Why do we cross the line? : assessing the effect of norm violations on imitating behavior within an organizational context

(1)

Why do we cross the line?

Assessing the effect of norm violations on imitating behavior

within an organizational context

Master thesis

Name student: Britt van Ballekom Student number: 5876079

MSc Business Studies – Leadership & Management track Version: Final version

Date: June 30th, 2014 Supervisor: N.M. Blaker

(2)

Abstract

Corporate malfeasance has earned a place among the defining themes of the last decade. Numerous organizations participate in norm-violating behavior, often leading to millions of deceived consumers or employees within the organization. Therefore, this study aims to get a better understanding of the consequences of norm violations within an organizational context. The main effects between norm violations, the psychological barrier, and imitation are examined. Furthermore, the mediating role of the psychological barrier and the moderating role of status motivation is tested. In order to examine this topic, an online survey is distributed among 63 Dutch respondents. The results show a positive effect of norm violations of others on the psychological barrier to exercise the same behavior and a positive effect of the psychological barrier on imitation. Also, a mediating effect of the psychological barrier is found when it is tested for full mediation. No significant moderating effect of status motivation is found. More research on norm-violating behavior in the organizational context is needed. Little attention has been paid to the potential negative impact of norm violations on the overall organizational performance.

(3)

2.3 Leadership ... 14 2.3.1 Influence ... 15 2.3.2 Power ... 16 2.4 Status motivation ... 18 2.5 Psychological barrier ... 19 2.6 Synthesis ... 21 3. Method ... 23 3.1 Research design ... 23 3.2 Data collection ... 24 3.3 Participants ... 26 3.4 Measures ... 26 3.4.1 Differential treatment ... 26 3.4.2 Norm-violating scenarios ... 27 3.4.3 Psychological barrier ... 27

3.4.4 Imitation of norm violations – Dependent variable ... 27

3.4.5 Status motivation ... 28 4. Analyses ... 29 4.1 Analysis strategy ... 29 4.1.1 Mediation analysis ... 29 4.1.2 Moderation analysis ... 30 5. Results ... 31 5.1 Descriptive statistics ... 31 5.1.1 Sample characteristics ... 31 5.2 Reliability ... 32 5.3 Correlations ... 32 5.4 Differential treatment ... 33 5.5 Mediation analysis ... 34 5.6 Moderation analysis ... 35

(4)

6. Discussion ... 41

6.1 Discussion of relationships between variables ... 41

6.1.1 Relationship between norm violations of others and the psychological barrier ... 41

6.1.2 Relationship between the psychological barrier and imitation ... 44

6.1.3 Mediating effect of the psychological barrier ... 46

6.1.4 Moderating effect of status motivation ... 47

6.1.5 Prosocial versus selfish norm violations ... 49

6.2 Theoretical implications ... 50

6.3 Managerial implications ... 50

6.4 Limitations and suggestions for future research ... 51

7. Conclusion ... 52 References ... 53

(5)

1. Introduction

Many theories have been generated and advanced over the years in order to explain why people behave as they do. A fundamental moment in the development of behavioral theory was the introduction of Bandura’s Social Learning Theory (Bandura, Ross & Ross, 1963; Bandura, 1969; Bandura, 1977). This theory emphasizes that new responses may be rapidly acquired and existing behavioral repertoires may be considerably changed as a consequence of observing the behavior and attitudes exhibited by other people acting as ‘models’. This type of learning is generally labelled as “imitation” in behavioral theory. Although Social Learning Theory and imitating behavior are both well-examined constructs, little attention had been paid to these topics with regard to norm-violating behavior and power affordance and, more specifically, in a business context. If indeed leaders or other employees within an organization can act as role models, as the Social Learning Theory states, it is necessary to examine the consequences if these role models practice norm-violating behavior. Obviously, the imitation of desired behavior has positive consequences for the organization, but the copying of norm-violating behavior could result in the deterioration of overall organizational performance.

In order to examine imitating behavior, it is important to find reasons why people would copy the behavior of others. The motivation to gain status or power could be one of those reasons. People having the incentive to develop themselves into a more powerful position within an organization, might use their leaders as role models and eventually copy their behavior to gain status. The question of what makes people rise to power has long been of interest to social scientists (Van Kleef et al., 2012). Earlier research has attempted to find an answer to this question by examining static features of the individual, such as physical appearance, gender, and ethnicity (Anderson et al., 2001; Berger et al., 1972). Other studies

(6)

examined the effect of leaders’ vocal attractiveness on leader effectiveness (DeGroot et al., 2011).

On the contrary, increasingly more studies argue that power affordance is a dynamic process instead of a static one (Anderson et al., 2001; Keltner et al., 2008). These studies emphasize the idea that individuals are granted power based on their social behavior. For example, Van Kleef et al. (2012) examined the possibility that power is afforded to individuals who are willing to break the rules to benefit their group. They conclude that specific types of norm violations can lead to power affordance.

Weiss (1977) discussed Bandura’s (1969) Social Learning Theory, with regard to the imitation of work behavior. He found that the degree of behavior similarity displayed by subordinates and their direct superiors was positively correlated with subordinates' perceptions of their superiors' success and competence. This implies that subordinates are more likely to imitate their superiors’ behavior when they perceive them as successful and competent. However, Weiss (1977) did not find a relationship between behavior similarity and subordinates’ perceptions of superior power, which indicates that perceived superior power, unlike perceived superior competence and success, is not sufficient for imitating behavior.

However, little attention has been paid to imitating behavior in the context of norm violations. As discussed, the possibility exists that norm-violating behavior leads to power affordance (Van Kleef et al., 2012). By combining the social learning theory and the concept of norm-violating behavior, the current study attempts to examine whether norm violations could not only result in power affordance, but also in the imitation of this behavior by others. If norm-violating behavior can lead to power affordance, then copying behavior by others could be a consequence. So, where Weiss (1977) discusses the insufficiency of perceived superior power to trigger imitating behavior, this research attempts to examine if the possibility for power affordance can motivate subordinates to imitate behavior.

