Final thesis Executive Program in Management Studies – Strategy track
What is the effect of
firm-orientation on stakeholder’s
punishment and reward
behavior?
Student: MSc. S. Bosch
Statement of Originality
This document is written by Student MSc. Sanne Bosch, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented
in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
Stakeholder theory is a widely discussed framework first introduced by Freeman (1979). All parties related to a corporation that affect or are affected by the actions of a corporation are stakeholders. According to stakeholder theory, a stakeholder-oriented firm bases their conduct on and prioritizes fairness towards stakeholders. Profit maximizing firms prioritize their performance. How a firm displays its orientation leads to stakeholders’ perception of their goals. Stakeholder perception of these goals may elicit a stakeholder response, positive (reward behavior) or negative (punishment behavior), which may impact the business (gains or losses).
The behavior of stakeholders is puzzling and this research will attempt to create an understanding of this phenomenon. To investigate this, a paradigm was set up where four groups of subjects were exposed to a unique combination of two different vignettes. One described a stakeholder-oriented company and the other described a profit-oriented company, after which some questions were posed with regards to their intended punishment and reward behavior. Profit orientation seemed to elicit more punishment behavior and less reward behavior. The emotional responses were not different, except for negative responses after good behavior, which was significantly higher in the profit oriented group as opposed to the stakeholder oriented group.
The results may provide students and managers with means of improving their understanding and handling of stakeholder reactions to firms’ actions.
Index
Statement of Originality ... 2
Abstract ... 3
Index ... 4
1. Introduction ... 6
2. Literature review ... 9
2.1 Punishment and reward behavior ... 9
2.2 Firm conduct and punishment reward behavior ... 10
2.3 Emotional response on corporate conduct ... 11
2.4 Firm-‐orientation and punishment reward behavior ... 12
2.5 Relevance ... 15
3. Theoretical framework ... 18
3.1 Company conduct and stakeholder punishment reward behavior ... 19
3.2 Company conduct and stakeholder emotional response and punishment behavior ... 22
3.3 Strategic-‐orientation and stakeholder punishment and reward behavior ... 23
3.4 Company’s orientation and stakeholder emotional response ... 24
3.5 The effect of reciprocity ... 24
4 Methods and design ... 27
4.1 Method ... 27
4.2 Data collection and sample ... 28
4.2.1 Data collection ... 28
4.2.2 Sample ... 29
4.3 Procedure and structure ... 29
4.4 Stimuli and measured variables ... 30
4.4.1 The main manipulation ... 30
4.4.1.1 Company introduction and-‐orientation (Moderating Variable) ... 30
4.4.1.2 Company conduct (Independent variable: Stimulus to provoke dependent variable) .... 33
4.4.2 The Mediating Variable ... 35
4.4.3 The Dependent Variable ... 36
4.4.4 Moderating Variable ... 36
4.4.5 Control variable ... 36
4.5 Statistical methods ... 37
5 Analysis and Results ... 38
5.1 Preliminary analysis ... 38
5.1.1 Reliability, validity, normality testing, and scale computation ... 38
5.1.2 Correlation ... 43
5.2 Hypotheses testing ... 46
5.2.1 The effect of corporate conduct on punishment and reward behavior (assumed) ... 46
5.2.2 The effect of positive and negative moral emotions on punishment and reward behavior48 5.2.3: The moderating effect of corporate-‐orientation on punishment and reward behavior .... 54
5.2.4 The effect of corporate-‐orientation on emotional response ... 56
5.3 Additional analysis ... 58
6 Discussion and Conclusion ... 62
6.1 Implications ... 62
6.1.1 implications for literature ... 62
6.2 Limitations ... 66
6.3 Recommendation for future research ... 67
6.4 Conclusions ... 68
7 References ... 69
8 Appendices ... 77
8.1 Cronbach’s Alpha Analyses ... 77
8.2.1 Chronbach’s Alpha Emotional response scale ... 77
8.2.2 Cronbach’s Alpha Reward Scale ... 77
8.2.3 Trust Scale ... 78
1. Introduction
Stakeholder theory is a widely discussed framework first introduced by Freeman (1979). All parties related to a corporation that affect or are affected by the actions of a corporation are stakeholders (Freeman, 1979). According to stakeholder theory a typical large corporation has employees, customers, management, owners, suppliers and community as stakeholders. Just as in the stockholder theory, stakeholders have rights to make certain claims. Stakes require action and conflicting stakes need to be resolved. Stakeholder theory redefines the purpose of the firm opposed to stockholder theory, which argues that the purpose of the firm is to
maximize the welfare of stockholders (Freeman, 1979). In contrast, stakeholder theory argues that a firm can have three orientations. The stakeholder, Profit maximizing and
mixed-orientation. Stakeholder-oriented firm base their conduct on and prioritize fairness towards stakeholders. Profit maximizing firms prioritize their performance. Mixed-orientation firms combine both fairness- and performance-based conduct (Bridoux and Stoelhorst, 2014; Quintelier et al., 2016).
How a firm displays its’ orientation leads to stakeholders’ perception of the goals. Stakeholder perception of these goals will possibly elicit a stakeholder response, positive (reward behavior) or negative (punishment behavior), which may impact the business (gains or losses) (Power, 2013; Sweetin et al., 2013). For instance, a study by Hoffman et al., (2015) shows that profit maximizing actions are likely to elicit less cooperation among stakeholders. Since stakeholder punishment and reward behavior can contribute to the success of a
corporation, it is crucial to unravel which factors impact this behavior.
The behavior of stakeholders is a puzzling phenomenon and this research will attempt to create an understanding. Quintelier et al., 2016 suggested that punishment and reward
behavior may be directed by the orientation of the firm. This is an interesting thought because it implies that the firm can adapt to stakeholder responses even before an action of good or
bad conduct is taken. Therefore, the focus of this study lies on how the orientation of the firm may influence punishment and reward behavior by stakeholders towards the firm.
Quintelier et al., 2016 suggests that this behavior potentially moderated by the personal reciprocal character trait of the stakeholder (Quintelier et al., 2016). For instance, one who has a true ‘you reap what you sow’ mentality might be inclined to punish, whereas one with a more strategic, pragmatic vision, might condone misconduct that leads to profitability and will not be inclined to punish.
This study investigates how the reciprocal character trait of the stakeholder may influence its punishment and reward behavior and what the possible role of emotional response is. To investigate this, a paradigm was set up where four groups of subjects were exposed to a unique combination of two different vignettes. One described a stakeholder-oriented company and the other described a profit-oriented company, after which some questions were posed. These questions served to measure the baseline emotional response to the company that was portrayed. Subsequently, the subjects were exposed to a vignette that described either a good conduct by the company, or a bad conduct by the company. Then some questions were posed to measure their emotional responses, their intention of punishment and reward, followed up by questions with regard to their reciprocal personality.
The results may provide students and managers with a means of improving their
understanding and handling of stakeholder reactions to firms’ actions. For instance, a firm that knows portraying a stakeholder-orientation will elicit higher stakeholder punishment response when they misbehave, might want to reconsider their strategy. For example, socially responsible conduct might be costly for a company, like purchasing clothes made in Europe versus Asia, or the enforcement of protocols of fair labor in the suppliers’ Asian factory. If a company cannot afford such costs, or will lose their customers when they increase their product prices, a stakeholder-orientation image may not be appropriate.
