Will employees suffer from autocratic leadership behavior or is this behavior sometimes justified? : a closer look at group-leader prototypicality

(1)

1

Will employees suffer from autocratic leadership behavior or is

this behavior sometimes justified? A closer look at group-leader

prototypicality.

Master Thesis

Faculty of Business and Economics

University of Amsterdam

Supervision: dr. A.H.B. de Hoogh

Authors:

S.D. Verhoef 10636145

(2)

(3)

3

Abstract

Autocratic leader behavior is quite often seen as negative for team spirit and is also known to be responsible for creating different forms of stress among teams. However, some theories debate the negativity of autocratic leader behavior. Some researchers argue that autocratic behavior sometimes provides guidance through centralized decision making and even boost team morale. Social differences in group norms and values between team members and leaders are presumed to influence outcomes of autocratic leader behavior. The way in which team members and their leaders are similar is called group leader prototypicality. We propose that when leaders scores high on prototypicality this should make the relationship between autocratic leader behavior and procedural fairness more positive compared to leaders who score low in prototypicality. Subsequently, the relationship between procedural fairness and counterproductive work behavior should become less positive through these conditional effects of prototypicality. We tested the expectations in our study which consisted of N = 266 members of the Dutch working society. We did not found support for the moderating role of group leader prototypicality. We did find support for a mediating effect of procedural fairness between autocratic leader behavior and CWB. Unsuspectedly group leader prototypicality was found to mediate the relationship between autocratic leader behavior and counterproductive work behavior as opposed to our moderation expectations.

(4)

4

Introduction

Due to demanding job climates of large companies and multinationals today, it is not strange that autocratic leadership is still used by many managers. When workers find themselves in an ambiguous and highly demanding environment they tend to foster a more autocratic leadership style which facilitates more direction for employees (Bass & Bass, 2008). Given that leaders in demanding climates tend to act in a more selfish and competitive way, they also tend to use less follower oriented leadership behaviors (Van de Vliert, 2006). Businesses in demanding markets focus heavily on strategies to optimize profits and shareholder value. Certain behaviors suit these environments better than other behaviors and according to Pearce et al. (2003) these behaviors are of a more autocratic nature. Often leadership behaviors are characterized by making decisions without consulting employees, heavily criticizing poor work, ensuring that subordinates follow procedures and ensuring subordinates work up to their capacity.

In most of the cases autocratic leadership behavior is perceived as negative by the workers and resulting in lower motivation and satisfaction and therefore less productivity (De Cremer, 2006). Power centralization during decision making may activate team members' feelings of being undervalued and wronged (Adams, 1965; Anderson & Brown, 2010). These feelings may increase the perception of inequity (Muller, 1985; De Hoogh, Greer & Den Hartog, 2015) and thereby reduce perceptions of procedural fairness. The way employees perceive decision making procedures and allocation of resources by their managers determines the amount of procedural fairness (Lind & Tyler, 1988; Tyler, 1989).

De Cremer and Van Knippenberg (2003) illustrate that a leaders’ influence on cooperation is determined by the perceived fairness by employees of the managers’ decision making procedures. If followers feel treated unfair by their superiors they tend to be less motivated and satisfied to perform their jobs. In many cases followers react with a form of resistance against being treated unfairly (Priesemuth, Arnaud & Schminke, 2013). This might result in behaviors like counterproductive work behavior (CWB; Cohen-Charash & Spector, 2001). CWB consists of acts which are harmful to the company and the people working at the company, or behaviors countering the interests of the company

(6)

6 (Sackett & DeVore, 2002; Spector & Fox, 2005). It is at the interest of every company to minimize these harmful acts and behaviors because they inhibit productivity (Kelloway, Francis, Prosser & Cameron, 2010).