(7)

It can be argued that individuals perceive the copying of norm violations of others, as less damaging than initiating the norm violation themselves. In other words, it is examined if norm-violating behavior of others lowers the barrier for people to practice that same behavior. Two reasons come to mind that might explain a lowered barrier for copying behavior. First, people observing others that practice norm-violating behavior, could conclude that they are allowed, or maybe even obligated, to act in the same manner, despite their knowledge about the inappropriateness of this behavior. In this case, the observed person acts as a model for the “appropriate” behavior, regardless of what is actually wrong or right within the organizational context. Second, subordinates might expect and believe that the way other people in the organization behave is the right way. In this case, the observer changed his attitude about the norm-violating behavior and does not perceive it as the “wrong” behavior anymore. The current study examines if this psychological barrier can act as a mediator on the relationship between a leader/co-worker practicing norm-violating behavior and individuals practicing norm-violating behavior themselves.

It is interesting to examine this topic, paying attention to both prosocial and selfish norm violations, and to imitation of leader norm violations, as well as non-leader norm violations. Furthermore, attention is paid to the possibility that the motivation to gain status and power by means of norm-violating behavior can act as a moderator on the effect between norm-violations by others and a lowered barrier to exercise the same behavior.

(8)

2. Theoretical framework

In this section, the existing literature related to the research problem will be addressed in order to define the research topic more precisely. First, the literature concerning norm-violating behavior and related constructs will be discussed. Then the concept of imitation will be elaborated on, including the discussion of some principle theories on this topic. Because leadership and the consequences of leader behavior might be of great influence with regard to imitating behavior of subordinates, the third section is concerned with leadership in general. The next section will introduce the concept of a psychological barrier, in order to explain what occurs in people’s mind that motivates imitating behavior in a certain context. Furthermore, it is discussed if the motivation of individuals to gain status or power can influence the chances of them practicing norm-violating behavior. Finally, a synthesis is provided which emphasizes the need for this research, including the examined hypotheses in the current study.

2.1 Norm-violating behavior

Norms can be defined as “collective expectations for the proper behavior of actors within a given identity” (Shannon, 2000). In other words, a norm is a group-held belief about how members should behave in a given context. These contexts come in various different forms; the business environment, cultural or religious groupings, but also smaller group units such as the grocery store, a team, or an office (Jackson, 1965). Shannon (2000) explains that every norm consists of two components: prescription and parameters. The prescription is the part of the norm explaining what to do (or what not to do). The parameters of a norm indicate under what situations the norm's prescription applies.

(9)

2.1.1 Prosocial and selfish norm violations

Norm violation, or norm-violating behavior, is acting or doing something that is against what the society, culture, or grouping believes is correct. Inappropriate behavior violating the values, attitudes and beliefs of a particular context is believed to be violation of norms. Van Kleef et al. (2012) distinguish between two types of norm violations: prosocial and selfish norm violations. Prosocial norm violations can be described as breaking the rules for the benefit of others, whereas selfish norm violations represent inappropriate behavior with benefits for the person performing the norm violation.

2.1.2 Norm-violating behavior vs. ethical behavior

Norm-violating behavior seems to be closely related to the concept of ethical business conduct and ethical behavior. The number of corporate ethical scandals that have harmed millions of employees, consumers, and investors has shown to be at a peak at the beginning of this 21st century (Camps & Majocchi, 2010). Vitell, Keith, and Mathur (2011) describe moral reasoning as “the cognitive activity of processing information about issues to make moral judgments” and argue that it is presumed to be one of the strongest predictors of ethical behavior. It can be argued that people less capable of moral reasoning are more likely to engage in norm-violating behavior.

Another concept important to discuss with regard to norm violations is moral justification. A person who engages in moral justification practices behavior that can be seen as “wrong” or “inappropriate” but makes that negative behavior personally and socially acceptable by somehow portraying it as serving a valued norm. In short, moral justification is defined as “the act of validating or rationalizing unethical behavior” (Vitell, Keith, & Mathur, 2011). As discussed, the current study aims at providing insight on imitating norm-violating behavior. The theory of moral justification could be one explanation for the fact that people

(10)

copy certain types of behavior that can be described as unethical or norm-violating. To further elaborate on this topic, the next section will discuss the process of imitation.

2.2 Imitation

History shows us an interesting aspect of people performing norm-violating behavior: often, others deliberately participate in this kind of behavior. Think of the countless corporations where employers and employees participate in business scandals leading to millions of deceived consumers or other employees. For example, the Dutch Rabobank recently had to pay a 774 million euro fine for engaging in the so-called Liboraffaire. Managers of the international department in London manipulated the Libor-rent in order to increase profits. A reconstruction of the events shows that when the Dutch department became aware of this business scandal, it deliberately pursued the same practices for more than two years. Why does a person decide to engage in behavior that is obviously not appropriate and sometimes clearly wrong? Or, in other words, what is it that makes people imitate the (norm-violating) behavior of others?

2.2.1 Social Learning Theory

To stay within a business context, Weiss’ (1977) study on subordinate imitation of supervisor behavior describes the concept of “Organizational Socialization”. This is a process of ‘learning the ropes’, getting to know the nature of norms within the organization, as well as the organizational perspectives and expectations. Furthermore, according to Bandura’s (1977) Social Learning Theory (SLT), leaders can act as models for their subordinates. Leaders have certain attributes such as status, power, and perceived capability, which lead subordinates to believe that the model’s behavior has been rewarded in the past or is appropriate in a particular situation. These beliefs, in turn, lead subordinates to expect that when they engage

(11)

in or imitate the same behavior as their leaders, they might get rewarded as well or are at least engaging in appropriate behavior.

Social learning theory emphasizes antecedent learning as well as learning by consequences (Manz & Sims, 1981). Antecedent learning refers to behavioral change as a consequence of a stimulus that occurs before the behavior. For example, if an employee has a specific goal, such as a higher salary or a different position within the organization, this could act as the antecedent for a certain type of behavior. Learning by consequences works the opposite way and is best defined by operant theory. This theory explains that individuals have the tendency to increase behavior that results in positive consequences, and will try to decrease the frequency of behavior that has negative consequences. Building on these two types of learning, two forms of imitating norm-violating behavior could be identified as well. Employees within an organization could imitate a certain type of behavior because they receive reinforcement for practicing that behavior. When considering the Rabobank-example, employees might have received monetary reinforcements (positive consequence) that motivated them to engage in the Liboraffaire. Using another person in the organization as a role model, for example a supervisor, is a typical example of antecedent learning. The supervisor can be considered as the stimulus prior to behavioral change.