In summary, this thesis investigates what effect stakeholder-orientation has on stakeholder punishment and reward behavior. Research suggests a role for firm orientation. Additionally, it is suggested that reciprocal character traits of stakeholders and emotionality might amplify or influence punishment and reward behavior.
2. Literature review
2.1 Punishment and reward behavior
Adults trying to raise an infant are probably most aware of this premise: almost all behavior rests on reinforcing good and punishing of bad behavior. That is how people, of all ages, learn. Rewards lead to repetition of previous behavior and a penalty leads to avoidance of previous choices. These effects have been described in multiple journals as the reinforcement learning model (Kubanek et al., 2015).
Reinforcement and punishment behavior are also ways for stakeholders to encourage or discourage firms’ policy and decision making. Research into stakeholder behavior with regard to corporate decision making shows that stakeholders tend to punish misconduct and reward good behavior, but the mechanism to what extend a stakeholder is willing to punish or reward is still unclear (Sweetin et al., 2014, Groening and Kanuri, 2013, Podsakoff et al., 2006). This mechanism can serve as a natural guideline for a corporations’ decision making and can be used to determine firms’ strategy and orientation (Groening and Kanuri, 2013). Therefore, it is important to unravel this mechanism and discover which factors affect punishment and reward behavior.
A challenge in unraveling this mechanism is that severity of misconduct or good behavior does not seem to be perfectly correlated to the stakeholder response (Kubanek et al., 2015). Therefore, it seems likely that there are multiple factors mediating and moderating
stakeholder response. So which factors mediate stakeholder punishment and reward behavior? Considering the power of good ‘word-of-mouth’ on social media, positive feedback and reward behavior is a true asset. Studies show that consumers’ action in social media, which would be a reciprocal effort with an intermediate cost (posting reviews about the firm and its products) can positively influence firms’ market value (Chung et al, 2015, Luo et al., 2013, Culnan et al., 2010).
Studies show that punishment behavior like bad-word-of-mouth and boycots negatively impact firm’s value (Saeidi et al., 2015).
Thus, stakeholder punishment and reward behavior can damage or create firm value. This means that there is value in being able to predict punishment and reward behavior. A firm’s strategy can be adapted in anticipation of these behaviors and potentially create profit. To be able to predict these behaviors it is import to investigate the factors that might mitigate these behaviors.
2.2 Firm conduct and punishment reward behavior
Studies show that dilemmas are usually solved by institutions that coordinate punishment, like authorities, police, and the justice system (Guala, 2012). In the world of stakeholder theory, these institutions are the stakeholders. They are the framework to guide company conduct. Stakeholders are always human actors and use punishment and reward to lead the way. Punishment and reward behavior is studied most in reciprocity studies by experimental economists that are interested in costly punishment. An often used paradigm is the Ultimatum Game (Guth et al., 1982).
The Ultimatum game consists of two players having the opportunity to share some money. The first player can make the first offer, the second player can only accept or reject this offer, creating a power asymmetry. If the second player accepts the offer they can both keep the proposed amount. However, if the second player rejects, they both earn nothing. For player one it is most profitable to offer the smallest amount possible, to keep the highest amount possible. For the second player, the best move is to accept any amount that is higher than zero. In theory, this should lead to inequitable distribution of resources. In actuality, this is not the case (Guth et al., 1982).
Camerer (2003) shows that offers perceived as fair (30-50% of the endowment) are usually accepted. Offers that are perceived as unfair however (less than 30% of the endowment), are
rejected about half of the time. This is as punishment to player one, even at their own costs. When actors perceive unfairness, they tend to punish, while perceived fairness leads to
rewards for both parties. Actors in stakeholder theory fulfill a useful social function by setting the expectation for unfair players and pushing them away from future misconduct (Guth et al., 1982).
Unfortunately, the ultimatum game is not the right paradigm to measure variation in punishment, as there is only the option of accepting and rejecting for player two. Research into variation in punishment and reward behavior needs a measurement system that provides more granularity.
Studies suggest that perception of fairness and unfairness leads to emotional responses (Shinada et al., 2004; Grappi et al., 2013). Therefore, these emotions may have a mediating effect on punishment and reward behavior.
2.3 Emotional response on corporate conduct
Moral emotions seem to be the initial response after perceiving an act as fair or unfair. Shinada and colleagues (2004) state that punishment and reward behavior and following actions are formed by emotional responses to the actions of a corporation. These emotional responses come forward from feelings of anger, unfairness and guilt (Shinada et al., 2004). Ferh and Gachter (1999) studied cooperation and punishment in public goods experiments. They demonstrated that bad conduct leads to negative emotions among cooperators and proposed that these emotions may influence severity of punishment. They provided evidence for reciprocal punishment and that punishment follows the pattern of bad conduct. The worse the conduct, the heavier the punishment. Furthermore, they show that one is more likely to cooperate to punish, which in case of stakeholders, could mean organize activist movements or strikes. Therefore, Fehr and Gachter believe that negative emotions among stakeholders can be seen as actual threats to a company.
Grappi et al., (2013) constructed a framework to explain stakeholder protest behavior against corporate irresponsibility. The study shows that corporate irresponsibility leads to negative emotional response, including sadness, anger and disgust and that this leads to “bad-word-of-mouth” and other punishment behavior.
For these reasons, it is important to consider and emotional response of the subjects while researching punishment behavior in stakeholders.
2.4 Firm-orientation and punishment reward behavior
From the previous sections we can conclude that, in the interest of a firm’s value, it is
important to explain variation in stakeholder punishment and reward behavior. As discussed in the introduction, stakeholder theory is a widely discussed framework first introduced by Freeman (1979). Firm orientation determines how decisions are made.
Stakeholder-oriented firms prioritize motives to attend to all stakeholders’ needs and interests.
Even at the cost of performance, they will voluntarily practice a fair and beneficial distribution of value to satisfy stakeholders’ needs. (Quintelier et al., 2016).
Profit-oriented companies prioritize motives to increase returns, typically engaging in
practices that increase performance even at the detriment of some stakeholders (Quintelier et
al., 2016).
In the interest of study clarity, we are disregarding companies with a mixed orientation (Quintelier et al., 2016).
Many studies describe stakeholder-oriented companies as having great corporate
responsibility. However, corporate responsibility seems like a strategy that many companies want use because of potential economic incentives and its strong positive effect on the behaviors of customers, employees, and investors (Maden et al., 2012).
Sweetin and colleagues (2013) used an experimental design to study willingness-to-punish, willingness-to-reward, brand attitude, and purchase intention for corporate social
irresponsibility (CSI), or misconduct, within consumers. They demonstrated that consumers dealing with misconduct were more likely to punish and less likely to reward than the control group.
Quintelier et al., 2016 theorizes that companies that project a stronger stakeholder-orientation (other regarding, moral) are more likely to induce moralistic cognitive decision making. Conversely, firms projecting a relatively strong profit-orientation are more likely to elicit instrumental cognitive decision-making.