However, some studies show possible evidence of autocratic leadership being perceived as less negative compared to earlier research (Van Knippenberg, 2005; Giessner, Van Knippenberg & Sleebos, 2009; Giessner, van Knippenberg, van Ginkel & Sleebos, 2013; Koivisto & Lipponen, 2015). The study by Giessner et al. (2009) helps explaining the concept of group leader prototypicality. In their study they argue that the greater the trust in the leadership given to a prototypical leader compared to a non-prototypical leader is an important mechanism through which non-prototypical leaders receive more follower endorsement on their decision making procedures. In the same study they show evidence for a higher leader effectiveness for group prototypical leaders compared to non-prototypical leaders.

These results can be explained by the social identity theory of leadership which outlines how group membership contributes to self-definition. Merging and aligning one’s self-concept into the groups identity creates group prototypes and thoughts (Pierro et al., 2005). The more contingent the leaders behaviors are with these group prototypes the more prototypical leaders are and thus trusted more by their followers (Hogg, 2001; Van Knippenberg 2011). This results in prototypical leaders being perceived as more effective after failure, compared to their non-prototypical counterparts (Giessner et al. 2009). Combined, these findings might suggest that autocratic leaders which are group prototypical might be trusted more and perceived to be higher in procedural fairness are more effective compared to autocratic leaders who are non-prototypical to the group, even in a case of a failure.

To date, little effort has been done to directly look at the role of leader group prototypicality in the link between autocratic leadership, procedural fairness and CWB. In this study we especially focus on the possible moderating effects of leader prototypicality on the relationship between autocratic leadership and perceived procedural fairness and subsequent follower CWB.

Autocratic leadership is still part of many company cultures these days (Van de Vliert, 2006) and is perceived as more negative than positive (De Cremer, 2006). Therefore, it is at the interest of every company to investigate the degree to which followers perceive decision making procedures as fair and to which degree this is affected by the prototypicality of their leaders. Because CWB is shown by

(7)

7 employees who feel treated unfairly (Cohen- Charash & Spector, 2001), and CWB is harmful to the company (Spector & Fox, 2005) it is imperative to keep this outcome as low as possible (Kelloway et al., 2010).

In this study we want to investigate the possible moderating effects of leader group prototypicality on the relationship between autocratic leadership and procedural fairness and the relationship between procedural fairness and CWB (see Figure 1). Based upon past research (Cohen-Charash & Spector, 2001; Pierro et al., 2005; Van Knippenberg, 2005; De Cremer, 2006; Giessner, Van Knippenberg & Sleebos, 2009; Koivisto & Lipponen, 2015) we presume that when an autocratic leader scores high on group prototypicality, this might result in a higher perceived procedural fairness and a lower amount of subsequent CWB.

First we will explain the different concepts mentioned above in more detail as part of the literature review. Beginning with the conceptual model which is at the beginning of the literature review. It is key to understand all the relationships between the different concepts in detail. The literature review is followed by the research design in which we will explain our research model, our method to test our hypothesis, information about our sample and the procedures and the scales we used to measure the concepts and will help us answering our research question. The Results will consist of the statistical analysis of our measurements which is followed by a discussion in which we debate theoretical and managerial implications and limitations to our study. This is followed by the final conclusion in which we answer the research question.

(8)

8

Literature review

Figure 01: Theoretical model

Historical perspective of autocratic leadership

The concept of autocratic leadership stems from early experimental studies on schoolboys by Lewin, Lippitt, and White (1939) and Lippitt (1940). Autocratic leadership is usually characterized by behaviors focused on the centralization of decision-making and concentrating power towards allocation of resources (Jago, 1982) through which the leader controls every aspect of his followers and their activities, without considering the input of the follower (Sauer, 2011). Specific behaviors illustrated by Bass and Bass, (2008) include ordering group members around, telling them what to, making decisions without consulting the employees, ensuring that employees follow exact procedures, overly criticizing poor work, and ensuring that employees work up to their capacity.