Despite the fact that few researchers paid attention to imitating behavior in the context of norm-violating behavior, there is a history of literature on both constructs separately. Flanders (1968) focuses on reviewing experiments of imitating behavior, and argues that the imitation of people who are older, more skillful, or who possess high social status, is more likely than the imitation of people lacking these attributions. He emphasizes that too little is known about either the relative effects of two or more independent variables on imitation or the strength of association between any independent variables and imitation. Therefore, this research will focus on imitation in the context of variables that have not been examined yet.

(12)

To extend the current literature on imitating behavior, this study focuses on the imitation of norm-violating behavior in an organizational context. It can be argued that the social learning theory would predict the same imitating effects for norm-violating behavior as it does for behavior in general. If superiors initiate or engage in norm-violating behavior, this can lead their subordinates to believe that it must be the correct way to behave in a particular situation, even if they know that in essence this behavior is wrong. The leader acts as a role model, serving as an example for people within the organization. The situation of superiors engaging in norm-violating behavior might also lead others to think that this kind of behavior is appropriate. In other words, it might positively change the attitude of others towards norm-violating behavior.

2.2.2 Vicarious Learning

To further elaborate on the concept of learning, Manz and Sims (1981) focus on one particular form of learning: vicarious learning. Modeling is a type of vicarious learning that plays a prominent role in Bandura’s social learning theory (1977). The basic elements of this type of learning are well captured by one of Bandura’s (1977) statements: “By observing a model of the desired behavior, an individual forms an idea of how response components must be combined and sequenced to produce the new behavior. In other words, people guide their actions by prior notions rather than by relying on outcomes to tell them what they must do” (Bandura, 1977a, p. 35). Bear, Peterson, and Sherman (1971) argue that this type of learning could also be described as “imitative behavior”.

Manz and Sims (1981) explain three different types of modeling. The first is learning a new behavior by observing a model, which explains itself. The second type of modeling is concerned with the consequences for the model. It holds that when a model gets reinforced for a certain type of behavior, the observer is likely to be disinhibited and is more eager to exercise this behavior him- or herself. If the model’s behavior is punished, the observer is

(13)

likely to be inhibited and the chances of imitating behavior are smaller. The behavioral facilitation effect is the third type of modeling. The behavioral facilitation effect occurs when a model acts as a cue for others to imitate a previously learned behavior. Manz and Sims (1981) display the example of the person with the highest organizational position sitting down first in a meeting, after which the rest of the room will imitate this previously learned behavior.

All three types of modeling have the potential to explain why individuals within an organization imitate norm-violating behavior. The first option is that they have never engaged in norm-violating behavior before, but they learned this as a new behavior by observing a model. An individual could perceive a behavior learned by a model as the correct and appropriate behavior, regardless of what is in fact “right” in a specific situation. Second, employees could decide on their own behavior by examining the consequences of particular types of behavior for the observed model. For example, when the observer notices that the model works a lot of overtime and knows the model gets a monetary reinforcement for these extra hours, the chance of the observer imitating this behavior increases. The same mechanism could take place when individuals observe models getting reinforced for violating behavior. The third possibility is the behavioral facilitation effect causing norm-violating behavior. This means that the norm-norm-violating behavior is previously learned by the individual, but that a certain model first needs to practice this behavior in order for the individual to imitate it. It seems reasonable that in certain situations individuals might not initiate norm-violations themselves, but are willing to break the rules when they observe a model practicing this type of behavior.

(14)

2.2.3 Organizational Learning Capability

In line with the current study, Camps and Majocchi (2010) examine norm-violating and unethical behavior in a business context. By discussing the concept of organizational learning capability (OLC) they attempt to find causes of unethical behavior. Much research has noted the impact of multiple aspects of organizational contexts on individuals’ ethical behavior. Camps and Majocchi (2010) explain the importance of organizational learning capability with regard to this research topic. They argue that employees’ perceptions of the organization’s ethical climate relate to individual ethical attitudes and behavior. Furthermore, it is argued that individuals also relate to certain peers within the organization when it comes to ethical behavior. This implies that leadership could play an important role in the establishment of desired (and undesired) behavior and attitudes and that leaders can act as a role model for their employees. The fact that individual behavior relates to that of certain peers within the organization could be the consequence of individuals (deliberately or unconsciously) imitating those significant others.

2.3 Leadership

The concept of leadership is an extensively examined topic. Most definitions of leadership reflect the assumption that it involves a social influence process whereby intentional influence is exerted by one person over other people to structure the activities and relationships in a group or organization (Yukl, 1994). Despite the countless definitions, this broad description of leadership seems to cover most, if not all, commonalities that can be found between the definitions.

Yukl (1994) explains a difference in literature on leadership that is important to note here. Some theorists argue that the definition of leadership should be limited with regard to the outcomes and consequences of that leadership behavior. They argue that leadership can

(15)

only be considered as “real” leadership when the leader’s influence results in enthusiastic commitment by followers. Proponents of this view argue that a person who uses authority and control over rewards and punishments cannot be considered as someone practicing leadership. However, opponents argue the short-sightedness of this perspective, stating that it excludes influence processes that are important for understanding why a manager or leader is effective or ineffective in a given situation. With regard to the current research, it is important to take into account both effective and ineffective leadership, because it is yet to discover what causes people to imitate norm-violating behavior and how individuals are influenced to act in a certain way.

Another important point made by Yukl (1994) is the incorporation of all influence attempts made by a leader. Some research argues that only attempts related to task objectives or group maintenance should be included in leadership. However, “it is evenly important to include influence attempts that are extraneous or even detrimental to the group, such as a leader’s attempts to gain personal benefits” (Yukl, 1994). This statement is in line with the current study, in that it is examined here whether the influence of leader could result in certain forms of norm-violating behavior.

Terms such as influence and power have been used in various contexts by different writers. Both concepts will be discussed in the following sections.

2.3.1 Influence

In order to define the concept of influence as detailed as possible, this section will elaborate on Kelman’s (1958) conceptualization of three different social influence processes. First, he describes a process where a target person (e.g. an employee) carries out a requested action in order to obtain a reward or to avoid a punishment that is controlled by the agent (e.g. a leader). This type of influence process is called instrumental compliance. To relate this to the

(16)

current study, an employee could be motivated to engage in norm-violating behavior, having the knowledge that this could result in a (monetary) reinforcement.