This theory is also supported by neuroscientific research. The “Dual Process Theory”, proposed by Piaget in 1948, claims that humans have the ability to use emotion based and rationality based cognitive subsystems (Vaisey, 2009). In simple terms, these two systems are active and competing in every decision one makes. Research shows more activity in the prefrontal cortex, an area that is highly associated with planning and computing (Corricelli & Nagel, 2009, van den Bos, 2009) when strategic, rational decisions are being made.
Conversely, moral decision-making is associated with activity in the anterior insula, the cingulate cortex and the limbic system in general, which are highly associated with negative affect, emotional conflict, vicarious emotions, and behavioral control (Majdandzic et al., 2012; van den Bos et al., 2009).
Additionally, neuroscientific studies show that moralistic evaluations also elicit moralistic, reciprocal decision making and punishment (van den Bos et al, 2009). Delgado and
colleagues (2005) studied differences in (neural) responses when participants had to make decisions based on one others’ prior behavior. Participants were informed about either the moral or immoral behavior of the other person prior to those decisions and were then asked to reward or punish the actions of the other person. They posed three different hypotheses. 1) No differences in decisions of the participant will be seen after being informed about the moral background of the other person, due to rational economic decision making. 2) Information
about moral characteristics behind the other person’s choice creates expectations of the participant, and failed expectations might increase the magnitude of the responses in the trial and error learning brain circuit. 3)The bias that was created by being informed about the moralistic evaluations of the other person will modulate the brain mechanisms associated with trial and error learning, which would make the participant less reliant on feedback or willing to discount that information. Their study shows evidence that moral beliefs can affect economic decision making due to the modulation of the human caudate nucleus.
Phan et al., (2010) showed, using fMRI and a trust game with fictional partners who have different reputations for reciprocity, that positive actions yield positive returns. They also show that positive reciprocity engages the ventral striatum and orbitofrontal cortex, which is absent when the partner has poor reputation, which is the assumption that the partner would be immoral. Thus, prior information about reciprocity changes brain activity, indicating a Reuber and Fisher (2016) examine why stakeholders may overlook some, even ethical
breaching actions. They developed a model in which they appoint perception of firm’s actions and defenses, stakeholder motivation and media coverage as factors that mediate punishment behavior. The perception of the firm’s actions and extent of punishment and reward response is believed to be associated with the perceived threat of the stakeholder, thus if the bad conduct may affect the stakeholder any way the punishment would be affected
correspondingly.
Studies also show that firms can potentially restore their legitimacy with stakeholders when they go through a transgression. They would have to provide an explanation of their behavior and serve an appropriate punishment. In this framework, the firm would be resetting the expectations of the stakeholders and elicit a moral thought process and give moral
of forgiveness by stakeholders is higher when they make rehabilitative changes towards the stakeholder demands (Zavyalova et al., 2016).
Thus, research suggests that company-orientation may affect stakeholder punishment reward behavior. Does this mean, when a company behaves irresponsibly, and is driven by economic incentives, the company will elicit reciprocal punishment behavior in the stakeholder? Would the consumer use own resources to punish this company? Would this punishment be harder when they have portrayed a stakeholder-oriented position, creating a mismatch between actions and stakeholder expectations? Are good actions of a stakeholder-oriented also reciprocally awarded, even at stakeholders’ own costs?
Neuroscientific research shows that bad conduct leads to higher emotional responses than good conduct, therefore the payoff of being stakeholder-oriented and good conduct might not be as high as the costs of good conduct. Studies also show that people weigh bad news more heavily than good news (Folkes and Kamins, 1999; Mizerski, 1982), therefore bad conduct may have a stronger effect on stakeholder punishment behavior than good conduct on
stakeholder reward behavior. Then, maybe portraying a profit-orientation is overall a safer bet because it elicits instrumental cognitive decision making, which elicits less emotional
responses and therefore less punishment but also less reward? These are all questions that remain unanswered in current literature, and therefore, a gap in research exists in explaining variation in stakeholder behavior.
2.5 Relevance
In the light of the popular topic Corporate Responsibility (CR), many firms engage in in CR initiatives to contribute to society in a positive manner. A study by Maden et al., (2012) confirms that CR has a strong positive effect on reputation and behaviors of stakeholder groups like customers, employees and investors. A firm can display this prosocial behavior and present itself as stakeholder-oriented. However, the motivation behind this behavior can
be out of profit-oriented viewpoint. For example, when a company takes good care of their suppliers in the cacao industry and promotes and advertises this. The company then actively promotes their stakeholder-orientation which may elicit positive feelings and stakeholder reward behavior. However, this can also be a powerful marketing strategy, conversely, if the corporation does less for the suppliers then they lead their stakeholders to believe, this may lead to negative feelings, skepticism and stakeholder punishment behavior (Skarmeas and Leonidou, 2013).
Evidence that the intention behind the prosocial behavior of firms is important in the
stakeholders’ reactions is supported by the study of Darnall and Aragon-Correa (2014). They studied the effect of ecolabels (programs designed to signal information to a stakeholder to reduce information asymmetries) on the firms’ sustainability strategy and stakeholder behavior. They found that these labels were only successful if the stakeholders perceived them as being credible. Their behavior was dependent on how they perceived the credibility of the labels.
It can be hypothesized that when stakeholders perceive firm’s intentions as disingenuous, it might elicit feelings of unfairness, negative emotional responses and reciprocal behavior in stakeholders. This may influence the attitude towards the firm and speculatively lead to aversive behavior and punishment, and potentially damage the firm in reputation and/or financially.
These findings were supported by the study of Skarmeas and Leonidou (2013), who argue that CR claims and incidents of corporate misconduct create skepticism under stakeholders, which can lead to decreasing profitability of the firm (Saiedi et al., 2015).
A gap in research exists in what explains the variation of punishment behaviors in
the effect of business-orientation on stakeholder punishment and reward behavior, in which a distinction will be made between stakeholder-oriented and profit-oriented firms.
Therefore, this thesis will attempt to answer the following research question:
How does firm-orientation affect stakeholder punishment and reward behavior?
Furthermore, research suggests that emotional responses and individual reciprocal character traits of the stakeholders to good and bad actions of firms are moderators in stakeholder punishment and reward behavior. Research investigating the interplay of these factors is lacking.
Stakeholder punishment behavior often leads to a decrease in profitability of the firm (Saiedi
et al., 2015, Gatzert, 2015), therefore it is important for firms to know what kind of behavior
to expect when the company conduct itself in a certain way, and avoid circumstances that can lead to punishment behavior. The results may provide students and managers with means of improving their understanding and handling of stakeholder reactions to their actions.
3. Theoretical framework
In this chapter, the theoretical framework of this thesis will be elaborated on. In figure 1. The conceptual model of expected relationships is displayed. Subsequently, hypotheses will be discussed in detail. The study will focus on answering the research question:
How does firm-orientation affect stakeholder punishment and reward behavior?
Additionally, this research will also try to answer the following sub questions:
How does a negative emotional response affect firm-orientation-induced stakeholder punishment and reward behavior?
How does a positive emotional response affect firm-orientation-induced stakeholder punishment and reward behavior?
How does firm-orientation affect emotional response?
