The relationship between firm-orientation and
the reasoning mode of stakeholders:
An empirical approach.
Master Thesis submitted in partial fulfilment of the requirements for the degree of Master of Science.
Djoni de Vos
Graduate student at:
Universiteit van Amsterdam - 10474668 Vrije Universiteit - 2523788
Supervisor:
Dr. Ir. Jan-Willem Stoelhorst Universiteit van Amsterdam
ABSTRACT
Drawing on their theory about the sociocognitive foundations of stakeholder reciprocity (Quintelier, Stoelhorst, & Bridoux, np), we examine if and how firm-orientation activate different reasoning modes in individuals. Through 3 experiments, we investigate if firm-orientation changes reasoning mode (experiment 1), and if this is associated with activation in
the neurological pathways of the Default Mode Network or Task Positive Network (experiment 2 and 3). The results suggest that indeed firm-orientation might influence the
reasoning mode of individuals. Being confronted with a profit-oriented firm caused a significantly lower moral judgement of the business decision of another firm to engage in
misconduct than being confronted with a stakeholder-oriented firm, and this mediated willingness to buy from this firm. The differences are only significant for severe misconduct, and no significant effect on how reasonable this business decision was found. Two subsequent
experiments did not show that this effect was caused by DMN or TPN activation. This thesis contributes to the empirical knowledge of stakeholder theory, provides a possible explanation for the pervasiveness of the separation fallacy of business and ethics, and helps build a more
1. INTRODUCTION
“I’ve often noticed that we are not able to look at what we have in front of us, unless it’s inside a frame.” ~Abbas Kiarostami
A central theorem in stakeholder theory is that stakeholder-oriented firms are able to create more value than profit-oriented firms (Donaldson & Preston, 1995; Jones, 1995; Sachs & Rühli, 2011), because although both firms can establish cooperative relationships with their stakeholders, the stakeholder-oriented firms benefit more from this (Freeman, 1984; Harrison, Bosse, & Phillips, 2010). Stakeholder theorists recognize that human behavior is boundedly self-interested instead of solely self-interested (Jolls, Sunstein, & Thaler, 1998; Bosse, Phillips, & Harrison, 2009), and that most stakeholders are driven by a norm for reciprocity, meaning that they are willing to incur a personal cost to reward fair and punish unfair
behavior (Gouldner, 1960). Stakeholder firms can benefit more from this reciprocal behavior as they usually have fairer policies (Bosse et al., 2009), thus explaining their advantage over profit-oriented firms in creating value. However, in contrast to this idea, stakeholder-oriented firms are more likely to become targets of stakeholder-activism than profit-oriented firms are(Schrempf-Stirling, Bosse & Harrison, 2013), and they encounter both increased reward and increased punishment from their stakeholders, even for similar practices as profit-oriented firms (Briscoe, Chin & Hambrick, 2014; Janssen, Sen & Bhattacharya, 2015). If stakeholders were purely reciprocal regardless of firm orientation, this would not happen, as they would assess the negative firm practice without taking firm orientation into account. But apparently that is not the case, and stakeholder-oriented firms seem to face more reciprocal behavior than profit-oriented firms. An explanation for this could lie in these firm-orientations themselves, and specifically in how they frame the mind of stakeholders.
An affluent actor donating half of his fortune to a charity could be framed as a benevolent hero saving the sick and poor, or as an hypocrite trying to capitalize on a good deed to save his public image. How we frame people and situations influences how we assess them and behave in reaction to them. How we understand a situations affects our expectations of others, our interpretations of the outcome of situations, and our behavior (Messick, 1999). This is the inward out effect of framing, it lies a filter on how we interpret the world. There is also an outward in effect, with contextual cues activating different frames, or different modes of reasoning if you will (e.g. Tenbrunsel and Messick, 1999 who show that bringing money
into a dilemma situation makes people less cooperative and more likely to frame it as a business decision instead of an ethical decision). Using the idea that people in different reasoning modes act differently, Quintelier, Stoelhorst, & Bridoux (np) explain why stakeholders show qualitatively different behaviors in response to firms of different
orientations. Building on knowledge from the neurosciences, they theorize that different firm-orientations put stakeholders into different reasoning modes, and consequently these
stakeholders make different assessments of a firm and their practices, leading to different behaviors in response to practices and events of the firm. These different reasoning modes are associated with two anti-correlated neural pathways, the Default Mode Network (DMN), associated with socioemotional reasoning, and the Task Positive Network (TPN), association with analytical reasoning (Jack et al., 2013). Quintelier et al. (np) argue that through their communications and practices, stakeholder-oriented firms elicit a social context, thus
activating the DMN, whereas profit-oriented firms elicit a non-social context, thus activating the TPN. The activation of these different reasoning modes has consequences for how stakeholders assess actions of a firm: either via an instrumental assessment in the analytical reasoning mode associated with TPN, or via a moral assessment in the socioemotional reasoning mode associated with DMN. Consequently, stakeholders exhibit reciprocal behavior in reaction to a firm’s action in the socioemotional reasoning mode, and self-interested reactions in the analytical reasoning mode. Thus explaining the increased stakeholder activism for stakeholder-oriented firms.
The theory of Quintelier et al. (np) contributes to the literature by providing possible microfoundations for stakeholder behavior. Microfoundations are micro-level behavioral explanations of macro-level phenomenon, and the field of strategy is in need for them (Felin & Foss, 2005). The necessity for understanding the mechanisms underlying stakeholder behavior are clear: aggregate stakeholder behavior can influence firm performance, understanding the individual mechanism can help us predict how they will react on firm practices and thus how the firm will be affected. However, the theory of Quintelier et al. is still just a theory, an idea. Theories are important for building knowledge and cohesion in the field, but until it is tested, a theory is like an ornament on the building of management
knowledge. It is nice to have, but too instable to use as a fundament to expand the building. However, the field of management science is a field that seems to be captivated by theory, some even say it has an overemphasize on theory, leaving to little room for empirical testing (Hambrick, 2007). As an illustration, out of a sample of articles, 46% of theories published in AMR and AMJ were not empirically tested (Kacmar &Whitfield, 2000). The building of
management science seems to have many ornaments, but perhaps too little solid foundations. Therefore, this thesis will empirically test parts of the theory of Quintelier et al. (np), helping further the empirical knowledge of the field of stakeholder theory, and helping build
microfoundations for the field by trying to understand the mechanisms at play. Thus, in this thesis, we will answer the question: “Does firm-orientation trigger different reasoning modes in people, specifically the activation of DMN by stakeholder-oriented and TPN by profit-oriented firms?” We conduct three experiments to do so. In the first experiment, the focus lies on investigating an influence of firm-orientation on reasoning mode, explored via the
variables of moral judgement, reasonableness, and willingness to buy in response to the negative event of a firm. In the remaining two experiments we try to see if the found effect is caused by DMN or TPN activation by respectively stakeholder- and profit-oriented firm, or perhaps by some other mechanism.
The results of the 3 experiments suggest that there is some cause to conclude that different firm orientations elicit different reasoning modes in people who are confronted with them. Experiment 1 for instance shows that individuals confronted with a profit-oriented firm will judge less harsh over the misconduct of a firm than those confronted with a stakeholder-oriented firm. However, the results of experiment 2 and 3 are contradictory, making it impossible to conclude that this is caused by DMN or TPN activation.