The second influence process is referred to as internalization. Here, the subordinate “becomes committed to support and implement proposals espoused by the leader because they are perceived to be intrinsically desirable and correct in relation to the target’s values, beliefs, and self image” (Yukl, 1994). In other words, the subordinate has feelings of loyalty towards the leaders’ ideas, not towards the leader communicating the ideas. However, internalization is more likely to occur when the leader communicating the ideas is skilled in making rational and emotional appeals. This implies that certain types of leadership could result in a change of subordinates’ attitudes towards specific (norm-violating) plans, procedures, or even entire strategies.

The last social influence process is identification, and might be of great importance with regard to the current study. It reflects employees wanting to imitate the behavior of their leader in order to fulfil their needs for acceptance and esteem. Important to note here is that this process not necessarily has to be a conscious one. By this is meant that identification may occur in response to an influence attempt by the agent, but also in the absence of such an attempt. To add some insight to the current research, this means an individual could (deliberately or unconsciously) engage in norm-violating behavior solely to identify him- or herself with a norm-violating leader.

2.3.2 Power

Power is generally referred to as an individual’s capacity to influence a target person (Mintzberg, 1983; Yukl 1994). Where some research merely focuses on power influencing a person’s behavior, other studies emphasize the importance to examine the influence of power on a person’s attitudes as well. This difference in definition is important for the current study, because it is examined if leaders’ influence could result in a change of subordinates’ beliefs

(17)

about the “appropriateness” of norm-violating behavior. In other words, it will be examined if leaders could change the attitudes of subordinates towards norm-violations.

French and Raven (1959) developed a taxonomy to classify different types of power. Some of them closely correspond to the previously discussed influence processes. The first two types of power are reward and coercive power, which relate to Kelman’s (1958) instrumental compliance. An individual complies in order to obtain rewards or avoid punishments (coercive power) he or she believes are controlled by the agent.

Legitimate power occurs when an individual has the opinion that the leader has the right to make a certain request and that it is the individual’s obligation to comply. With regard to this study, the leader could request a certain type of norm-violating behavior, resulting in the subordinate to comply because he or she feels obligated to do so.

The fourth type of power is expert power and is primarily associated with Kelman’s (1958) internalization process. This power type holds that an individual complies because he or she believes that the leader has specific knowledge about the best way of doing something. This means that the individual sets aside personal beliefs and attitudes, because it is believed that the leader knows best. As discussed earlier, people observing others that practice norm-violating behavior, could conclude that they are allowed, or maybe even obligated, to act in the same manner, despite their knowledge about the inappropriateness of this behavior.

The fifth and last power type is referred to as referent power. This power type strongly corresponds to Kelman’s (1958) identification influence process. In this case, the subordinate complies because he or she admires or identifies with the leader and wants to gain the agent’s approval. When the leader engages in norm-violating behavior, this could result in the subordinate’s imitation of this behavior because of the need for identification with the leader.

(18)

2.4 Status motivation

The previous section discussed various possibilities of how leaders can influence and motivate subordinates to practice certain types of behavior. Another interesting construct to examine concerning the motivation of people to imitate norm-violating behavior, is the motivation to gain status. In other words: is it possible that employees practice norm-violating behavior in order to gain status and power themselves?

This status motivation could be complemented by the earlier discussed concept of moral justification. A person who engages in moral justification practices behavior that can be seen as “wrong” or “inappropriate” but makes that negative behavior personally and socially acceptable by somehow portraying it as serving a valued norm (Vitell, Keith, & Mathur, 2011). When the individual considers his or her motivation to gain status a valued norm, a process of moral justification could occur.

Magee (2009) examines the hypothesis that social targets who display a greater action orientation are perceived as having more power. The study shows that action orientation has dramatic effects on inferences about social targets’ power. Extensive deliberation and the failure to act convey little power, and focusing on goal implementation and bold action signal that one is doing what one wants and thus possesses power. This could lead other people in the organization to imitate this kind of behavior, because they are motivated to gain status and power as well. The aim of this research is to examine whether these action signals encourage individuals to imitate this kind of behavior, for example to gain power within the organizational (or social) context.

As previously discussed, Weiss (1977) concludes that subordinates are more likely to imitate their superiors’ behavior when they perceive them as successful and competent. However, his research did not find a relationship between behavior similarity and subordinates’ perceptions of superior power. This implies that perceived superior power is not

(19)

sufficient for imitating behavior. However, it does make it interesting to examine whether people do imitate behavior if they have the motivation to gain status and power themselves.

This line of thought is complemented by other research (Bandura, 1969; Bandura, 1971; Weiss, 1977; Keltner et al., 2008). Van Kleef et al. (2012) examined the possibility that power is afforded to individuals who are willing to break the rules to benefit their group. They conclude that prosocial norm violations can lead to power affordance. If norm-violating behavior can lead to power affordance, then copying behavior by others could be a consequence. Therefore, it is expected that the motivation to gain status moderates the effect between norm-violations by others and a lowered psychological barrier to exercise the same behavior. In other words, this research examines whether people that have the motivation to gain status and power are more likely to imitate the norm-violating behavior of others within the organization.

2.5 Psychological barrier

To elaborate on what takes place in people’s mind that motivates imitating behavior in a certain context, the concept of a psychological barrier is introduced. Chugh, Bazerman, and Banaji (2005) describe the psychological barrier in terms of bounded ethicality in decision-making. Bounded ethicality places a critical constraint on the quality of decision-making, especially in a situation where issues of ethics are in order. It emphasizes that individuals view themselves as moral and competent, and this view obstructs their ability to see and recognize conflicts of interest when they occur.

This study has a slightly different perspective on this psychological barrier. The concept of a barrier is used here to explain that people have certain beliefs about what is right and what is wrong. In the context of this research, a lowered psychological barrier could have two explanations. First, people observing others that practice norm-violating behavior, could

(20)

come to the conclusion that they are allowed to act in the same manner, even if they know that this behavior is not appropriate. In this case the observed person acts as a model for the “appropriate” behavior, regardless of what is actually wrong or right within the organizational context. Second, when subordinates observe norm-violating behavior of others, their psychological barrier might be lowered because they expect and believe that the way other people in the organization behave is the right way. In this case, the observer changed his or her attitude about the norm-violating behavior and does not perceive it as the “wrong” behavior anymore. The lower the psychological barrier, the easier it is for the individual to exercise norm-violating behavior. It is expected that the psychological barrier can be increased or decreased by norm violations of others, and therefore it is examined if this barrier can act as a mediator on the link between norm violations by others and performing the norm violation. In other words, it is expected that norm violations of a leader lower the psychological barrier of subordinates to exercise the same behavior.