How does individual’s reciprocal character affect firm-orientation-induced stakeholder punishment and reward behavior?
Figure 1. The conceptual model
3.1 Company conduct and stakeholder punishment reward behavior
Previous studies suggest that when companies conduct themselves in a bad way they can hurt their success due to stakeholder punishment (Fehr and Gachter, 1999; Sweetin et al., 2013). Good conduct can lead to stakeholder reward behavior (Phan et al., 2010; Reuber & Fisher 2015), which in turn can lead to company success (Hahn, 2015).
Most literature on this subject describes punishment-reward behavior through reciprocity, which is described as a character-trait. One that responds in a kind manner to action that are perceived to be of kind intend, and in a hostile manner to actions that are perceived to be of hostile intend (Rabin, 1993; Segal and Sobel, 2004; Dufwenberg and Kirchsteiger, 2004; Falk and Fischbacher, 2005). Preferences depend on both material payoff and intention why an agent chose a certain action (Fehr and Schmidt, 2006). On a neural level, evidence of this was
posed by Watanabe et al., 2014, who found that neural mechanisms underlying indirect reciprocity and suggests that pay-it-forward reciprocity may not occur as myopic profit maximization but elicit emotional rewards.
There are many factors that might explain the variation in stakeholder punishment and reward behavior. Barnett (2014) argues that the stakeholder has to notice the misconduct before he can act on it and proposed that stakeholders might be limited in their ability to deter firm misconduct through social control. This would explain the variability of stakeholder punishment behavior.
A study by Groening and Kanuri (2013) shows that nearly 50 percent of the reactions on positive and negative corporate social events do not match between investors and other stakeholders. This means that there are many situations that investors do not reward a firm even though stakeholders perceived actions as positive. There were also cases when a firm conducted an action that was perceived negatively by other stakeholders but were encouraged by investors.
In terms of punishment and reward behavior one can distinguish variety in severity. There are several theories describing reciprocal punishment and reward behavior. For this study, strong reciprocity theory was adapted (Carpenter, 2009; Guala, 2012), which assumes altruistic punishment and reward behavior. According to this theory, a strong reciprocator would always cooperate with others and punish non-cooperators, to the extent that the punishment may be costly to the actor (Gintis et al., 2000). This in mind, the highest level for punishment, would be what we know as altruistic punishment. Altruistic punishment is a strong form of reciprocal punishment and is described in behavioral economics, anthropology and
psychology research (Guala, 2012). In the case of strong altruistic punishment, one would punish and not benefit from it, and it would be costly for the punisher (Fehr and Gachter, 2002). This is a deviation of the textbook example assumptions of the homo economicus. The
homo-economicus acts as a rational and self-interested agent that would only pursue something when they can gain something from it. They would always attempt to maximize utility and profit, so would never invest in something that they would not benefit from (Henrich et al., 2001).
An example of this is actively protesting misconduct of a company towards a third party, for example suppliers of a firm; cocoa farmers in Africa. This is costly for the activists (for example: take a day off from work, stand in the rain, use up their voice) but personally, they will gain nothing directly from it. If the company decides to treat the farmers better, the stakeholder will still buy the same chocolate in the store (DeConinck et al., 1997).
There is also a strong reciprocal way to reward, which is when a stakeholder would perform costly actions in favor of the firm that would not benefit themselves. For example, when somebody would act as an ambassador for a company. The person will encourage good conduct by actively supporting the firm, although they will not gain anything from it (DeConinck et al., 1997) .
A lesser form of punishment and reward is also seen as strong reciprocal, but against less costs. For example, when a stakeholder who would actively discuss a firms’ actions in conversations without being asked for their opinion. It costs extra effort to bring up the topic of conversation, while it will lead to no gain (DeConinck et al., 1997).
A lesser action is to give an opinion when asked, where almost no costs or gains are linked to the action. These options are all open if the stakeholder recognizes the conduct decides to take action (DeConinck et al., 1997). Furthermore, a stakeholder can always decide to take no action (Shinada et al., 2004). Of course there is more nuance and variability in stakeholder punishment and reward behavior.
Punishment behavior can potentially be a big threat to profitability of a company (Power, 2013). Skarmeas and Leonidou (2013) show that attributions of egoistic motives elicit
skepticism towards a firms’ intention. This skepticism hurts retailer equity, places the firm in a difficult-to-defend position and it facilitates a bad reputation and word of mouth, which inevitably leads to decreasing returns.
Therefore, it is predicted that we will see higher reward and lower punishment after stakeholders are exposed to good conduct, and higher punishment with lower reward after they are exposed to bad conduct. Interestingly, studies also show that people weigh bad news more heavily than good news (Folkes and Kamins, 1999; Mizerski, 1982), so bad conduct may have a stronger effect on stakeholder punishment behavior than good conduct on stakeholder reward behavior.
This leads to the following hypothesis about firm conduct and stakeholder punishment and reward behavior.
H1a: Good (as opposed to bad) conduct leads to more reward H1b: Good (as opposed to bad) conduct leads to less punishment
3.2 Company conduct and stakeholder emotional response and punishment behavior Studies have suggested that emotional responses have a mediating role on punishment and reward behavior (Shinada et al., 2004; Fehr and Gachter, 1999). Shinada and colleagues (2004) state that punishment and reward behavior and actions are formed by emotional responses to the actions of a corporation. It is suggested that these negative responses are feelings like anger, fear, sadness and disgust (Grappi et al., 2013). The positive responses come from feelings like joy (Shinada et al., 2004).
As discussed before, Ferh and Gachter (1999) show that bad conduct leads to negative emotions among cooperators and propose that these emotions may influence severity of
punishment. In addition, they believe that negative emotions among stakeholders can be seen as actual threats to a company.
Grappi et al., (2013) tried to find an explanation for consumer word of mouth and protest behaviors against corporate irresponsibility. They show that consumer’ negative moral
emotional responses to bad conduct lead to negative word of mouth and protesting. It seems a trend in these findings that emotional responses mitigate the behavior of people. Therefore, I hypothesize that positive response will increase reward behavior and that negative emotional responses will increase punishment behavior.
For these reasons the following hypothesis are posed:
H2a: The positive effect of good conduct (as opposed to bad conduct) on reward is mediated by positive moral emotions.
H2b: The positive effect of bad conduct (as opposed to good conduct) on punishment is mediated by negative moral emotions.
3.3 Strategic-orientation and stakeholder punishment and reward behavior
When reviewing research on the effect of strategic-orientation it becomes apparent that stakeholders’ responses are affected by the information they have about the company prior to judging the company on good or bad conduct. For instance, Phan et al., (2010) shows in his research that, when subjects had to play a trust game with fictional partners, they responded differently when the fictional person was told to be of reciprocal or self-regarding nature. Quintelier et al., (2016) uses stakeholder theory to apply this on stakeholders versus
companies. She proposes that the way a company is oriented (Profit -oriented, in the previous example self-regarding, or Stakeholder -oriented, in the previous example reciprocal) will affect stakeholder emotional response and will moderate the punishment and reward behavior. This leads to pose the following hypotheses:
H3a: Stakeholder-orientation (as opposed to profit orientation) will positively moderate the positive mediation effect of good conduct on reward.