This thesis contributes to the building of empirical knowledge in the field of strategy, it provides a possible explanation for the prevalence of the separation fallacy, and helps build a more accurate model of men, also adding understanding to why stakeholders show more activism towards stakeholder-oriented than towards profit-oriented firms.
It contributes to the building of empirical knowledge by developing methods to infer mechanisms that guide behavior and assessment of situations without using an MRI machine. Although it did not necessarily success in findings a method to infer DMN and TPN
activation, by reflecting critically on what happened and what could be improved, steps are made.
The separation fallacy is the concept that business and ethics can be separated, with as a consequence that business decisions and ethical decision can be thought of as different domains (Freeman, 1994). The findings of experiment 1 contributes to a possible explanation of the pervasiveness of the separation fallacy. They suggest that that perhaps there is a
reasoning mode in which humans are less prone to moral thinking. If humans are indeed more likely to reason without ethics in certain reasoning modes, as induces by profit-oriented firms,
this means that perhaps it comes natural to people to separate business and ethics. That could explain why so many people separate business from ethics.
In providing some indication that different reasoning modes can be activated by firm-orientation, and that this impacts how individuals assess moral situations, this thesis
contributes to a more realistic model of men. It adds to the idea of bounded self-interest that this is not a set state, but can be influences by contextual elements. It also details what these contextual elements can be: firm-orientation. This could also help explain why stakeholders show more activism towards stakeholder-oriented firms: they are in a different reasoning mode, making them less susceptible for the moral dimensions of the transgressions of a profit-oriented firm.
This thesis starts out with an overview of the theory leading up to the theory of Quintelier et al. (np). Subsequently we will discuss the importance of empirical testing, and the theory leading up to the conceptual model, including the development of variables to infer reasoning mode without using an MRI. After that, the experiments will be discussed in their totality (hypotheses development, methods, results, discussion) in a consecutive order. We end with an overall discussion of the three experiments taking together, the contributions to the literature and directions for future research. The limitations of each individual experiment are discussed in their own discussion section.
2. THEORY
2.1 The effect of firm-orientation on stakeholder behavior Stakeholder-oriented firms create more value through cooperative relationships
Instrumental stakeholder theory tries to establish a relationship between how a firm sees its stakeholders, and common objectives of firms such as growth and profitability (Donaldson & Preston, 1995). It shows the importance of stakeholders in driving firm performance
(Clarkson, 1995). The idea that stakeholders can influence firm performance is logical, as stakeholders are defined as those who are affected by, and/or can affect the operations and achievements of the firm (Freeman, 1984). They are also described as “the group that the firm needs in order to exist, specifically customers, suppliers, employees, financiers, and
communities” (Dunham, Freeman, & Liedtka, 2006:25). Building on the idea that these stakeholders can influence firm performance, a central theorem in stakeholder theory is that stakeholder-oriented firms are able to create more value than profit-oriented firms (Donaldson & Preston, 1995; Jones, 1995; Sachs & Rühli, 2011), because they are able to benefit more
from the cooperative relationships they built with their stakeholders (Freeman, 1984;
Harrison, Bosse, & Phillips, 2010). This suggests that stakeholders behave differently towards a firm depending on the orientation of this firm, and that these different behaviors can
influence value creation. This raises the question, what is this difference in behavior and how does it benefit firm performance?
Human behavior is reciprocal, not purely self-interested
According to stakeholder theory, the answer to this question lies in reciprocal stakeholder behaviors (Bosse et al., 2009; Bridoux & Stoelhorst, 2016). Bosse et al. describe that, in contrast to the assumption of the completely self-interested actor that is prevalent in economics, a growing number of studies show that most people actually deviate from this purely self-interested model (2009). A more accurate assumptions of individuals is that there are boundaries on the desire to maximize utility caused by norms of fairness, and that they behave in a manner based on reciprocity (Jolls et al., 1998). The observation that most humans actually behave in a reciprocal manner can help us understand why stakeholder-oriented firms are able to benefit more from the cooperative relationships with their stakeholders. In its essence, reciprocal behavior is rewarding kind actions and punishing unkind actions (Gouldner, 1960). In contrast to purely self-interested behavior, reciprocal behavior can come at a cost to the self, were one self-sacrifices for the sake of one’s principle (Bosse et al., 2009). And importantly, in reciprocity individuals are not only concerned about their own reward, but also concerned about the rewards for the other parties (de Quervain, Fischbacher, Treyer, & Schellhammer, 2004; Van Lange, 1999). They take the effect their behavior has on the pay-off of the other party into account, and this effect is not only positive. Whether they want this effect to be positive of negative is dependent on the behavior of the other party (Fehr & Gächter, 2002). Positive reciprocity occurs when reciprocating
individuals perceive the actions of a party as fair, and exhibit rewarding behaviors that increase the pay-off of that party (Rand, Dreber, Ellingsen, Fudenberk & Nowak, 2009). Negative reciprocity occurs when reciprocating individuals perceive the actions of a party as unfair, and exhibit punishing behavior that decreases the pay-off of that party (Fehr & Gächter, 2022). As opposed to self-interested behavior, reciprocating individuals engage in this punishing and rewarding behavior even if it decreases their own pay-off (Henrich et al, 2006). Positively reciprocating stakeholders can drive firm performance, because their behaviors have a positive effect on the firm’s pay-off, over and above they pay-off they formally need to give the firm (Bosse et al., 2009). Therefore it is beneficial for a firm to have
positively reciprocating stakeholders (Bosse et al., 2009; Harrison et al., 2010). On the other hand, negative reciprocating behavior can harm firm performance as findings suggest that individuals are willing to incur greater costs in negative reciprocity than in positive reciprocity (Offerman, 2002). Fair acts are less rewarded then unfair acts are punished.
What are different firm orientations?
As stakeholders reciprocate in reaction to acts of an actor, it seems important how a firm treat its stakeholders. This can be seen as a continuum from completely stakeholder-oriented to completely profit-oriented (Jones, Felps, & Bigley, 2007). Although this orientation can be seen as a continuous scale, research suggests that most firms are either largely profit- oriented or largely stakeholder-oriented, and that this orientation is stable over time (Brickson, 2005; Maignan & Ralston, 2002). Stakeholder-oriented firms aim to distribute their created value fairly and beneficially to all stakeholders, and thus voluntarily implement policies and practices towards this goal (Bridoux & Stoelhorst, 2014). Profit-oriented firms strive to maximize value for the firm owners, usually in the form of maximizing profit (Friedman, 1970). To this extend, they interact with their stakeholders in a much more instrumental manner, and see them as a means-to-an-end.