To examine if there is a difference for subordinates to imitate the behavior of other employees or the behavior of their supervisor, this study will also include non-leaders in the research. For the same reasons as for the effects of norm-violating behavior of leaders on subordinates, it is expected that norm violations of a non-leader lower the psychological barrier of others to exercise the same behavior. However, because leaders can act as a role model for subordinates and because subordinates might consider leader’s behavior as the correct behavior, it is expected that – in comparison with imitating the behavior of non-leaders – employees are more prone to copy behavior of their non-leaders.

As discussed, it can be expected that the lowered psychological barrier to imitate norm-violating behavior will increase the chance of exercising this kind of behavior. In other words, individuals experience the norm-violating behavior of others, create a more positive idea about this behavior (or a less negative idea), which eventually results in a greater chance

(21)

of exercising norm-violating behavior themselves. Therefore, it is expected that a lowered psychological barrier to imitate norm-violating behavior increases norm-violating behavior.

This study also aims to examine the possible differences between prosocial and selfish norm violations. Van Kleef et al. (2012) argue that their results provide evidence that prosocial norm violations fuel perceptions of legitimate power, whereas selfish norm violations do not. However, they raise the question whether individuals who enjoy elevated power due to prosocial norm violations will be motivated to ensure that future norm violations also benefit others, or whether they start breaking the rules in an indiscriminate fashion, with no regard for the social consequences for others. It is expected that subordinates find it easier to copy or imitate norm-violating behavior when they know this will benefit others, aiming for the wellbeing of the group, than when the norm-violating behavior has a selfish purpose.

2.6 Synthesis

This section provides a summarized overview of the discussed theory leading to the hypotheses that will be examined in the current study. The previous sections explained various ways of how a leader can influence subordinates to exercise a certain type of behavior. Subordinates may want to identify themselves with their leader, or they may believe that their leader has special knowledge about the best way of doing something. This could lower their psychological barrier to also engage in norm-violating behavior. Furthermore, the motivation to gain status and power could also lead subordinates to imitate specific behaviors.

Two explanations for a lowered psychological barrier were identified. First, people observing others that practice norm-violating behavior, could conclude that they are allowed, or even obligated, to act in the same manner. Second, when subordinates observe

(22)

norm-violating behavior of others, their psychological barrier might be lowered because they expect and believe that the way other people in the organization behave is the right way.

These and other discussed theories lead to the following expectations about imitating- and norm-violating behavior:

Hypothesis 1: Norm violations of a leader are more likely to lower the psychological barrier of subordinates to exercise the same behavior compared to norm violations of a non-leader.

Hypothesis 2: A lowered psychological barrier to imitate norm-violating behavior increases norm-violating behavior through imitation.

Hypothesis 3: Norm violations of a leader are more likely to lead to imitation of that behavior by subordinates compared to norm violations of a non-leader.

Hypothesis 4: The motivation to gain status moderates the effect between norm violations by others and a lowered psychological barrier to exercise the same behavior (this means that it is expected that a higher motivation to gain status leads to a lowered psychological barrier).

(23)

3. Method

The previous section discussed the existing literature on the concepts that will be examined in this research. This led to the formation of four hypotheses. The research design and method through which these hypotheses will be tested are discussed in this section. First, the experimental design of the current study will be discussed. This is followed by a clarification of the method for data collection. Next, the participants that took part in this study will be discussed, which is followed by a clear explanation of the measures used in this study. The specific analyses that will be used in order to test the hypotheses will be provided in the next chapter.

3.1 Research design

For this research an experiment is used in order to collect the data needed to test the hypotheses that are discussed in the previous section. An experiment is a test of cause-effect relationships by collecting evidence to demonstrate the effect of one variable on another (Breakwell & Hammond, 1995; Saunders et al., 2009). In its simplest form, two groups of subjects are treated in exactly the same way except for one. This is the experimental or differential treatment and any observed difference between the groups is then attributed to the different treatment. Experiments can be used to establish causal relationships that are generalizable (Barabas & Jerit, 2010; Breakwell & Hammond, 1995; Gaines, Kuklinski, & Quirk, 2007). By randomly assigning respondents to treatment and control conditions it can be examined whether one factor causes another; a causal relationship.

The experimental design is considered best for this research, because it provides the opportunity of establishing causal relationships. Because little attention has been paid to imitating behavior in the context of norm violations, an experiment is feasible because it creates the possibility of examining causal relationships concerning this research topic.

(24)

For the core experiment, the respondents were randomly allocated to two groups of equal size and each group was presented with one of two versions of the survey question about norm-violating behavior within a business context. One half was asked to imagine working in an organization and to fill out the survey focusing on their (imaginary) supervisor/employer. The other half was also asked to imagine working in an organization but instead they were asked to fill out the survey focusing on other coworkers – other employees just like them. In other words, the differential treatment in the current study is the leadership condition versus the non-leadership condition. Allocating subjects to different conditions within an experiment means that a between-subjects design is used for this research. Furthermore, a within-subject design is used for examining a potential difference in relationships with regard to prosocial and selfish norm violations.

By means of an experimental design, this research aims to examine if employees are more likely to imitate norm violations when they observe their leaders practicing this behavior in comparison with a situation where they observe their coworkers practicing norm-violating behavior.

3.2 Data collection

For the execution of the experiment, a survey was designed including all measures that need to be examined in order to test the hypotheses. A survey makes it easier to gather data from a large amount of people and to reach a lot of respondents in a quick and affordable way (Saunders et al., 2009). The method used to reach respondents is the internet-mediated method. The website www.qualtrics.com was used to develop the survey, because it provides many options for different types of questions and it is easy to use for respondents. This method saved a lot of time, since the data did not have to be entered manually (Saunders et al., 2009). Furthermore, the internet-mediated method saved money as well, because no

(25)

printing costs were involved in the process (Wright, 2005). Also, this method makes it easy to reach people in different geographical locations and it can reach a lot of respondents in a relatively short time. To ask for respondent’s participation, emails were sent with information about the research and also social network sites were used to reach and inform potential participants. The internet-mediated method is especially useful for experiments because Qualtrics provides the opportunity to randomly assign respondents to different conditions (control- and treatment group) and to continuously change the order of questions to establish validity (randomisation).