H3b: Stakeholder-orientation (as opposed to profit orientation) will positively moderate the positive mediation effect of bad conduct on punishment.
3.4 Company’s orientation and stakeholder emotional response
Given the hypothesis that company orientation moderates and the emotional response mediate punishment behavior, it is possible that these variables influence each other as well.
Stakeholder emotional response might be affected by corporate orientation because the orientation creates an expectancy of conduct. Neuroscientific studies show that when
expectancy is violated, the emotional response increases (Ask & Landstrom, 2010; Steinbeis
et al., (2006); Shapiro et al., 1999) and provokes ‘threat’ responses in the receivers (Mendes et al., 2007). Therefore, I hypothesize that emotional response would be amplified if the
expectancy was violated, e.g. if the action those not match theorientation; Stakeholder -oriented x bad conduct or profit--oriented x good conduct.
H4a: If a company is stakeholder-oriented (as opposed to profit-oriented) and shows good conduct the stakeholder positive emotional response will increase
H4: If a company is stakeholder-oriented (as opposed to profit-oriented) and shows bad conduct the stakeholder negative emotional response will increase
The role of human motivation is a broadly discussed topic in management literature, for instance in realizing the value creation potential of resources (Bridoux et al., 2011). It is also an important topic in the stakeholder theory.
Economic theories mainly assume homo economicus: everyone is self-interested and the overarching goal of individuals and companies is to maximize profit. Many studies have been done showing that the rational egoists’ assumption is a sufficient predicting model for the outcome in competitive market situations and auctions (Kagel and Roth, 1995). However, behavioral economic studies show that not everybody is driven by self-interest. People exhibit diverse motives to cooperate (Ostrom, 2000, Bridoux et al, 2011) and even show predictable behavioral patterns in simulated settings.
Research describes an outcome transformation model with three dimensions; weight assigned for payoffs to oneself, weight assigned to payoffs for others and the weight assigned to reciprocity of fairness (Bridoux et al., 2011; De Cremer & van Lange, 2001; Stouten et al., 2005). On this basis, it was found that most individuals can be classified as either
self-regarding or reciprocator and that these traits are stable characteristics (Bridoux et al., 2011). It has been shown that only 20-40% of people are completely self-regarding, therefore, the assumptions of many economic theories do not reflect the real world (Bridoux et al., 2011). Self-regarding people try to maximize their own payoffs and only care about payoffs to others as long as it also influences their own pay off. Other people show reciprocity, as long as they feel they are treated fairly, they will cooperate and contribute to the common good. Some of these reciprocators will even use their own resources to punish self-interested people; they punish at their own expense (Bridoux et al., 2011; De Cremer and Van Lange, 2001).
In summary, how and to what extent a stakeholder would respond to a corporate action would likely be affected by the reciprocal character traits of the individual. According to this theory
only 20-40% of the people would respond in a manner that could be predicted by the homo economicus model.
Understanding reciprocal stakeholder behavior can contribute to a better understanding of why companies comply with stakeholder demands for prosocial corporate behavior that goes beyond legal obligations (Hahn et al., 2015).
Thus, literature has suggested that stakeholder punishment and reward behavior differs when a background behind the decision is made is different. Like the study of Phan et al., (2010) who showed that personal information about the fictional decision maker, if he was reciprocal or self-regarding, changed the pattern of punishment and reward. Recent studies show that human behavior is not always self-centered, but reciprocal (Quintelier et al., 2016; Bosse and Phillips, 2016; Bridoux and Stoelhorst, 2014; Hahn, 2015; Harrison et al., 2010; Hahn and Albert, 2015). I expect reciprocal stakeholders to reward a firm when they evaluate the treatment towards the other stakeholders fair, and that they will punish the firm to own costs when they evaluate the behavior towards other stakeholder as unfair (Quintelier et al., 2016; Bosse et al., 2009; Bridoux & Stoelhorst, 2014). Therefore, the following hypotheses are posed:
H5a: If a stakeholder-oriented company (as opposed to a profit-oriented company) shows good conduct stakeholder reward is positively moderated by reciprocal personality of the stakeholder
H5b: If a stakeholder-oriented company (as opposed to a profit-oriented company) shows bad conduct stakeholder punishment is positively moderated by reciprocal personality of the stakeholder
4 Methods and design 4.1 Method
In order to test the hypotheses, a quantitive approach was used. A vignette method was applied (Finch, 1987; Hill, 1997). Vignette study is a reliable method to study behavior and attitudes of individual after perceiving actions and is typically short scenarios written, intended to elicit responses (Alexander & Becker 1987; Hill, 1997).
The researcher can standardize the stimuli with this method and that allows for fair
comparison between outcomes. Reading the vignette leads to a authentic perception of the subjects and will not likely bias the subjects’ responses, and these responses are likely to correspond with real situation (Alexander & Becker 1987; Hill, 1997). Therefore, it is a suitable method for this study.
Participants were randomly assigned to an experimental group by the following procedure: First, subjects were randomly assigned to Vignette 1 (Stakeholder -oriented company, see stimuli section for vignettes) or Vignette 2 (Profit -oriented company). 50 percent of subjects that entered the study were assigned to read vignette 1 and answer questions regarding their baseline emotions. 50 percent of people that entered the study were assigned to read Vignette 2 and answer questions regarding their baseline emotions towards the company. Which vignette a subject was exposed to was randomly assigned by the survey software.
Subsequently, participants were randomly assigned to vignette 3(a sketched situation where the company conducted itself in a good way) or vignette 4 (a sketched situation where the company conducted itself in a bad way). 50 percent of subjects that entered the study were assigned to read vignette 3 and answer questions regarding their emotions and intended punishment reward behavior towards the described good conduct -oriented company. 50 percent of people that entered the study were assigned to read vignette 4 and answer questions
regarding their emotions and intended punishment reward behavior towards the bad conduct, ultimately creating 4 different experimental groups; Group 1 (n=54) first read a vignette introducing a stakeholder-oriented company followed by a vignette regarding displayed good conduct. Group 2 (n=56) first read a vignette introducing stakeholder-oriented company and were then exposed to a vignette regarding the company displaying bad conduct. Group 3 (n=48) read a vignette introducing a profit-oriented company followed by a vignette regarding the display of good conduct by that company. Group 4 (n=62) read the vignette introducing a profit-oriented company followed by a vignette regarding the display of bad conduct. A schematic presentation is shown in fig. 2. The size of the groups ended up to be unequal due to drop out.
Figure 2: Schematic representation of the four groups
4.2 Data collection and sample
4.2.1 Data collection
The data and output for this paper were generated using Qualtrics software, Copyright © 2017 Qualtrics. Qualtrics and all other Qualtrics product or service names are registered trademarks or trademarks of Qualtrics, Provo, UT, USA.
Stakeholder-‐ oriented vignette
Good conduct Group 1 (n=54)
Bad conduct Group 2 (n=56)
Profit-‐oriented vignette
Good conduct Group 3 (n=48)
In order to collect data, the study was distributed through social media, using a fan page site created for this study in order to distribute the survey more widely through the option of facebook advertisments, on which €10 was spent. €25,- was used two times for Bol.com gift-cards. These gift-cards were used to create a raffle, one among the people who shared the link, and one among the people that completed the study.