Why stakeholder-oriented firms benefit more from reciprocity
Bosse et al. argue that firms that manage their stakeholders well and fairly, will reap the positive effects of reciprocity (2009). They argue that firms perceived as distributional, procedurally, and interactionally fair by their stakeholders create more rent and thus have a larger surplus of compensation available for distribution amongst their stakeholders. In other words, more value is created. Distributional fairness is concerned with the pattern of
allocation of material outcomes. Procedurally fairness is concerned with the fairness of rules and procedures in a firm. Interactional fairness is concerned with how actors inside the firm are treated. The fairer stakeholders perceive these three aspects to be, the more they will positively reciprocate and the more benefit the firm will reap from its cooperative relationship with these stakeholders. Experimental research underwrites their point that stakeholder
behavior can be affected by firm behavior and communication. An example of how
motivation and orientation affects behavior is found in Hoffman, Yoeli and Nowak, who show that profit-maximizing actions are likely to elicit less cooperation (2015). They show that we put more trust in those who cooperate without first calculating the costs of cooperation. In a more applied example of how firm behavior influences stakeholders, Cropanzano and
Mitchell show that employees tend to exhibit more positive behaviors and attitudes towards firms if they perceive their exchange relationship with the firm to be more social than transactional (2005). Thus, their practices and communications make that firms that are stakeholder-oriented are more likely to be perceived as fair in a distributional, procedural, and interactional way by their stakeholders (Bosse et al., 2009), and thus reap the positive benefits of stakeholder reciprocity.
Intermediate summary
Thus far we have seen that most stakeholders do not behave purely self-interested, but boundedly self-interested and in a manner of reciprocity (Jolls et al., 1998). As a
consequence, most stakeholders reward fair behavior and punish unfair behavior in firms. It is beneficial for a firm to be rewarded by stakeholders, as they will then increase the pay-off of the firm over and above what is strictly necessary. Firms that have a stakeholder-orientation are concerned with the wellbeing of their stakeholders, and behave and communicate in a way that is usually perceived as fair and thus rewarded by reciprocating stakeholders (Bosse et al., 2009). Because of the benefits of these cooperative relationships with their stakeholders, stakeholder-oriented firms are able to create more value than profit-oriented firms (Donaldston & Preston, 1995; Jones, 1995; Sachs & Rühli, 2011).
Stakeholders behave differently in reaction to the same firm practices, depending on firm-orientation
There is an interesting phenomena, a difference between stakeholder reactions on stakeholder-oriented and profit-stakeholder-oriented firms, that cannot be explained by the reciprocity theory outlined above. Stakeholder-oriented firms are more likely to become targets of stakeholder-activism than profit-oriented firms are (Schrempf-Stirling et al., 2013). Moreover, regardless of their orientation, stakeholder-oriented firms encounter both increased reward and increased punishment from their stakeholders, even for similar practices as profit-oriented firms (Briscoe et al., 2014; Janssen et al., 2015). In other words, it seems like stakeholders react differently to stakeholder-oriented firms then to profit-oriented firms, even if they engage in similar practices. This could indicate that stakeholders engage less in reciprocal behavior for profit-oriented firms than for stakeholder-oriented firms. If stakeholders were reciprocal regardless of firm-orientation, and only looked at concrete practices and actions of firms, they would reward and punish profit-oriented firms the same as stakeholder-oriented firms. It could also be that stakeholders actually assess the practices of stakeholder- and profit-oriented
firms differently, and consequently react differently. As if the stakeholders are in different reasoning modes when interacting with stakeholder- versus profit-oriented firms, therefore make a different assessment of the practices, and consequently act differently.
It is not unthinkable that contextual cues, like the orientation of a firm, changes the frame and reasoning mode with which individuals assess situations and base their behavior on. For example, Molinsky, Grant, and Margolis show in a set of three experiments that activating economic schemas in participants reduces the compassion that they express to others in need (2012). In their experiment, this effect is mediated by dampened feelings of empathy. In other words, the economic schema reduced their feelings of empathy, and this reduce the
compassion they express. It could be that profit-oriented firms activate different schemas in participants as stakeholder-oriented firms, similar as in the experiment of Molinksy et al.
2.2 The effect of firm-orientation on the reasoning mode of stakeholders Different firm-orientations activate different reasoning modes in stakeholders
A theory of Quintelier, Stoelhorst, & Bridoux argues exactly this (np.). Building on
neuroscientific knowledge, they theorize that different firm-orientations put stakeholders into different reasoning modes, and consequently they make different assessments of a firm and their practices, leading to different behaviors in response to practices and events of this firm. They acknowledge, like Bosse et al. (2009) that it is in the nature of humans to be reciprocal, but nuance this view by using neuroscientific evidence that suggests that people are more likely to react reciprocal or self-interested depending on the reasoning mode that is activated (Quintelier, Stoelhorst & Bridoux, np). As such, they theorize that it is not the case that people are either self-interested or reciprocal, but that all people can react both self-interested and reciprocal, depending on which reasoning mode is activated.
The neuroscientific background of different reasoning modes
Research in neuroscience suggests that in the way human brains are structures, two distinct, neural network exist (Jack et al. 2013a). A neural network is a cluster of connected brain areas that are consistently activated and deactivated by certain types of cognitive tasks (Rochford, Jack, Boyatzis & French, 2017). These two, distinct brain networks are anti correlated. Anticorrelated means that if one network is activated above baseline, the other is suppressed to below baseline (Fox et al., 2005). In other words, only one of these networks can be active at any moment.
Activation in one of these networks, the Default Mode Network (DMN) is associated with task requiring social cognition, and activation of the other network, the Task Positive Network (TPN) is associated with tasks requiring physical cognition (Jack et al., 2013). The DMN is called ‘default’ because it is also active when the brain is in resting state (Greicius, Krasnow, Reiss, & Menon, 2003; Schilbach, Eickhoff, Rotarska-Jagiela, Fink, & Vogeley, 2008). It is thus truly the default mode of our brains. The DMN helps us navigate the social world by supporting our socioemotional reasoning mode. This means thinking about one’s own (Xu, Yuan, & Lei, 2016) and other’s mental state and emotions (Amodio & Frith, 2006). It is reliably activated by social tasks like engagement in social narratives (Mars et al., 2012; Van Overwalle, 2011) and important in making moral decision (Jack et al., 2013). Thus, the DMN is a neural network that helps us reason socioemotionally, and is reliably activated by social tasks. Activation of this network by social tasks is coupled by simultaneous deactivation in regions associated with mechanical reasoning (Jack et al., 2013).
The TPN is reliably activated by tasks associated to mechanical reasoning, like visual-spatial reasoning (Corbetta et al., 1998), working memory, attention, goal-directedness (Fox et al., 2005) or abstractions (Baetens, Ma, & Van Overwalle, 2017). It supports our analytical reasoning (Friedman, Jack, Rochford, & Boyatzis, 2015). Activation of this network by mechanical reasoning tasks is coupled by simultaneous deactivation in regions associated with social reasoning (Jack et al., 2013).
As the pathways are anti-correlated, only one reasoning mode can be active at a given time The anti-correlation between the two neural pathways has behavioral implications. It means that individuals are either reasoning socially, or analytically, but not both at the same time. The behavioral implications of this anto-correlation is demonstrated by an experiment of Small, Loewenstein, and Slovic (2007). Participants in the experimental condition were given a set of mathematical problems to solve, in order to activate their analytical reasoning mode. Subsequently they were paid $5,- for their efforts, and a flyer with a picture and story of a distressed individual, and a charity request. Included was an envelope, in which they could anonymously donate to this charity. Participants whose analytical reasoning was activated by the mathematical problems gave significantly less to the charity then the control group, who got a different task to earn their $5,-. This illustrates the anticorrelated nature of both brain networks: if the analytical reasoning mode is activated, people seem to reason less socially and empathetically, and thus donate less to the social cause.
