unfair do you

consider Alpha’s practices regarding the implementation of their human resource policy?’.

This question has a 7-point Likert-Scale ranging from completely unfair (1), to completely fair (7).

Secondly, a self-reported measure for emotions is used. This self-reported measure consist of four questions and assesses the intensity of anger-like emotions regarding the described violation of the human resource policy on a 7-point Likert-Scale (1 = not at all, 7 = very strong). This measure is based on Bicskei et al. (2016) and includes anger-like emotions (ALEs) such as; anger, irritation, contempt. Moral disgust is added to these three emotions and is adapted from Hutcherson and Gross (2011).

(32)

32

4.4.5 Attention, Manipulation and Realism Check

First of all, an attention check was added to the survey of this thesis to make sure that all participants have read and understood the description in the vignette. This attention check includes a multiple choice question regarding what type of company Alpha is; a car rental, grocery store, or a consumer electronics store.

Secondly, manipulation checks were added in the profit-oriented and stakeholder-oriented vignettes (See Table 4). Manipulation checks are crucial in safeguarding construct validity (Cozby, 2009), and can be described as “a way of ensuring that an experiment

actually has been conducted (i.e., that the independent variable has been effectively manipulated)” (Sansone, Morf & Panter 2004, p. 244). The manipulation checks in this thesis

investigate whether the treatment has an effect on the relevant theoretical causal construct. In other words, does the participant understand and interpret the manipulation of the vignette as it is intended by the researcher. The manipulation checks were placed on the pages after the vignettes. Due to the added restrictions in Qualtrics it was not possible to go back and read the vignettes again when the respondents had to fill in the questions including the manipulation checks.

In addition to the attention and manipulation check, a realism check was performed (See Table 5). The realism check includes two items that evaluate if the described vignettes are realistic to the participants. These items are measured on a 7-point Likert-Scale ranging from (1) strongly disagree, to (7) strongly agree.

(33)

33

Table 4 – Manipulation Checks (Stakeholder Orientation & Profit Orientation)

Table 5 – Realism Check

# Item

1 The situation described was realistic

2 I had no difficulty imagining this situation

4.4.6 Control Variables

A control variable is any variable that can have an influence on the relationship between the dependent and independent variables. The validity of the results can be increased by testing the control variables, because the extraneous effects on the relationship between the dependent and independent variable can be ruled out (Saunders et al., 2009). See Table 6 for all items regarding the control variables.

The first variable that this study controls for is social value orientation (SVO). The (SVO) of the respondents is measured by the SVO-slider from Murphy et al. (2011). The first six primary items of this scale are used, because this study is only interested in differentiating

# Item Answer

1 Alpha is committed to: 1 = Improving its financial performance, because Alpha believes this is necessary to be a successful business 2 = Improving its stakeholders’ welfare, because Alpha believes this is necessary to be a successful business 3 = Improving its stakeholders’ welfare, because Alpha believes this is the morally right thing to do

2 Please indicate on a scale from 1 to 7 to what extent you think Alpha prioritizes its

commitments.

1 = welfare of the stakeholders (customers, employees, suppliers, or the community)

4 = both, or balance their commitments

7 = firm’s performance (firm’s profit, market share, or prestige)

3 Please indicate on a scale from 1 to 7 to what extent you think Alpha has mostly long-term or short-term stakeholder relationships or both.

1 = long-term relationships 4 = both

(34)

34

between prosocial and self-interested people. Each item consists of a variety of options regarding the allocation of money between the respondent and an unknown person. The respondent is asked to select the option with the allocation that he or she prefers. These six items enable the categorization of the respondents into four categories; altruists, pro-socials, individualists, and competitors (Murphy et al., 2011).

This research also controls for age and gender. This is based on the research of Van Lange et al. (1997) who found that younger people in general tend to behave competitive and individualistic, however when people get older their behavior becomes more pro-social and other-interested. Secondly, the differences in behavior between men and woman can be ascribed to the level of other-orientation. Men are considered to be more self-oriented, whereas women are more socially-orientated (Van Lange et al., 1997; Eckel & Grossman, 1998).

Moreover, nationality is used as control variable, due to the differences that arise between countries in terms of two dimensions: individualism and masculinity. Firstly, individualistic societies can be characterized as: “everybody is supposed to take care of

themselves” and “involvement with organizations is calculative” (Hofstede, 1980, p. 48). On

the contrary, in a collectivistic society “people are born in extended families or clans who

protect them in exchange for loyalty” and “involvement with organizations is moral”

(Hofstede, 1980, p. 48). Secondly, masculine societies can be described as materialistic, performance is what counts and independence is the ideal situation. On the other hand, feminine societies value the quality of life as well as the people and environment and interdependence is the ideal situation (Hofstede, 1980). Individualistic and/or masculine countries are more likely to adopt strategic evaluations rather than moralistic evaluations and collectivistic and/or feminine countries vice versa. Differences in nationality can therefore have an impact on the likelihood of reciprocal behavior as well as fairness perceptions.