The aim was to find 50 subjects per treatment group to ensure the study would have sufficient power (Saunders & Lewis, 2014; Campbell et al., 1996). Four ‘treatment’- groups equals a goal of 200 subjects.
4.2.2 Sample
385 people entered the study of which 219 subjects (n=219, 65 male, 153 female, 1
unspecified) completed the survey without missing values. All subjects were over the age of 18 (Min 19; Max 73; Mean: 40, Stdev. 13,59) and signed a consent form at the start of the questionnaire. Participants responded to emails or social media messages and ads.
4.3 Procedure and structure
Informed consent forms were signed after subjects were informed about the experiment. Surveys were filled out on electronic devices to subject’s preference, on either smartphone, tablet, laptop or pc. The survey started with a vignette followed by the emotional response questionnaire. The subjects were then asked to read a second vignette and fill out a second emotional response survey as a within subject measure and an intention of punishment and reward questionnaire. Then, the subjects were asked to fill out a reciprocity and trust questionnaire, followed by a socio-demographics questionnaire. Finally, the subjects could optionally fill out their mail address to participate in a raffle for a 25 euro gift card. The e-mail addresses were separated from the answers to keep participants anonymous, as soon as the data was exported from Qualtrics (Qualtrics, Provo, UT).
4.4 Stimuli and measured variables
For this study, vignettes were used as manipulating factors. The first pair of vignettes served to introduce and create an image of a corporation and their-orientation. The second pair of vignettes was used to provoke a response punishment or reward response, creating a measure of the moderating role of company-orientation on stakeholder punishment and reward
behavior.
4.4.1 The main manipulation
Vignettes were drafted with great precision. The examples were derived from news about an existing large corporation. The third party that judges how the company conducts itself in stakeholder’s interest is an existing organization, part of Oxfam Novib1. They scored large corporation on several stakeholder subjects, such as the country where the production facilities are located, how the company deals with women’s interest, farmers, employees, climate, water, palm oil and transparency (Hoffman, 2013; Marston, 2016). They also check to see what kind of demands the companies place on their suppliers to meet their standards. The subject of the vignettes was derived from a company that had one of the highest scores. The description of the company was for both stakeholder- and profit-oriented company the same. Variation was the part where the score, the stakeholder-oriented company had a fictitious high score and the profit-oriented company had a fictitious low score. Furthermore, there was a difference intend to improve their current policies.
4.4.1.1 Company introduction and-orientation (Moderating Variable) Vignette 1: Stakeholder -oriented company
[Lees de volgende tekst alstublieft aandachtig door en beantwoord daarna de vraag over de tekst.
VOEDEX is een multinationale onderneming op het gebied van voedingsmiddelen. Het bedrijf heeft circa 170.000 werknemers wereldwijd en heeft jaarlijks een omzet van ongeveer 50 miljard euro, met een winst van rond de 5 miljard.
Afgelopen jaar heeft een zeer betrouwbare partij het bedrijf gewaardeerd voor de WERELD-BELANGEN-INDEX (WBI). Deze index beoordeelt hoe bedrijven omgaan met de belangen van derden. Hiervoor is VOEDEX beoordeeld op diverse thema’s, bijvoorbeeld hoe het omgaat met de belangen van het land waar de fabrieken zich bevinden, maar ook hoe het omgaat met de belangen van vrouwen, boeren en werknemers, en met klimaat, water en transparantie. Ook wordt er gekeken naar de eisen die er op die thema’s aan toeleveranciers worden gesteld.
VOEDEX scoorde voor het jaar 2016 hoog op alle gemeten thema’s. De waardering van VOEDEX was met 90 van maximaal 100 punten de hoogste binnen haar bedrijfstak. Het bedrijf geeft aan verder in de verschillende thema’s te willen investeren om bij te dragen aan een betere wereld.] (165 words)
Translation: Please read the text carefully and answer the questions about the text. VOEDEX is a multinational corporation in the food industry. The company has about 170000 employees and has a yearly turnover of 50 billion euros, with a profit of about 5 billion.
Last year, a very trustworthy organization valuated the company on the World-Interest-Index (WII). This index valuates how the company conducts itself with regards to third parties, stakeholders. VOEDEX was scored on multiple themes, for example, how it treats the interest of the countries where the production facilities are located, how it treats women’s rights,
farmers, employees, climate, water and transparency. They also valuate the demands their suppliers have to meet.
VOEDEX scored high on all valuated themes. The valuation of VOEDEX was with 90 out of the max. 100 achievable points the highest ranked company within their industry. The
company says to want to invest in all themes to contribute to a better world.
Vignette 2: Profit -oriented company
[Lees de volgende tekst alstublieft aandachtig door en beantwoord daarna de vraag over de tekst.
VOEDEX is een multinationale onderneming op het gebied van voedingsmiddelen. Het bedrijf heeft circa 170.000 werknemers wereldwijd en heeft jaarlijks een omzet van ongeveer 50 miljard euro, met een winst van rond de 5 miljard.
Afgelopen jaar heeft een zeer betrouwbare partij het bedrijf gewaardeerd voor de WERELD-BELANGEN-INDEX (WBI). Deze index beoordeelt hoe bedrijven omgaan met de belangen van derden. Hiervoor is VOEDEX beoordeeld op diverse thema’s, bijvoorbeeld hoe het omgaat met de belangen van het land waar de fabrieken zich bevinden, maar ook hoe het omgaat met de belangen van vrouwen, boeren en werknemers, en met klimaat, water en transparantie. Ook wordt er gekeken naar de eisen die er op die thema’s aan toeleveranciers worden gesteld.
VOEDEX scoorde voor het jaar 2016 laag op alle gemeten thema’s. De waardering van VOEDEX was met 10 van maximaal 100 punten de laagste binnen haar bedrijfstak. Het
bedrijf geeft aan geen geld te willen investeren in verbeteringen die de omzet van het bedrijf niet ten goede komen.] (164 words)
Translation: Please read the text carefully and answer the questions about the text. VOEDEX is a multinational corporation in the food industry. The company has about 170000 employees and has a yearly turnover of 50 billion euros, with a profit of about 5 billion.
Last year, a very trustworthy organization valuated the company on the World-Interest-Index (WII). This index valuates how the company conducts itself with regards to third parties, stakeholders. VOEDEX was scored on multiple themes, for example, how it treats the interest of the countries where the production facilities are located, how it treats women’s rights, farmers, employees, climate, water and transparency. They also valuate the demands their suppliers have to meet.
VOEDEX scored high on all valuated themes. The valuation of VOEDEX was with 10 out of the max. 100 achievable points the lowest ranked company within their industry. The
company says only to want to invest in improvements that will contribute to the company’s profit.
4.4.1.2 Company conduct (Independent variable: Stimulus to provoke dependent variable) Vignette 1: Good conduct
[Begin 2017 kwam het volgende nieuwsbericht naar buiten.