Different firm-orientations activate different reasoning modes
Building on the knowledge of different neural pathways outlined above, Quintelier et al. theorized that different firm-orientations induce different reasoning modes and different behaviors in stakeholders through activation of different neural pathways. Through their communications and practices, stakeholder-oriented firms elicite a social context, which activates the DMN (and thus deactivate the TPN), whereas profit-oriented firms elicit a non-social context through their communication and practices, and thus activate the TPN and deactivate the DMN (Quintelier et al., np).
Different reasoning modes lead to different assessments of practices
The activation of DMN and TPN has, hypothetically, consequences for how individuals assess situations. DMN activation leads to a moral assessment of events, and TPN activation leads to an instrumental assessment of events (Quintelier et al, np). Activation in the DMN is associated with moral decision making (Jack et al., 2013) and with empathy and placing oneself in the shoes of somebody else (Amodio & Frith, 2006). To realize that behavior of firms or people has a moral dimension, it is necessary to understand the perspective and feelings of other people (Friedman et al., 2015). Thus, when the DMN is activated and individuals are in a socioemotional reasoning mode, they have all the ingredients to be aware of the moral dimensions of firm practices. On the other hand, TPN activation is associated with analytical thinking (Friedman et al., 2015) and deactivation of our social reasoning modes (Jack et al., 2013). As a consequence, the moral dimensions of firm practices will be less accessible for individuals in an analytical reasoning mode, and they will assess moral situations more mechanically, more instrumental.
In instrumental evaluations, individuals determine whether an act is good or bad, or if an actor is competent or incompetent. The act if then evaluated in relation to the extrinsic payoff to the evaluator. If this payoff increases with the action, it is evaluated as good. If it decreases, it is evaluated as strategically bad (Declerck, Boone, & Emonds, 2013).
Instrumental evaluations motivate behavior to maximizes payoff to the self (Lyons & Mehta, 1997). On the other hand, in a moralistic evaluation, one assesses whether an act is right (meaning, complying with the salient norm) or wrong (meaning, transgressing this salient norm). A salient norm can be based on rules, laws, and consensus (Treviño, Weaver, & Reynolds, 2006).
The different reasoning modes of stakeholders could explain why stakeholders show more activism in response to stakeholder-oriented firms than profit-oriented firms, even for the same misconduct. If the analytical mode of stakeholders is activated, they will be less likely to recognize the moral dimensions of a firm’s behavior. They will be, compared to when their socioemotional reasoning is activated, less interested in the needs of others, and show less concern for behavior that is harmful to others (Cropanzo, Massaro, and Becker, 2017). As such, they will show less stakeholder activism.
An addition/nuance to the reciprocity argument
The stakeholder literature (e.g. Bosse et al., 2009) tries to explain differences in firm performance by referring to the communications and actions of stakeholder-oriented firms, and noting that they trigger more beneficial cooperative relationships with stakeholders through the process of reciprocity. In practice the situation seems to be more nuanced, as stakeholders react differently on practices of a firm depending on the orientation of that firm. Quintelier et al. go one step deeper in their explanation, by theorizing that the
communications and practices of stakeholder-oriented firms not directly result it more beneficial cooperative relations through reciprocity, but that they activate a different reasoning mode then profit-oriented firms. This socioemotional reasoning mode allows for moral assessments of firm practices, which induces reciprocal behavior and thus results in more punishments and rewards. In the analytical reasoning mode, activated by profit-oriented firms, stakeholders make an instrumental assessment of firm practices, and show behaviors more fueled by purely self-interested motivations. As a result, profit-oriented firms are met with qualitatively different responses from their stakeholders then stakeholder-oriented firms, and the punishing and rewarding behavior of reciprocal stakeholders is irrelevant for profit-oriented firms, as their stakeholders are not in a reciprocal modus (Quintelier et al., np).
2.3 Testing parts of this theory empirically The importance of testing theories empirically
By building on the neurosciences, scholars are building on thoroughly tested findings. Neurosciences is typically a field of empirical testing, researchers back up their theories by experimental findings. This is in contrast to the field of management. Here, theory building is not only a very important building block of the body of knowledge, sometimes it seems the only building block. According to Hambrick (2007), the emphasis on theory building
academic sophistication by trying to gain legitimacy as a science with a commitment to theory. However, were other fields in business have now relaxed their emphasis on theory-building, management seems stuck. Hambrick illustrates this over-emphasize on theory by comparing the use of the word ‘theory’ in management journals with the journal of marketing, the journal of finance, and the journal of accounting. He found that 100 percent of the articles in AMJ, ASQ, and OS contained some variation of the word theory, versus 78 percent of the marketing, finance, and accounting journals. They also used the word more often: 18 times in management articles versus 8 times on average in the non-management journals (Hambrick, 2007). Of course, theory building is important. But until it is tested, a theory is just that, an idea. To return to the building blocks example, in the field of management we are building whole houses with stones that only exist in the minds of researchers. One does not need to be a builder to know that illusionary building blocks are not the best foundation. And it does seem to be the case that most building of knowledge in the field of management is done with untested theories. Kacmar and Whitfield (2000) did a study looking at the quality of citings in the field of management. Taking a sample of articles out of AMR and AMJ, they traced all the articles that cited these sample articles, and noted how the source article was used. Their research shows for instance that only 9% of articles that were citing an article out of AMR were testing this theory empirically, and only 6% of the articles that were citing an article out of AMJ were testing this theory empirically. So in 91% of the time for AMR and 94% of the time for AMJ, theories were cited just as a theory, as a building block of knowledge. More alarmingly, out of the sample of AMR articles that Kacmar and Whitfield tested, 46% were actually never tested empirically. This means that firstly, many theories in management literature are never tested, and secondly, that other scholars are citing these untested theories and using them as building blocks for new theories. It seems important to test theories in management, to discover what ideas are in line with empirical observations and which ideas are not. Therefore, I will not continue building on the theory of Quintelier et al., but will use this thesis to first test part of their theory.
Towards a research question
An essential part of their theory seems to be that different firm-orientations put stakeholders into different reasoning modes, namely DMN and TPN. And these different reasoning modes induce a chain of consequences in the form of different assessments of firm practices,
different stakeholder behavior etc. But the first step and biggest new insight for the current stakeholder and reciprocity theory is that different firm-orientations put stakeholders in
different reasoning modes. Therefore, the research question of this thesis will be: “Does firm-orientation trigger different reasoning modes in people, specifically the activation of DMN by stakeholder-oriented and TPN by profit-oriented firms?”
Inferring reasoning modes without an MRI
To test the activation of DMN and TPN in people empirically, it would be easiest to use an MRI to scan participants brains whilst they are confronted with a profit- or stakeholder-oriented firm. Unfortunately, for this thesis, access to an MRI was not possible. Because we cannot look at direct brain activity, we have to use other methods to infer if participants are in different reasoning modes. We do this using the knowledge there already is about being in the different reasoning modes. For instance, we know that DMN activation and being in the socioemotional reasoning mode is probably needed to make moral assessments of situations, and that individuals in an analytical mode have less access to moral dimensions of situations (e.g. Jack et al., 2013) and feel less concern for the other, thus be less bothered by behavior that is harmful to others (Cropanzano et al., 2017). Therefore, we are going to infer activation of the different reasoning modes from a moral judgement variable. Our assumption is that individuals in an analytical reasoning mode will judge situations as being less morally wrong than their socioemotional reasoning counterparts, as they are simply not in a modus to assess situations morally.