However, collecting data through the Internet also had its disadvantages. This method can lead to a systematic bias because people engage in self-selection. In other words, they choose whether they want to participate or ignore the invitation (Wright, 2005).

The survey is written in Dutch language, because potential participants were assumed to be Dutch. The layout was made attractive to make the questionnaire easy to process for the respondents. This is important to increase the response rate (Saunders et al., 2009). Furthermore, a process element was included, which showed what part of the survey the participant had already completed. Wright (2005) describes this as an important element for participants because it motivates them to complete the survey. The survey includes an introduction letter, which explains the purpose of the research and why the respondent’s opinion is important. The survey is closed with a short text by which the respondents are thanked for their contribution and provided an opportunity to contact the author with any questions or comments on the survey. Also, participants were given the opportunity to fill in their email address if they are interested in the outcome of the study.

(26)

3.3 Participants

The sample size was strived to be as large as possible, because the larger the sample size, the more generalizable the results are (Saunders et al., 2009). Because this research examines one between-subjects factor, being the leader-condition versus the non-leader-condition, at least 40 participants were needed for the validation of the results. Despite of money and time constraints often forming a limitation on sample size, 117 respondents participated in the experiment. However, after removing data from participants that filled out less than 85% of the questions, 63 useful surveys remained for the execution of the analyses. Within this group, 31 respondents were assigned to the non-leadership condition, and 32 respondents were assigned to the leadership condition. The participants were all contacted through email or social network sites.

3.4 Measures

This section discusses the variables used for this research. Responses for all items were given on a seven-point Likert scale (1 = ‘completely disagree’ to 7 = ‘completely agree’).

3.4.1 Differential treatment

As explained earlier, half of the respondents were assigned to the leader condition and the other half were assigned to the non-leader condition. The introduction of the survey clearly explained to half of the respondents that they had to imagine working in an organization and to fill out the survey focusing on their (imaginary) supervisor/employer. The other half was also asked to imagine working in an organization but instead they were asked to fill out the survey focusing on (imaginary) coworkers.

(27)

3.4.2 Norm-violating scenarios

The respondents were presented four different norm-violating scenarios. Two of the scenarios were a representation of prosocial norm-violating behavior, and the other two describe examples of selfish norm-violating behavior. The selfish norm-violating behaviors are: “A leader/supervisor (vs. a co-worker) uses the company car for private matters” and “A leader/supervisor (vs. a co-worker) handles private matters during work hours”. The prosocial norm-violating behaviors are: “A leader/supervisor (vs. a co-worker) acquires decoration (e.g. paintings, flowers) for the furnishing of the office, while personal decoration of the office is not allowed” and “A leader/supervisor (vs. a co-worker) sometimes works overtime at the office after official closing time of the organization, which is not allowed”.

3.4.3 Psychological barrier

This measure includes five items and it is examined if it can act as a mediator on the relationship between a leader/co-worker practicing norm-violating behavior and individuals practicing norm-violating behavior themselves. The five items used for the current study are: “When my leader/supervisor (vs. a co-worker) would practice the behavior in the above-mentioned scenarios, (1) this makes it easier for me to practice that same behavior, (2) this lowers the barrier for me to practice that same behavior, (3) this encourages me to practice that same behavior, (4) I tend to tolerate this behavior, (5) I expect this behavior to be accepted”.

3.4.4 Imitation of norm violations – Dependent variable

The dependent variable tested in the current study is the imitation of (prosocial and selfish) norm violations. The survey includes two items to test this variable. The two items of imitation are: “When my leader/supervisor (vs. a co-worker) would practice the behavior in

(28)

the above-mentioned scenarios, there is a great chance I would (deliberately or unconsciously) imitate this behavior”, and “When my leader/supervisor (vs. a co-worker) would practice the behavior in the above-mentioned scenarios, there is a great chance I would practice this behavior myself in the future”.

3.4.5 Status motivation

The last variable is status motivation. This measure tests the extent to which respondents have the motivation to gain status and power. The status motivation variable includes ten items and is based on the items used by Bennett (1988). The current study examines if status motivation can act as a moderator on the effect between norm violations by others and a lowered barrier to exercise the same behavior. An example item of this variable is: “I think I would enjoy having authority over others”. This measure includes five reverse coded items.

(29)

4. Analyses

The previous section discussed the design and measures of the current study. This section will provide the analysis strategy that addresses the hypotheses, which is followed by a presentation of the findings in the next chapter.

4.1 Analysis strategy

The aim of this study is to examine the relationships concerning norm-violating behavior of others and imitating behavior. As discussed earlier, the current study examines if norm violations of a leader are more likely to lower the psychological barrier of subordinates to exercise the same behavior compared to norm violations of a non-leader. Also, it is examined if an individual’s psychological barrier can act as a mediator on the relationship between a leader/co-worker practicing violating behavior and individuals practicing norm-violating behavior themselves. Finally, it is examined if the motivation to gain status and power by means of norm-violating behavior can act as a moderator on the effect between norm violations by others and a lowered barrier to exercise the same behavior. The mediation and moderation effects will be examined separately.

4.1.1 Mediation analysis

The data will be analyzed through a linear regression analysis in the program SPSS. Four models will be presented:

1. Main effects à Independent influence of non-leaders and leaders practicing norm-violating behavior on imitation of the norm-norm-violating behavior.

2. Main effects à Independent influence of non-leaders and leaders practicing norm-violating behavior on the psychological barrier to imitate the norm-norm-violating behavior.

(30)

3. Main effects à Independent influence of the psychological barrier to imitate norm-violating behavior on the actual imitation of the norm-norm-violating behavior.

4. Mediating effect à Mediating effect of the psychological barrier to imitate the norm-violating behavior on the relationship between non-leaders and leaders norm-norm-violating behavior and imitation of the norm-violating behavior.

4.1.2 Moderation analysis

Again, the data will be analyzed through a linear regression analysis in the program SPSS. Three models will be presented:

1. Main effects à Independent influence of non-leaders and leaders practicing norm-violating behavior on the psychological barrier to imitate the norm-norm-violating behavior. 2. Main effects à Independent influence of the motivation to gain status/power on the

psychological barrier to imitate the norm-violating behavior.