VOEDEX voert krachtig beleid op het gebied van ontbossing en palmolie. VOEDEX spoort overheden aan om in actie te komen en er zijn richtlijnen voor leveranciers. Vandaag
van de Maleisische OIL GROUP, omdat die een belangrijk keurmerk van duurzaamheid is kwijtgeraakt, nadat ze al meerdere malen op verbeterpunten waren gewezen. Vorige week besloot het overlegorgaan voor duurzame palmolie (RSPO), waaraan VOEDEX deelneemt, zijn keurmerk al in te trekken. Palmolie zit onder meer in margarine, frituurolie en mayonaise. (ANP)] (92 words)
Translation: Start of 2017 the press release below came out. VOEDEX has a firm policy with regards to the destruction of woods and palm oil. VOEDEX encourages governments to take action and has specific guidelines for their suppliers to meet. Today, VOEDEX announced to stop purchasing from the Malaysian supplier OIL GROUP because they lost an important sustainability award, after they have been alerted on points of improvement multiple times. Last week, the board for sustainable palm oil, of which VOEDEX is a member, decided to withdrawal their award. Palm oil is used for, amongst others, the production of margarine, deep-frying oil and mayonnaise.
Vignette 2: Bad conduct
[Begin 2017 kwam het volgende nieuwsbericht naar buiten.
VOEDEX stopt mistanden bij de productie van palmolie in de doofpot. Volgens een vandaag verschenen rapport van Milieudefensie gebruikt VOEDEX op zoek naar nieuwe grond vaak gewelddadige technieken om de inheemse bevolking van haar land te verjagen. Families die voorheen in staat waren in hun eigen behoeften te voorzien worden overgehaald om hun land op te geven in ruil voor nieuwe ontwikkelingsprojecten en werkgelegenheid. Maar in
werkelijkheid verliezen zij hun land en krijgen ze daar slecht betaalde banen voor terug, terwijl hun bos verandert in een plantage. Pesticiden en kunstmest vervuilen het drinkwater. (ANP)](94 words)
Translation: Start of 2017 the press release below came out:
VOEDEX is covering up for maladministration of the production of palm oil. According to a milieu defense report that came out today, VOEDEX uses violent manners to ban native inhabitants from their country in search of new soil for palm oil production. Families that were capable of supplying for their own needs are being convinced to give up their land in trade of new development projects and employment opportunities. In reality, they lose their land for poorly paid jobs while their forest turns into a plantation. Pesticides and fertilizers used on the plantation spoil the drinking water.
4.4.2 The Mediating Variable
Emotional response- As Grappi et al., 2013 demonstrates, consumers’ and stakeholders’ negative moral emotional responses to bad conduct lead to negative word of mouth and protest toward the corporation, which are examples of altruistic punishment. These emotions included anger, contempt and disgust. They also showed that other-regarding, or stakeholder--oriented behavior lead to beneficence, equality and cooperation. These findings were
supported by the study of Shinada et al., 2004. The emotional response was measured by asking participants to rate their emotions on a 7-point Likert scale (To what extent did you experience this emotion? : 1-not at all, 7-extremely), this was limited by the 6 base- emotions (Positive emotion: Happiness; Negative emotions: Anger, Sadness, Fear, Disgust,
Astonishment). The used value for positive response was the measure of happiness, the value used as negative response was the mean of the negative scores, leading to a score range of 1-7 for both negative and positive emotions. These base emotional response questions were derived from Ekman (1994) and translated to Dutch. Questionnaire can be found in the appendix.
4.4.3 The Dependent Variable
Punishment behavior - This construct was measured using a modified version of the four item scale by Hunt & Vasquez-Parraga (1993). Each item was measured using a 7-point Likert scale (1 = very unlikely, 7 = very likely). Example of these questions is: How likely would you give this company a positive review (on social media)? Questions were adapted from DeConinck et al. (1997) and translated to Dutch. Used score was the sum of the four items, resulting in a scale of min. 4 – max. 28 points. Questionnaire can be found in the appendix.
Reward behavior - This construct was measured using a modified version of the four item scale by Hunt & Vasquez-Parraga (1993). Each item was measured using a 7-point Likert scale (1 = very unlikely, 7 = very likely). Examples of questions were ‘How likely would you actively recommend this company?’ Questions were adapted from DeConinck et al. (1997) and translated to Dutch. Used score was the sum of the four items, resulting in a scale of min. 4 – max 28. points. Questionnaire can be found in the appendix.
4.4.4 Moderating Variable
Reciprocity- This construct was measured using a questionnaire on reciprocal character, derived from Maximiano (2010). The questionnaire was adapted to a seven item 7-point Likert scale, asking subjects to rate their personality with questions with on whether they agree or disagree (1-firmly disagree – 7 firmly agree) on statements with regards to
reciprocity (If I humor someone, I expect that person to humor me), and trust (Most people will try to use you when they get the chance). Furthermore, the questions were translated to Dutch. The questionnaire can be found in the appendix.
4.4.5 Control variable
Socio-demographics – Subjects’ gender, age, level of education and status of employment were recorded, since these might affect punishment and reward behavior (Andreoini &
Vesterlund, 2001; Eckel & Grossman, 1996; Gintis et al., 2003). Employment status was measured using three categories; unemployed, employee, own-employed. Level of education was measured using four categories: ‘Middelbare school’ (high school), MBO (secondary vocational education), HBO (higher vocational education) and Academisch (Academical education). A standard questionnaire was derived from the Qualtrics Questionnaire Library (Qualtrics, Provo, UT). Questionnaire can be found in the appendix.
4.5 Statistical methods
Analysis was conducted using PASW Statistics 18, Release Version 18.0.0 (SPSS, Inc., 2009, Chicago, IL, www.spss.com). To analyze the data, first a MANOVA was done as a
manipulation check. Then a correlational test was done, identifying predictors in the model. Then independent T-tests were used comparing direct effects for groups. Then a two-step approach using the SPSS plug-in PROCESS by Hayes (2012) was used. First, the PROCESS Macro Model 4 was used to examine the mediating effects of emotional response and model 7 was used identifying the moderating effect of business orientation on punishment and reward behavior.
5 Analysis and Results
5.1 Preliminary analysis
5.1.1 Reliability, validity, normality testing, and scale computation
The data set was first screened for missing data and errors. Of the 385 respondents that entered the study, 219 subjects finished the survey without missing values. The 219 cases were used for further analyses. All other unfinished cases were deleted.
To examine the measurement instruments for construct validity and reliability Cronbach’s Alpha was used (Cronbach & Meehl, 1955). The golden rule for Cronbach’s Alpha is that it should be higher than α = 0.70 and the ‘corrected item-total correlation’ should be higher than 0.30 (Bernstein & Nunnally, 1994).
The emotions scale showed an overall α = 0.726 (n = 6), which is reliable. However, if you check the corrected item-total correlation the items that measure ‘happiness’ does not meet the 0.30 norm (happiness = 0.045, Cronbach’s Alpha if item deleted = 0.804). The measure for happiness seems to not be correlated with the rest of the questions, therefore this item is separately used to measure positive emotions.
The reward scale showed an overall α = 0.708 (n = 4), which shows that this scale is reliable. The question ‘how likely would you act as an ambassador for this company had a corrected item-total correlation of <0.30 (0.29), leaving the question out however, would not improve the scale (Cronbach’s Alpha if item deleted =0.67). Therefore, it was kept in the
questionnaire.