We will also infer reasoning mode from the idea that people in a socioemotional reasoning mode find immoral business decisions less reasonable then those in an analytical reasoning mode. The idea being that the socioemotional reasoning mode will make people identify with the victims and feel empathy for, and therefore they will not find the immoral decision of a firm reasonable. Those with an activated analytical reasoning mode, however, will reason more ‘coldly’ and understand why the firm made the decision it made.
Lastly, a variable of behavioral intention is included in the form of Willingness to Buy (WtB). If individuals in a socioemotional reasoning mode are indeed more prone to
reciprocity, then this could affect their Willingness to Buy from an immoral company, as they can engage in punishing behavior by reducing their buys. Individuals with an active analytical reasoning mode should not be prone to this type of reasoning, and thus the immoral actions of a company should not affect there WtB. Moreover, a mediation effect of the moral judgement and the reasonable decision variable could be present on WtB.
Figure 1, conceptual model
The research question actually has two parts: does firm orientation trigger different reasoning modes in people, and is this caused by activation of DMN and TPN? Therefore, we have designed 3 experiments to address these two different questions. The first experiment will focus mainly on the different reasoning modes, and the remaining two will try to infer if different firm-orientations activate DMN and TPN. Each experiment will be discussed in its entirety (hypotheses, methods, results, discussions) in a subsequent order.
3. EXPERIMENT 1
3.1 Hypotheses development
The aim of experiment 1 is to answer part of the research question, by investigating the effect of firm-orientation on the reasoning mode that is activated in people. For clarity, I will first introduce the propositions that we are using in this experiment, and subsequently I will
introduce how we have operationalized these propositions in the form of concrete hypotheses. We want to find out if different firm orientations lead to the activation of different reasoning modes, and based on theory we expect stakeholder-oriented firms to elicit a social context and thus activate a socioemotional reasoning mode, and profit-oriented companies to elicit a non-social context and thus activate an analytical reasoning mode. This leads to our first two propositions:
P1: Stakeholder-oriented firms activate a socioemotional reasoning mode in individuals confronted with them.
P2: Profit-oriented firms activate an analytical reasoning mode in individuals confronted with them.
Inferring reasoning modes without an MRI
To test the activation of DMN and TPN in people empirically, it would be easiest to use an MRI to scan participants brains whilst they are confronted with a profit- or stakeholder-oriented firm. Unfortunately, for this thesis, access to an MRI is not possible. Because we cannot look at direct brain activity, we have to use the knowledge already present in literature about the different reasoning modes to infer in what reasoning mode participants are.
We know that DMN activation and being in the socioemotional reasoning mode is probably needed to make moral assessments of situations, and that individuals in an analytical mode have less access to moral dimensions of situations (e.g. Jack et al., 2013) and feel less concern for the other, thus are less bothered by behavior that is harmful to others (Cropanzano et al., 2017). . Because we are mainly interested in stakeholder judgement with regards to firms, we are going to use a morally wrong event induced by a firm (a negative event) as a situations our participants have to judge, to infer reasoning mode. Cropanzano et al. (2017) argues that a sense of empathy for those effected by a firm’s behavior is necessary to apply moral principles to evaluate a behavior of firms. Therefore, our assumption is that individuals in an analytical reasoning mode will judge situations as being less morally wrong than their socioemotional reasoning counterparts, as they are simply not in a modus to assess situations morally. This gives us the first hypothesis:
Hypothesis 1: Being confronted with a stakeholder-oriented (profit-oriented) firm will
increase (decrease) the negative moral judgement a stakeholder gives of the negative event of another, neutral firm.
A second variable to infer reasoning mode is how reasonable of a business decision
participants find the negative event of this company. As the socioemotional reasoning mode assists empathy and putting oneself in the shoes of others, we assume that individuals in this reasoning mode will be more likely to think the decision of the firm is not reasonable. However, as individuals in the analytical reasoning mode will reason more mechanical, they might understand why the firm made the decision it made regardless of the negative
Hypothesis 2: Being confronted with a stakeholder-oriented (profit-oriented) firm will decrease (increase) how much of a reasonable business decision a stakeholder finds the negative event of another, neutral firm.
The choice for a negative event of a neutral firm
In the hypotheses mentioned above, participants are not judging the immoral action of a profit- or stakeholder-oriented firm, but the immoral actions of a neutral firm. This merits an explanation. The aim of this research is to discover if firm-orientation causes a difference in reasoning mode. The experiment should thus be designed in such a way that the potential effects found on moral judgement and reasonable business decision are mainly caused by a different reasoning mode, and not by some confounding variable. To avoid potential confounding effects of firm-orientation on moral judgement, we have chosen to have
participants judge the negative event of another, neutral firm, and not the stakeholder-oriented or profit-oriented firm they are confronted with earlier in the experiment. If firm-orientation really activates different reasoning modes, then the differences in assessment should also apply to the negative event of another neutral company, since participants are set into a certain mindset and this neutral company is probably not taking them out of that mindset. In doing so, we really test the effect of reasoning mode only, and not possible reactions to the discrepancy between what a firm says it finds important (e.g. stakeholders) and what it
actually does (e.g. something negative to its stakeholders). Thus, it seemed the cleanest design choice to go for another neutral company as these possible effects are avoided and the
likelihood that a difference in behavior and assessment is caused by a different mode of reasoning is increased.
Reciprocal punishing behavior
A second aspect we want to address with this experiment is the reciprocal punishing behavior of people, caused by activation of different reasoning modes through differences in firm-orientation.
P3: As a result of an activated socioemotional reasoning mode, stakeholders will engage in punishing behaviors towards companies that engage in morally wrong practices
P4: As a result of an activated analytical reasoning mode, stakeholders will not engage in punishing behaviors towards companies that engage in morally wrong practices
To operationalize this, we included a variable to measure behavioral intention, the Willingness to Buy (WtB). One way of punishing behavior is to boycotting, in which
stakeholders refuse to buy products from a certain company, with a personal cost of having to go somewhere else to buy the products they need (Hahn & Albert, 2017). The idea of not buying products to punish a firm is operationalized in the WtB variable. Because this
boycotting behavior is in the theory caused by the difference in reasoning mode, this leads to the following hypotheses:
Hypothesis 3: The more negative the moral judgement of a stakeholder is, the lower his/her willingness to buy
Hypothesis 4: The more reasonable a stakeholder finds a business decision, the higher his willingness to buy
Thus, the model of Experiment 1 looks like this:
Figure 2, model of experiment 1
3.2 Methods Materials
To operationalize the idea of being confronted with a stakeholder- or profit-oriented firm, a vignette was designed. Quintelier et al. (np) theorize that DMN and TPN are activated by firm
oriented through the differences in communication and practices of firms in different orientations. As a vignette can describe both he practices of a firm but also use the type of communication and words a firm would use, it seems very suitable. Moreover, it allows for both a highly controlled study, but offers more realism then some other techniques (e.g. a survey) by allowing for the representation of contextual elements (Atzmüller & Steiner, 2010). Both conditions needed to be comparable, thus as similar as possible, but vary in the most important aspect: the orientation of the firm.