3. Moderation effect à Interacting effect of non-leaders and leaders practicing norm-violating behavior and the motivation to gain status/power on the psychological barrier to imitate the norm-violating behavior.

(31)

5. Results

The results of the research will be provided in this section. First, some descriptive statistics will be discussed. After that, the reliability of the scales is shown by calculating Cronbach’s Alpha. Finally, the hypotheses will be tested using linear regressions.

5.1 Descriptive statistics

In order to get an idea and overview of the sample, this section presents the characteristics of the sample used and provides a description of the data collected.

5.1.1 Sample characteristics

As previously discussed, the survey ought to be completed by at least twenty respondents for the leader condition, and by another twenty for the non-leader condition. In total, 117 respondents filled out the questionnaire. However, after removing data from participants that filled out less than 85% of the questions, 63 useful surveys remained for the execution of the analyses. Within this group, 31 respondents were assigned to the non-leadership condition, and 32 respondents were assigned to the leadership condition. It can be concluded that this amount of respondents is sufficient for the research.

The survey has been completed by 21 males (33.3%) and 42 females (66.6%). The minimum age of the respondents is 19, the maximum age is 37, and the average age of the respondents is 22.8. There are five educational level options in the survey which resulted in the following information about the employees: 4.8% = College; 11.1% = MBO; 27.0% = HBO; 57.1% = University; 0% = Other. Furthermore, 75% of the respondents has a part-time job, so consequently 25% works fulltime.

(32)

Table 1: Descriptives and Correlations between the variables (Cronbach's Alphas on diagonal)

M SD 1 2 3

1 Psychological barrier 4.63 0.869 (0.879)

2 Imitation of norm-violating behavior 4.53 1.134 0.815** (0.864)

3 Status motivation 4.62 1.085 0.129 0.089 (0.906)

Note. N = 63. * p < 0,05. ** p < 0,01.

5.2 Reliability

To be able to analyse the data, variables were created that are based on the items derived from the questionnaire. In order to make sure that these new variables are reliable, Cronbach’s Alpha test was used. This is the most common measure of scale reliability (Field, 2009, p. 674). The new items are accepted as reliable when Cronbach’s Alpha is above 0.7 (Field, 2009, p. 675). Cronbach’s Alpha measures if the items from the questionnaire that are used to create a variable, indeed do measure the same thing.

By calculating the Cronbach’s Alphas, the reliability of the three variables measured in this paper was found. Good reliability for the psychological barrier measure (α = 0.88) and the imitation measure (α = 0.86) was found (α ≥ 0.8). The status motivation measure is even highly reliable, with a reliability of 0.91 (α ≥ 0.9). As can be seen in table 1, all three measures have a Cronbach’s Alpha higher than 0.7, which means that they are reliable and can be used in this research.

5.3 Correlations

Next to the Cronbach’s Alphas, also the correlations between the variables are calculated. A correlation measures the relationship between variables (Field, 2009, p. 167). The current study used Pearson’s R to calculate the correlations, and the results are presented in table 1.

(33)

As can be seen, a highly positive (r = 0.815) and highly significant (p < 0.01) correlation was found between the psychological barrier and the imitation of norm-violating behavior. No significant relationships were found between the motivation to gain status and one of the other two variables.

5.4 Differential treatment

To test hypothesis 1, table 2 presents the main effects of norm violations of a leader on the psychological barrier and the main effects of norm violations of a non-leader on the psychological barrier.

Hypothesis 1 predicted that norm violations of a leader are more likely to lower the psychological barrier of subordinates to practice that same behavior, compared to norm violations of a non-leader. The results show a positive (β = 0.300) and significant (p = 0.017) effect and therefore support hypothesis 1. It can be concluded that norm violations of a leader, compared to those of a non-leader, are more likely to lower the psychological barrier of subordinates. This outcome is in line with predictions. Furthermore, Model 1 has an explained variance (R2) of 9 percent.

Table 2: Regression Results of Main Effects of Leader and Non-leader Norm Violations on the Psychological Barrier (Model 1).

Psychological Barrier DV Model 1

Coefficient SE Beta t

Non-leader vs. leader norm violations 0.518* 0.211 0.300 2.458

(34)

5.5 Mediation analysis

To test if the psychological barrier can act as a mediator, first the main effects between the variables are tested using a linear regression. To test hypothesis 1, the previous section already discussed the main effects between the norm violations of leaders and non-leaders and the psychological barrier. Table 3 presents the main effects between the psychological barrier and the imitation of norm-violating behavior.

Hypothesis 2 states the expectation that a lowered psychological barrier to imitate norm-violation behavior increases norm-violating behavior through imitation. Table 3 shows that the results support this hypothesis. A strong positive (β = 0.815) and highly significant (p < 0.001) relationship between the psychological barrier and imitation was found. These results are in line with predictions. However, this exceptional strong relationship (β = 0.815) could also indicate that the variables were perceived as being similar. This will be further discussed in the next chapter. Finally, Model 2 has a high explained variance (R2 = 0.664).

The third hypothesis predicted that norm violations of a leader are more likely to lead to imitation of that behavior by subordinates compared to norm violations of a non-leader. The main effects for this relationship are presented in Table 4.

Table 3: Regression Results of Main Effect of the Psychological Barrier on Imitation (Model 2)

Imitation DV Model 2

Coefficient SE Beta

Psychological Barrier 1.063*** 0.097 0.815

(35)

The results show that there is a positive effect of leaders’ and non-leaders’ norm violations on the imitation of this behavior. However, this relationship is not significant (p = 0.235). It can be concluded that hypothesis 3 is not supported.

In order for the psychological barrier to act as a mediator on the relationship between a leader/co-worker practicing violating behavior and individuals imitating that norm-violating behavior, all main effects between the variables need to be positive and significant. Because the relationship between leader and non-leader norm violations and imitation is not significant, it can be concluded that the psychological barrier does not act as a mediator on the relationship between a leader/co-worker practicing norm-violating behavior and individuals imitating that norm-violating behavior.

5.6 Moderation analysis

To test if the motivation to gain status can act as a moderator on the effect between norm-violations by others and a lowered barrier to exercise the same behavior, the main effects of non-leaders and leaders practicing norm-violating behavior on the psychological barrier to imitate the norm-violating behavior need to be examined. Furthermore, the main effects of the motivation to gain status on the psychological barrier to imitate the norm-violating behavior need to be tested. Finally, an interaction-variable based on norm-violations of others and the

Table 4: Regression Results of Main Effects of Leader and Non-leader Norm Violations on Imitation (Model 3).