(35)

35

Furthermore, the area of study is another control variable. Evidence from previous literature suggests that business and economics students, compared to students from other areas, are less altruistic and less cooperative in various experimental settings (e.g. public good games (Marwell and Ames, 1981), prisoner’s dilemma (Frank et al., 1993) and ultimate bargaining (Carter and Irons, 1991; Kahneman et al., 1986)).

Table 6 – Control Variables

Control Variable Item

SVO 6 items with each 9 options regarding resource

allocation Gender Male Female Age Under 18 18 – 24 25 – 34 35 – 44 45 – 54 55 – 64 65 – 74 75 – 84 85 or older Nationality Dutch North-Western European Eastern European Southern European Asian African North American South American Australian/New Zealand

Education Primary School Secondary School

Vocational Education (MBO) College (HBO)

University Other

Study Area Business/Economics Law

(36)

36

5. Results

This chapter will discuss the results of the survey and consist of three parts. The first part starts with the preliminary steps of preparing the data for analysis, including; data cleaning, scale reliability testing, descriptive analysis and normality testing. Followed by the second part, including hypotheses testing for study one. The last part includes the hypotheses testing for study two. All the statistical analysis of the data were executed with SPSS.

5.1 Data Cleaning

A total of 240 participants started the survey. Only the data of the participants who completely finished the survey will be considered. Based on this criterion 13 participants were eliminated, leaving 227 participants whose survey was analyzed. The corresponding dropout rate is therefore 5,7 %. Moreover, the participants who failed the attention check will also be excluded. Based on this criterion 19 participants were eliminated, resulting in 208 analyzable surveys.

5.2 Reliability

This research includes three variables that use multiple questions to measure a specific construct: the dependent variable Word of Mouth, the mediator Anger-like emotions and the realism check. In order to test the internal reliability of these questions the Cronbach’s alpha was measured. The internal consistency is investigated as a determinant of the scale reliability, an alpha score equal or above .7 is considered to be an acceptable score (Saunders et al., 2009). The results of the reliability test can be found in table 7.

The dependent variable Word of Mouth is measured by five questions, the combined Cronbach’s alpha is .869. The mediating variable Anger-like emotions is measured by four questions, the combined Cronbach’s alpha is .855. The variables WoM and ALE are therefore considered as reliable, because the alpha is >.7. Lastly, the realism check is measured by two

(37)

37

questions with a combined Cronbach’s alpha of .684. This score is marginally acceptable because this Cronbach’s alpha is very close to .7 and will therefore be considered as reliable in this research.

5.3 Descriptive Statistics

The data was collected by making use of the online software of Qualtrics. No corrections for missing data were necessary because respondents could not proceed in the survey without answering each question, due to the settings in Qualtrics. From the 208 remaining surveys, 54 answered the control vignette; 46 the profit-oriented vignette; 57 the stakeholder-oriented vignette; and 51 answered the experiment vignette. The respondents can be categorized in terms of their social value orientation, 46 (22,1%) respondents are individualistic and the other 162 (77,9%) respondents can be identified as prosocials.

Considering gender and age, the survey was completed by 120 (57,7%) women and 88 (42,3%) men. The majority of the respondents is between 18 – 24 (67,3%) and 25 – 34 (26,9%) years old, and only a small minority (5,8%) is older than 34 years. The nationality of the respondents is unevenly distributed, because 185 (88,9%) of the respondents are Dutch or from Northern Europe 10 (4,8%). Furthermore the level of education, the majority of the respondents, 150 (72,1%) completed an university degree, secondly 39 (18,8%) respondents completed HBO. Moreover, about half of the respondents, 111 (53,4%), studied Business/Economics. The other half, 92 (44,2%) respondents selected ‘other’ and only 5 (2,4%) respondents studied law. See Table 7 for the descriptive statistics of the categorical variables.

(38)

38

Table 7 – Descriptive Statistics of the Categorical Data

Variable Level Frequency % Cumulative %

Vignette Control group 54 26 26

Experiment 51 24.5 50.5 Stakeholder 57 27.4 77.9 Profit 46 22.1 100 SVO Individualistic 46 22.1 22.1 Prosocial 162 77.9 100 Gender Male 88 42.3 42.3 Female 120 57.7 100 Age 18 - 24 140 67.3 67.3 25 - 34 56 26.9 94.2 35 - 44 1 0.5 94.7 45 - 54 7 3.4 98.1 55 - 64 3 1.4 99.5 85 or older 1 0.5 100 Nationality Dutch 185 88.9 88.9 North-Western Europe 10 4.8 93.8 Eastern Europe 2 1 94.7 Southern Europe 6 2.9 97.6 Asian 3 1.4 99 North American 1 0.5 99.5 South American 1 0.5 100

Education Secondary School 16 7.7 7.7

Vocational Education (MBO) 1 0.5 8.2

College (HBO) 39 18.8 26.9

University 150 72.1 99

Other 2 1 100

Study Area Business/Economics 111 53.4 53.4

Law 5 2.4 55.8

Other 92 44.2 100

5.4 Normality Analysis

To ensure that the variables are normally distributed normality tests were conducted for: the dependent variable WoM, and the mediating variables ALE and Fairness Perceptions. The independent variable is manipulated in the vignettes and therefore requires no normality testing. Moreover, the skewness and kurtosis of the aforementioned variables will be discussed in this paragraph. See Appendix 8.2 for the normality tests.