Thus, 4 profit and 4 stakeholder vignettes were developed. . The vignettes are +- 100 words long, and about anonymous firms (e.g. ‘an airline company…’). To make the vignettes as naturalistic as possible, we sourced words and sentences that are actually used on the websites and in yearly reports of profit-oriented and stakeholder-oriented firms. To decide of a company was profit- or stakeholder-oriented, we looked at what the companies said about themselves, and what the media said about these companies. We looked at if and how they talked about their company goals, their employees, their stakeholders, how elaborate
information about stakeholders was, and with what kind of news they were in the media. One example was found in prior literature, were Bridoux and Stoelhorst (2014) classify Rayainar as a profit-oriented firm and Southwest as a stakeholder-oriented firm. To make the
conditions as similar as possible, we used the same types of firms in both conditions (an airline, a financial service provider, a food company, and an oil and gas company). Ryanair, Goldman Sachs, Kraft-Heinz and Exxon-Mobile were selected as profit-oriented companies. SouthWest airlines, Triodos Bank, Unilever and GoodFuels were selected as their stakeholder counterparts. We copy-pasted sentences from different places on their website into one, smoothly readable vignette. To ensure that the participants read the vignettes attentively, a multiple choice question about the content of the vignette was asked after each vignette. The multiple choice question asked after each vignette was the same for both conditions (i.e., the question and answer possibilities for the stakeholder airline vignette was the same as that of the profit airline vignette). The Flesch reading ease score for the profit vignettes was 40.3 and for the stakeholder vignettes it was 45.0. All the vignettes can be found in Appendix 8.1 and 8.2.
Apart from the vignettes, two negative events were written for neutral companies. They were based on real events of real companies, but altered for the sake of the experiment.
Company X is a clothing manufacturing company. They have a shop close to your house and you regularly check out their collection. You are satisfied with the value for money you get there, and the style suits you.
Three weeks ago, extreme weather in Southeast Asia has damaged one of the factories of company X. As a result, toxic chemicals leaked into a river. Company X discovered the leak the next day, and took measures to close it. However, they did not inform local authorities of the leak, afraid of bad publicity. The World Watch Institute has now discovered this, and is publishing in the media about it.
The Company Z negative event
Company Z is a grocery retailer selling products you need on a weekly basis. The price and quality of their products meet your expectations. It is also conveniently located close to your house.
Recently, Company Z received some media attention. An investigation by Greenpeace discovered that the vegetables that the company sells as ‘locally produced’ are actually shipped in from Spain. This is cheaper for Company Z, and helps them to keep vegetable prices low.
Procedure
Participants were seated in a computer lab, in separate cubicles to ensure the least possible distraction. The computer program randomly assigned them to either the profit or the
stakeholder condition, in which they read the 4 vignettes of the orientation of their condition. After each vignette they had to answer a multiple choice question. Continuation to the next vignette was only possible when this question was answered correctly.
After reading the 4 vignettes, participants were confronted with the negative events of two other, neutral firms. They were asked, on a scale of 0 – 100, how morally wrong they thought this action was (as used by Wheatley & Haidt, 2005; Eskine, Kacinik, & Prinz, 2011), and on a scale of 0 – 100 how reasonable they thought the business decision of this company was. Subsequently their willingness to buy products of this company was assessed using a 4-item scale that can be found in the appendix (adapted from White, MacDonnell, & Ellard, 2012).
The order of the 4 vignettes was randomized. The order of the negative events was counterbalanced.
The control condition consisted of a separate set of students who only read the two negative events and answered the questions about them. This was combined with experiment 2 and 3, so the participants of experiment 2 and 3 first got the negative events of experiment 1 to act as the control group, and subsequently continued with experiment 2 or 3.
Participants
121 participants were recruited to participate, all students of the business faculty of the University of Amsterdam. 64 were randomly assigned to the stakeholder group (36 males, 27 females, and 1 other gender, age range 17 – 27) and 57 were randomly assigned to the profit group (31 male and 26 female, age range 18 – 29). All participants provided written consent for their participation, and were rewarded in the form of credits, which they have to collect as part of their degree requirements. The group consisted of both Dutch nationals and
internationals, and the experiment was conducted in English. The majority of participants did not have English as a native language, but as all the students followed courses in English, their level was sufficient. On a scale of 1 – 7, average self-reported English proficiency was 6.03 (SD = 0.84).
3.3 Results Control group
The control group consisted of 2 separate groups of 60 and 58 participants, who subsequently continued with experiment 2 and 3 respectively. T tests on all the variables of interest show that no difference in mean between the two groups is significant. Therefore the two datasets are added together, and jointly form the control condition data with N=118. The t tests and means of both groups can be found in the appendix.
Data cleaning
There were no extreme values in the data for any of the conditions. That makes sense, as all the questions used either a 1 – 7 Likert scale or a 0 – 100 slide scale, and answers were given in all of the possible answering ranges. As the scales were ‘closed off’ (lowest and highest value are set), extreme outliers are unlikely.
Histograms to visually see the spread of answers for all the different variables for all the different conditions can be found in appendix 8.10. Figure 3 below gives an example of difference in spread found for the moral judgement variable belonging to the negative event of Company X.
Figure 3, histograms of the moral judgement score for Company X
As an example, in figure 3 some differences in spread between the groups becomes apparent. For instance, nobody of the stakeholder group found the negative event to be not morally wrong at all (scoring between 0 – 20) whereas 9 people in the profit group found the event not morally wrong at all (scoring between 0 – 20). And many more people in the stakeholder group found the event very morally wrong (scoring between 80 – 100). The histograms show that the data is not exactly normally distributed, but skewed to the left. A Shapiro-Wilk test for all the different variables shows that they are indeed not normally distributed (p<0.05 for all variables). However, since the t-test is used, this is not a major problem. Although the assumption of normality is taught widespread in statistics textbooks, the belief that a t-test is only valid on normally distributed outcomes is incorrect (Lumley, Diehr, Emerson, & Chen, 2002). A t-test usefulness lies in that is produces valid results even for small samples if they are normally distributed. However, if a sample size is sufficiently large, the t-test will work for any distribution (Lumley et al., 2002). Sullivan and d’Agostino (1992) for example, found that the t-test produces significance levels very close to the normal significance levels even for small samples (50 or less) that are very skewed (half of the sample scoring only a threshold value of 0). Therefore, we will proceed with using the t-test to compare means of the different conditions, even though the data is not normally distributed.
Reliability of WtB variable
As WtB was measured on a 4-item scale, Cronbach’s alpha was used to measure the
consistency of the scale. In all conditions and for both negative events Cronbach’s alpha was >.88for the WtB items, thus the scale can be seen as consistent. Therefore, only the mean score of the 4 items will be reported as a measure of WtB.