Coefficient SE Beta

Non-leader vs. leader norm violations 0.067 0.056 0.152

(36)

motivation to gain status is created and used to test a potentially moderating role of status motivation. The results are presented in Table 5.

The results in Table 5 show that there is a positive (β = 0.133) but insignificant (p = 0.522) effect between the motivation to gain status and the psychological barrier. Furthermore, a positive (β = 0.129) but insignificant (p = 274) effect between norm violations of others and the psychological barrier was found. Finally, the results show a positive (β = 0.974) but insignificant (p = 0.095) effect of the interaction-variable on the psychological barrier. Unfortunately, it must be concluded that the motivation to gain status does not act as a moderator on the effect between norm-violations by others and a lowered barrier to exercise the same behavior.

Table 5: Regression Results of Main Effects and Moderation Effect on the Psychological Barrier (Model 4).

Coefficient SE Beta

Status motivation 0.106 0.165 0.133

Non-leader/Leader * Status motivation 0.345 0.203 0.974

(37)

5.7 Prosocial versus selfish norm violations

Interesting to examine is whether prosocial versus selfish norm violations have different effects on the psychological barrier. Furthermore, a lowered psychological barrier for prosocial versus selfish norm violations could have a different effect on imitation for these different norm violations types. Therefore, this section will examine these effects. Table 6 presents the results for the effect of norm violations of others on the psychological barrier with regard to selfish norm violations. Table 7 presents the results for prosocial norm violations. It is expected that prosocial norm violations are more likely to lower the psychological barrier than selfish norm violations, because this kind of behavior has a social benefit for others. Individuals could perceive it as less damaging to imitate this kind of norm violations instead of selfish ones.

The results show that there is a positive (β = 0.240) but insignificant (p = 0.058) effect of others practicing selfish norm-violations on the psychological barrier.

Table 6: Regression Results of Main Effects of Leader and Non-leader Norm Violations on the Psychological Barrier concerning Selfish Norm Violations (Model 5).

Coefficient SE Beta

Non-leader and leader norm violations 0.452 0.234 0.240

(38)

The results in Table 7 show a positive (β = 0.244) significant (p = 0.020) effect of the prosocial norm-violating behavior of others on the psychological barrier. Therefore, it can be concluded that the main effect of norm-violating behavior of others on the psychological barrier is mainly explained by prosocial norm violations. This outcome is in line with predictions.

As discussed, a lowered psychological barrier for prosocial versus selfish norm violations could have a different effect on imitation for these different norm violations types. Table 8 and Table 9 show the effects of the psychological barrier for both types on the imitation of these norm violations.

Table 7: Regression Results of Main Effects of Leader and Non-leader Norm Violations on the Psychological Barrier concerning Prosocial Norm Violations (Model 6).

Coefficient SE Beta

Non-leader and leader norm violations 0.584* 0.244 0.293

Note. N = 63. R² = 0.086. * p < 0.05. ** p < 0.01. *** p < 0.001

Table 8: Regression Results of Main Effects of the Psychological Barrier on Imitation concerning Selfish Norm Violations (Model 7).

Coefficient SE Beta

(39)

The results show that the psychological barrier for both selfish and prosocial norm violations has a positive (respectively β = 0.783 and β = 0.795) and highly significant (p < 0.001) effect on the imitation of those norm violations. As predicted, prosocial norm violations are more likely to be imitated but the results show that there is only a small difference between the two types.

Finally, the moderating role of the motivation to gain status is examined with regard to prosocial versus selfish norm violations. No moderating effect of status motivation on the relationship between norm violations and the psychological barrier was found. To further examine these variables, the moderating effect of status motivation on the two different norm violation types is tested. The results are presented in Table 10 and 11.

Table 9: Regression Results of Main Effects of the Psychological Barrier on Imitation concerning Prosocial Norm Violations (Model 8).

Coefficient SE Beta

Note. N = 63. R² = 0.632. * p < 0.05. ** p < 0.01. *** p < 0.001

Table 10: Regression Results of Main Effects and Moderation Effect on the Psychological Barrier concerning Selfish Norm Violations (Model 9).

Psychological Barrier Selfish DV Model 9

Coefficient SE Beta

(40)

The results show a positive (β = 0.418) but insignificant (p = 0.070) effect of the interaction-variable on the psychological barrier with regard to selfish norm violations. Furthermore, the results show a positive (β = 0.666) but insignificant (p = 0.256) effect of the interaction-variable on the psychological barrier to practice prosocial norm violations. It must be concluded that the motivation to gain status does not act as a moderator on the effect between norm violations by others and a lowered barrier to exercise either prosocial or selfish norm violations.

Table 11: Regression Results of Main Effects and Moderation Effect on the Psychological Barrier concerning Prosocial Norm Violations (Model 10).

Psychological Barrier Prosocial DV Model 10

Coefficient SE Beta

Non-leader/Leader * Status motivation 0.273 0.238 0.666

(41)

6. Discussion

This section will discuss the findings of the research including supported hypotheses, hypotheses that were not supported, and additional findings. All findings are linked to the existing literature by comparing them with the findings of established research. After that, the theoretical contributions and practical implications of the research are mentioned. Finally, the limitations of the study and suggestions for future research are discussed.

6.1 Discussion of relationships between variables

First, attention will be paid to the tested hypotheses and the outcomes of the research will be discussed. This section starts with the relationship between norm violations of others and the psychological barrier. After that, the influence of the psychological barrier on imitation is discussed. This is followed by a discussion about the main effect of norm violations of others on imitation. Finally, the mediating effect of the psychological barrier on the relationship between norm violations of others and imitation will be addressed.

6.1.1 Relationship between norm violations of others and the psychological barrier It was expected that norm violations of others lower the psychological barrier to exercise the same behavior. Furthermore, it was discussed that norm violations of a leader are more likely to lower the psychological barrier compared to those of a non-leader because leaders are more likely to act as a role model for subordinates. This study found significant support for these predictions.

Shannon (2000) describes norm violations as “collective expectations for the proper behavior of actors within a given identity”. Few researchers have paid attention to norm-violating behavior with imitating behavior as a potential consequence. Therefore, the current

Why do we cross the line? : assessing the effect of norm violations on imitating behavior within an organizational context