(39)

39

The mean of the dependent variable WoM is 3.84, with a 5% trimmed mean of 3.85, showing that the extreme values do not have a substantial influence on the mean. The skewness and kurtosis are respectively -.16 and -.13. A perfect normal distribution has skewness and kurtosis of zero (Saunders et al., 2009). These scores are close to zero which indicates that the values of the dependent variable WoM are normally distributed.

The mean of the first mediating variable, ALE, is 4.29 and a 5% trimmed mean of 4.32. This indicates that the extreme values do not have a substantial influence on the mean. The skewness and kurtosis are respectively -.59 and .35. The skewness of ALE is negative, however in a fairly acceptable range. Indicating a slightly negatively skewed distribution, put differently, the distribution’s tail on the left side longer than on the right side. Furthermore, the kurtosis of .35 indicates that the distribution is slightly leptokurtic, meaning that the distribution is more peak-topped than a normal distribution.

The second mediating variable, fairness perception, has a mean of 2.58 and a 5% trimmed mean of 2.49. Indicating that the extreme values have some but not substantial influence on the mean. The skewness and kurtosis are respectively .98 and .88. These scores are within the acceptable range between -1 and 1. Although these scores are within the acceptable range it shows that the distribution is positively skewed and has a fairly positive kurtosis distribution. Put differently, the distribution’s tail on the right side is longer than on the left side and the distribution is peak-topped.

(40)

40

5.4 Realism & Manipulation Analysis

The last step in the data preparation consists of analyzing the realism and manipulation checks. Two questions were asked in order to check if the vignettes were perceived as realistic by the participants. The questions were measured on a 7-point Likert scale. Moreover, the descriptive statistics (See Table 8) of the four groups combined (M = 5.54; SD = 1.03) show that the vignettes overall were found realistic.

Table 8 – Descriptive Statistics Realism Check

Vignette N M SD Control 54 5.50 1.09 Experiment 51 5.62 1.05 Stakeholder 57 5.58 .99 Profit 46 5.45 1.00 Total 208 5.54 1.03

Secondly an one-way ANOVA was conducted to test if the results differ significantly. The results of the one-way ANOVA, F(3,204) = .278, p =.841, show that the two questions regarding the realism of the vignettes show no statistical significant difference between the four groups.

Moreover, three items were used to determine if the manipulation of the independent variable was successfully performed. The manipulation was only included in the stakeholder- and profit group, therefore was chosen to analyse the items with the use of independent samples t-tests. The first manipulation check consisted of a multiple choice question with three options (See Table 9). The first option represented a pure profit orientation, the third option represented a pure stakeholder orientation and the middle option represented a mix between a profit orientation and a stakeholder orientation.

(41)

41

Table 9 – Type of Vignette * Manipulation 1 Cross Tabulation

Profit Mix Stakeholder Total

Vignette Stakeholder Count 16 19 22 57

% within Type of Vignette 28,1% 33,3% 38,6% 100,0%

Profit Count 38 8 0 46

Total Count 54 27 22 103

A chi-square test of independence was performed to examine the relation between the type of vignette (stakeholder vs. profit) and the first manipulation check. The relation between these variables was found significant, χ2(2, N = 103) = 34.67, p<.001, φ = .58. These results indicate that the respondents in the stakeholder vignette answered the manipulation question significantly different than the respondents in the profit vignette, showing that the manipulation was performed correctly. Because the respondents with the profit vignette (M = 1.17; SD = .38) selected option 1 more often than the respondents with a stakeholder vignette (M = 2.11; SD = .82).

The second manipulation item asked the respondents about how they thought Alpha prioritized its commitments. Alpha prioritizes (1) ‘the welfare of the stakeholders’, (7) ‘firm’s performance’ or (4) ‘balances its commitments’. A low score represents a more stakeholder orientation and a high score represents a more profit orientation. The third manipulation item was structured similarly as the second manipulation item, however asked whether the participant thought Alpha had mostly (1) ‘long-term’ relationships, (7) ‘short-term’ relationship, or (4) ‘both’.

First the Levene’s test for homogeneity of the variances was performed for both items. This test was not found to be significant, with respectively, F(1,101) = .342, p = .560,

The impact of a firm’s orientation on negative reciprocal behavior of primary stakeholders