The negative events are treated as different
Participants are asked to give their moral judgement, reasonable business decision score, and WtB in response to two negative events of two separate companies. Because the nature of these events is different and because the mean scores for all variables are significantly different for the two events in the control group, (e.g., for moral judgement, t test on mean difference between Company Z and X: (t = 3.4816, df = 226.37, p-value = 0.000598), we will treat these events as different. Mean scores are thus reported for each variable for each event, instead of taking the mean scores of both events.
Mean scores for each variable
The mean scores of each group for each variable are summarized in table 1, for both negative events.
Table 1
Negative event Company X Negative event Company Z
Moral Reas. WtB Moral Reas. WtB
Stkhldr 78.58 (19.45) 48.28 (26.48) 3.34 (1.71) 68.17 (27.3) 51.29 (26.57) 3.59 (1.68) Profit 66.65 (29.26) 50.68 (29.54) 3.57 (1.75) 63.96 (25.8) 55.31 (27.6) 3.94 (1.84) Cntrl 76.34 (19.90) 44.03 (27.17) 3.48 (1.21) 66.36 (23.96) 53.45 (29.14) 3.84 (1.39) Moral judgement
For the event of Company X, the difference between the mean moral judgement score of the stakeholder and profit group is 11.93 points on a 100-point scale (78.58 – 66.65), and this difference is significant (t = -2.6076, df = 95.595, p-value = 0.01058). The mean moral
judgement score of the control group lies close with that of the stakeholder group, their 2.24 point difference is not significant (t = -0.73555, df = 131.95, p-value = 0.4633). The
difference between the control and the profit group is, with 9.69 points, larger and significant (t = 2.2606, df = 81.873, p-value = 0.02644).
The moral judgement with regards to the negative event of Company Z shows the same direction of effect, but the differences are smaller. The mean moral judgement score difference between the stakeholder and profit group is 4.21 points (68.17 - 63.96), this difference is insignificant (t = -0.87124, df = 118.57, p-value = 0.3854). The control group scored in between the stakeholder and profit group, and thus the differences between the control group and the stakeholder and profit group are smaller and also insignificant (t = -0.44696, df = 115.78, p-value = 0.6557 and t = 0.5879, df = 103.77, p-value = 0.5579 respectively). The differences are thus in line with hypothesis 1, but only significant for the negative event of Company X.
Reasonable business decision
For the reasonable business decision variable, the direction of effects lie in line with hypothesis 2, but the differences are small and insignificant for both negative events. The difference between the profit and stakeholder group is 2.40 points (50.68 – 48.28) for Company X (t = 0.46892, df = 113.27, p-value = 0.64) and 4.02 points (55.31 – 51.29) for Company Z (T-test: t = -0.87124, df = 118.57, p-value = 0.3854). For the event of Company X, the control condition scores lower then both the profit group (6.65 points lower) and the stakeholder group (4.25 points lower). These differences are insignificant (t = -1.434, df = 102.92, p-value = 0.1546 and t = -1.0258, df = 132.28, p-value = 0.3069 respectively). For the event of Company Z, the control condition scores in between the stakeholder group (2.16 points higher) and profit group (1.86 points lower). These differences are insignificant (t = 0.52602, df = 139.93, p-value = 0.5997 and t = -0.41165, df = 116.39, p-value = 0.6814 respectively).
Willingness to Buy
The direction of effects for WtB lies in line expectations for both negative events, but the differences are insignificant. For the negative event of Company X, the stakeholder group scores 0.23 points lower (on a 7-point scale) then the profit group, this is insignificant (t = 1.4614, df = 472.22, p-value = 0.1446). For the negative event of Company Z, the willingness to buy of the stakeholder group is 0.35 points lower than the profit group, this is insignificant
(t = -1.3826, df = 112.97, p-value = 0.1695). In both cases, the control condition scored somewhere in the middle between the profit and stakeholder group. The control condition scored 0.14 higher than the stakeholder group (t = 0.67223, df = 117.24, p-value = 0.5028) and 0.09 lower than the profit group (t = -0.41033, df = 92.612, p-value = 0.6825) for the negative event of Company X. For the negative event of Company Z, the control group scored 0.25 higher than the stakeholder group (t = 1.1954, df = 135.84, p-value = 0.234) and 0.10 lower than the profit group (t = -0.44491, df = 105.16, p-value = 0.6573).
Although only the moral judgement of Company X scored a significant difference, all the directions of effect found are in line with the hypotheses. Participants in the stakeholder group judge harsher over the negative events, have a lower willingness to buy from this company and find the business decision less reasonable compared to participants in the profit group. The differences between these groups are larger for the negative event of Company X compared to that of Company Z. The only real significant difference is between the moral judgement of stakeholder and profit, and control and profit in the negative event of Company X. Usually, the control group has a mean score between that of the profit score and
stakeholder score. The only exception is the reasonable decision judgement of company X, where the control group scores lower then both the profit and stakeholder group.
The relationship between Moral Judgement and Reasonable Business decision, and WtB A simple regression analysis was calculated to predict WtB based on Moral Judgement. A significant regression equation was found for both negative events, with (F(1,119) = 60.71, p<0.01), with an R2 of 0.34 for the negative event of Company Z, and (F(1,119) = 43.08, p<0.01) with an R2 of 0.27 for the negative event of Company X. Participants’ WtB from Company Z decreased 0.03 for each extra point of moral judgement, and participants’ WtB from Company X decreased 0.03 for each extra point of moral judgement.
Another simple regression analysis was calculated to predict WtB based on Reasonable Business Decision. A significant regression equation was found for both negative events, with (F(1,119) = 29.46, p<0.01), with an R2 of 0.20 for the negative event of Company Z, and (F(1,119) = 18.80, p<0.01) with an R2 of 0.14 for the negative event of Company X.
Participants’ WtB from Company Z increased 0.02 for each extra point of reasonable business decision, and participants’ WtB from Company X increased 0.02 for each extra point of reasonable business decision. These results are in line with hypotheses 3 and 4.
Besides these direct effects between the variables, the conceptual model for experiment 1 also include mediating effects of moral judgement and the reasonable business decision judgement on WtB. In a mediation analyses, first the main effect is tested. In our case, this is the effect of being in either the stakeholder of profit group, on the WtB from either Company X or Z. For Company X, the regression shows, in line with previous t tests, no significant effect of firm orientation on WtB (F(1, 199) = 0.784, p = .378), with an R2 of 0.007. The whole output for the following mediation analysis can be found in the appendix.
Some people argue that if the first step of testing a mediation is not significant, it is no use to test a mediation (Baron & Kenny, 1986). However, there are also researchers who say that if there is a theoretical ground to assume a mediation, it is fine to go ahead with the analyses (e.g. Shrout & Bolger, 2002). Since we have some theoretical grounds for this mediation, we will continue.
In the second step of testing a mediation, the link between the independent variable (stakeholder or profit condition) and the mediator (moral judgement of event) is tested. For Company X, this proved to be in line with our t test, and thus significant (F(1,199) = 7.114, p < 0.01) with an R2 of 0.056.
In the last step the effect of the mediator on the dependent variable and the effect of the independent variable on the dependent variable are tested in the same regression. Results indicate that firm-orientation was a significant predictor of moral judgement (b= -11.929, SE = 4.472, p<0.01), and that moral judgement was a significant predictor of WtB (b= -0.030, SE = 0.005, p<0.01). These results support a partial mediation effect of moral judgement on WtB as the magnitude of effect for condition on the willingness to buy did get smaller (from b = 0.2302 to b = -0.125).