The punishment scale showed an overall α = 0.693 (n = 4), which shows that this scale is almost reliable. No items could be deleted to improve this score. If, however, the data is dissected in different groups, it shows that the group that answered the question after reading the profit-oriented vignette in combination with a good conduct vignette causes the
Cronbach’s alpha to drop (α = 0.518), whereas the mean Cronbach’s alpha for the other groups is α = 0.725. Furthermore, all corrected item-total correlations are >0.3. Since the assumption of this paper is that good conduct will lead to reward and bad conduct to punishment behavior, it is plausible that the stakeholder punishment responses on good
behavior impacts the data. On this bases, it is decided that the punishment scale can be seen as valid.
The trust scale showed intermediate reliability with α = 0.616, however, no item could be deleted to improve the scale. In addition, all corrected item-total correlations are >0.3. Furthermore, debate is still open on the value of 0.7 for alpha, for instance, Churchill and Peter (1984) suggested an accepted level for the alpha coefficient. According to them a value of alpha below 0.60 is undesirable. According to these metrics, this scale would be considered reliable.
According to the standards set, the construct reciprocity questionnaire can be viewed as having low reliability and cannot be improved upon (α = 0.182, all corrected item-total correlations are <0.3).
The construct of the reciprocity appears to be problematic when testing for Cronbach’s Alpha. Therefore, it was decided to leave this variable out of the structure. For now, all hypotheses regarding reciprocity are rejected with low certainty. Many studies describe a plausible role for reciprocity in punishment and reward behavior (Falk and Fischbacher, 2006;
Fehr and Schmidt, 2006), more research with a different construct is needed to clarify the effect of reciprocal personality in combination with corporate-orientation is, on stakeholder punishment and reward behavior.
H5a: If a stakeholder-oriented company (as opposed to a profit-oriented company) shows good conduct stakeholder reward is positively moderated by reciprocal personality of the stakeholder
H5b: If a stakeholder-oriented company (as opposed to a profit-oriented company) shows bad conduct stakeholder punishment is positively moderated by reciprocal personality of the stakeholder
Summary of reliability values can be found in table 1.
Table. 1 Reliability of the measurement instruments Scale Number of Items Cronbach's Alpha Corrected item-total correlation Delta Cronbach's Alpha*** Emotional
response 6 0.726* lowest: 0.045** Item 1: 0.078 Reward 4 0.708* lowest: 0.461** N.A. Punishment 4 0.694 lowest: 0.455** N.A. Trust 3 0.616 lowest: 0.391** N.A. Reciprocity 3 0.182 lowest: 0.067 item 3: 0.035 * Cronbach's alpha > 0.7
** Corrected item-total correlation > 0.3
*** Maximum value the scale can be improved if item deleted
The variables were tested for outliers to prevent the data from being not normally distributed. Z-scores should be -3 <Z <3. In total, 13 outliers were identified and deleted, the remaining data set contained 206 cases.
Subsequently, the variables were tested for normality (Skewness and kurtosis). The skewness and kurtosis scores both should be between -1 and 1 and should not be larger than three times its standard error (Schreiber, 2008). Results of normality testing can be seen in table 2. Distributions of the measures of reward and positive emotional response before and after reading the conduct vignette are slightly skewed, caused by the groups and over-time
measures that act as implicit factors skewing the distribution. In table 3 and 4 it is shown that groups that were exposed to a bad conduct vignette showed non normal distribution in reward behavior, however they show normal distribution in punishment behavior. Groups that were exposed to a good conduct vignette showed non-normal distribution in punishment behavior
as opposed to normal distribution in punishment behavior. In the assumption that good conduct leads to reward behavior and bad conduct leads to punishment behavior, this division in normal distribution does not form a problem. Emotional response difference is a
differentiated value of the emotional responses and this is normally distributed.
Table 2. Normality testing
Skewness Kurtosis
Punishment 0,399 -0,739
Reward 1,295 1,204
Emotional response difference 0,217 0,464 Negative emotional response before 0,574 -0,957 Positive emotional response before 1,054 -0,286 Negative emotional response after 0,224 -1,151 Positive emotional response after 1,445 0,73
*Std error of skewness 0,165
*Std error of kurtosis 0,338
Table 3: Normality in punishment behavior separated by group
Group Vignette Conduct Skewness Std. Error Skewness Kurtosis Std. Error of Kurtosis Normally distributed?
1 Stakeholder-oriented Good 1,297 0,33 0,807 0,65 No
2 Bad 0,178 0,333 -0,966 0,656 Yes
3 Profit-oriented Good 0,383 0,347 0,169 0,681 Yes
4 Bad -0,138 0,322 -0,221 0,634 Yes
Table 4: Normality in reward behavior separated by group
Group Vignette Conduct Skewness Std. Error Skewness Kurtosis Std. Error of Kurtosis Normally distributed?
1 Stakeholder-oriented Good 0,281 0,33 0,11 0,65 Yes
2 Bad 1,646 0,333 2,033 0,656 No
3 Profit-oriented Good 0,655 0,347 -0,845 0,681 Yes
4 Bad 2,174 0,322 3,597 0,634 No
Then, the created variables were tested for outliers to prevent the variables from being non normally distributed using Mahalanobis distance. As a golden standard, the probability of the MD-score needs to be p>0.001 (Tabahcknick and Fidell, 2007). The Mahalanobis distance was calculated for the combination of the measures of the emotional response scale,
emotional difference between sets, reciprocity, and trust (df=7). One outlier was detected (subject 216, pMD > 0.001), and was removed from the data set. The remaining data set contained 205 cases.
Thus, no transformations to normalize the distributions of the variables had to be made. Tests pointed out that the distributions of the variables can be considered as normal, or when separated in smaller subsets by group could be considered as normal, and that their combined distributions were suitable for further analysis.
5.1.2 Correlation
Before performing the correlation analysis, two dummy variables were created to enable implementation in T-tests and regression analyses. The first was corporate orientation; (0= Stakeholder orientation, 1 = Profit orientation), which indicates the first vignette the subjects read. Then, one was created for corporate conduct (0=good conduct, 1 = bad conduct),
indicating which second vignette that was read. This created combinations sketched in table 6 for the different groups. In case of the correlation analysis, this translates to a negative
correlation with conduct is a correlation with the lowest number; 0, thus good conduct. A negative correlation with orientation means a correlation with stakeholder orientation.
Table 6. Schematic presentation of dummy variables
GROUP ORIENTATION CONDUCT
1 0 0
2 0 1
3 1 0
4 1 1
Table 5 presents the Pearson’s Correlation Coefficients of the variables. Variable correlations can be considered non-existent at 0.0, small at 0.1, medium at 0.3, and large around 0.5 or higher (Cohen, 1992). The results indicate that there is a positive correlation between punishment and orientation (r=0.26, p<0.01**), which means it is positively correlated with profit-orientation, conduct (r= 0.580, p<0.01**), which means it is positively correlated with bad conduct, which were assumptions of this study, and hereby conformed. Negative
emotions before reading the corporate-orientation vignette (Negative emotional resoibse before (r = 0.378, p<0.01**)) and a stronger correlation with negative emotions after reading