To test the significance of the indirect effect, a bootstrapping estimation approach with 500 samples was used. These results indicated that the indirect coefficient was significant (b=0.356, p = 0.016). Complete tables of all these tests can be found in the appendix.
For the negative event of Company Z, such an indirect effect is not found. The first step of the mediation analysis, the effect of firm-orientation on WtB, is as expected
insignificant (F(1,119) = 1.937, p = 0.167). The next step, the effect of condition on moral judgement is, in line with the earlier t-test, also insignificant (F(1,119) = 0.754, p = 0.387). Therefore, it is no use continuing with the third step to test a mediation, as there is no significant mediation present. The complete tables of these regressions can be found in the appendix.
Mediation analysis for reasonable business decision
The same holds for the mediation of the reasonable business decision variable for both events. The regression between firm-orientation and reasonable business decision is insignificant for both events (F(1,119) = 0.227, p=0.638) for the Company X negative event and (F(1,119) = -.696, p=0.406) for the Company Z negative event). Therefore a mediation is not present. The complete tables of these regressions can be found in the appendix.
Multiple mediation
Using structure equation modeling (SEM), we finally test multiple mediation for both
variables. In this path model both possible mediations paths (via reasonable business decision making and moral judgement) are included, thus the whole conceptual model as depicted in figure 1 is included. For the negative event of Company X, this analysis shows, like the previous mediation analyses, that the mediation path via moral judgement is significant (b=0.309, p = 0.015), and the mediation path via reasonable decision is not
(b=0.030,p=0.638). The total effect of both paths is also insignificant (b=0.230, p=0.372). The whole table can be found in the appendix.
The relationships between the variables for the negative event of Company X is summarized in figure 4.
The same analysis for the negative event of Company Z shows that, in line with the previous mediation analyses, no indirect paths are significant (b=0.105, p = 0.387 for the indirect effect of moral judgement, and b=0.044, p=0.426 for the indirect effect of reasonable decision), nor is the total path (b=0.353, p=0.160. The table for this analysis can be found in the appendix.
The relationships between the variables for the negative event of Company Z is summarized in Figure 5.
Figure 5 – relationships for the negative event of Company Z
The order effect of the events
Another interesting aspect of the data is the effect of the order in which the negative events are shown to participants. It could be that the effect of reading the vignettes diminishes over time. In other words, it could be that if the DMN of a participant is activated during the vignette-phase of the experiment, this activation dies changes after reading the negative event and answering the 6 questions that belong to that negative event. Or it could be that reading and thinking about a negative event puts people in another mode of reasoning.
As the order of events was counterbalanced (half of the participants of each condition first read the Company X negative event and subsequently the Company Z negative event, and half
vice versa), we can investigate its effect by for instance comparing the moral judgement scores of those that saw company X first and those that saw company X second. Table 2 includes all the mean scores for the different order groups.
Table 2
Negative event X – overall Order X – Z Order Z – X
Moral Reas. WtB Moral Reas. Wtb Moral Reas. Wtb Stkhldr 78.58 (19.45) 48.28 (26.48) 3.34 (1.71) 82.91 (17.07) 53.22 (27.15) 3.18 (1.79) 74.25 (20.99) 43.34 (25.25) 3.50 (1.34) Profit 66.65 (29.26) 50.68 (29.54) 3.57 (1.75) 70.07 (27.57) 50.75 (26.27) 3.32 (1.70) 63.11 (31.01) 50.61 (33.07) 3.83 (1.56) Cntrl 76.34 (19.90) 44.03 (27.17) 3.48 (1.21) 79.05 (20.74) 49.27 (28.6) 3.46 (1.16 73.63 (18.81) 38.78 (24.81) 3.5 (1.28)
Negative event Z – overall Order X – Z Order Z – X
Moral Reas. WtB Moral Reas. Wtb Moral Reas. Wtb Stkhldr 68.17 (27.3) 51.29 (26.57) 3.59 (1.68) 72.91 (25.3) 49.84 (27.74) 3.62 (1.78) 63.44 (28.77) 52.56 (25.72) 3.56 (1.59) Profit 63.96 (25.8) 55.31 (27.6) 3.94 (1.84) 68.45 (23.52) 53.86 (27.11) 3.79 (1.93) 59.32 (27.61) 56.82 (28.52) 4.1 (1.74) Cntrl 66.36 (23.96) 53.45 (29.14) 3.84 (1.39) 67.15 (22.33) 52.42 (27.78) 3.8 (1.37) 65.56 (25.66) 54.47 (30.65) 3.77 (1.45)
The data in table 2 does seem to point towards some kind of order effect for the moral
judgement score. The stakeholders who first read about the negative event of Company X (n = 32) score on average 8.66 points higher than the stakeholder group who read this event second (n=32, 82.91-74.25). This difference is not significant (t = 1.8122, df = 59.454, p-value = 0.075), but very close and the sample size for the order effect is not that large (n = +-30). The same direction of effect is seen in the profit group. Those who view Company X first (n=29) score on average 6.96 points higher than those who read about Company X second (n=28, 70.07 – 63.11). This difference is insignificant (t = 0.89468, df = 53.748, p-value = 0.3749). The control group echoes this finding, participants reading about Company X first (n=59) score on average 5.42 higher on the moral judgement than those reading about is second
(n=59, 79.05 – 73.63), also an insignificant difference (t = 1.4879, df = 114.91, p-value = 0.1395).
For the negative event of Company Z, the effect is reversed. In all groups with participants that read about Company Z first, it scores on average a lower moral judgement score than with the participants that read about Company Z second. In the stakeholder group, it scores 9.47 points lower if viewed first (n=32) compared to second (n=32, 72.91-63.44). This difference is not significant (t = -1.3981, df = 61.006, p-value = 0.1672). For the profit participants, the mean moral judgement score is 9.14 points lower when reading first about Company Z (n=29) compared to those reading about it second (n=28, 68.45-59.31). This difference is not significant (t = -1.3411, df = 53.002, p-value = 0.1856). In the control condition, the difference is 1.59 points between those who read it first (n=59) and those who read about it second (n=59, 67.15-65.56).
The difference between groups stays the same, regardless of order of events
It becomes apparent that although there might be an order effect, through this order effect the direction of difference in the means lies in in line with our hypotheses. The difference
between the experimental groups stays approximately the same, regardless of whether the event is assessed first or second. The mean difference between profit and stakeholder who first see Company X is 12.84, mean difference when seeing Company X second is 11.14. For Company Z this is 4.46 when first seeing company Z and 4.11 when getting Company Z second.
Order effect for WtB and reasonable business decision
The order effect with respect to the WtB seems to be a consequence of the order effect of moral judgement for Company X: were the moral judgement goes down as Company X gets judged second, the willingness to buy goes up. From 3.18 to 3.50 in the stakeholder group, from 3.32 to 3.82 in the profit group, and from 3.46 to 3.50 in the control condition. Curiously enough the difference is larger in the profit condition then in the stakeholder condition, but none of these difference is significant. For the negative event of Company Z, there is no clear pattern in the order effect on WtB.