Based on the literature on demanding environments and job pressure in today’s markets by Van de Vliert (2006) and Demerouti, Bakker and Bulters (2004) autocratic leadership behavior is still much prevalent. Therefore, in this modern era we should be looking for a possibility to embrace autocratic leadership behavior and try to use it as an advantage instead of portraying it’s as an enemy of modern businesses or fail trying to eliminate it.

Autocratic leadership, procedural fairness and counterproductive work behavior

Autocratic leadership is usually characterized by behaviors focusing on the centralization of decision making and power exertion (Jago, 1982). According to Pearce et al. (2003) these autocratic

(9)

9 leadership behaviors mostly consist of heavily criticizing poor work, ensuring that subordinates follow procedures and ensuring that subordinates work up to their capacity. Also, ordering team members around, telling them what to do and making decisions without any form of consultation are often seen as part of these specific behaviors (De Hoogh & Den Hartog, 2009). By acting in such way superiors can create a clear and defined group hierarchy (Bass & Bass, 2008).

In general, autocratic leadership is seen as a negative form of leading people due to the excessive use of power and its effects on team climate and performance (De Cremer, 2006). In a study measuring absenteeism and sickness among Swedish women Nyberg, Westerlund, Hanson and Theorell (2008) conclude that leadership in terms of being more centralized in decision making, self-centered and more autocratic by nature is regarded less effective in leading teams nowadays compared to being more participative.

Repeated negative leadership behaviors consisting of heavily criticizing poor work, ensuring that employees follow exact procedures and ensuring the employees work up to their capacity are associated with negative feelings (Tepper, 2000). Workers confronted with these kinds of leader behaviors might feel treated unfair and left out. In turn, this might result in misbehavior such as deliberately spilling supplies, gossip, delay specific tasks and other behaviors which are harmful for the company (Priesemuth, Arnaud & Schminke, 2013). This type of misbehavior can be seen as a result of heavily centralized decision making procedures without any room for discussion. The lack of possibilities to provide input and the lack of rewards for their duty may fuel resistance against the managers (Lind, 2001).

The literature on organizational justice defines the concept of procedural justice as the perceived fairness by employees in case of formal procedures used in resource allocation and decision making by superiors (Lipponen et al., 2005). Colquitt, Noe and Jackson (2002) describe the dimension as a certain work climate or environment, an emergent state which reflects how a team as a whole is treated in terms of resource allocation and decision making procedures by the leading figures. As they point out in their study, perceptions of fairness are mostly centered around the leader figure. Leaders are the figures that most often make the final decisions because they are responsible for enacting changes to procedures and company policies. For example, a leaders decision making behavior, which is often guided by his

(10)

10 personal values, should be in congruence with his team members values to be aligned e.g. having a low power distance. Power distance can be described as the extent to which leaders and followers evaluate the distance in rank and the distribution of power over decision making (Hofstede, 2001). This congruence of personal feelings results in team members having the feeling they are treated fair because none of their feelings and values expectations are violated. Conversely, if there is an incongruence of power distance e.g. personal values of leaders and team members are not aligned, there will be a violation of expectations on the side of the team members. This may ultimately lead towards the feeling they are treated unfair (Cole, Carter & Zhang, 2013; Colquitt, Rodell, Zapata, Scott, Long & Conlon, 2013). In the study by Cole et al., (2013) they showed that procedural fairness mediated the relationship between power distance and team effectiveness including reciprocal behaviors like CWB. In turn, drawing on the theory and empirical evidence that suggests when procedural fairness is operationalized at the team level, this might be a key predictor of the team effectiveness and reciprocal behaviors like CWB (Cole et al., 2013). Colquitt et al., (2013) found empirical evidence that one’s trust in- and commitment to the leader is mediating the relationship between procedural fairness and both task performance and CWB. This illustrates the importance of the feeling of being treated fair because this is an important mechanism to predict organizational outcomes like effectiveness which is positive and CWB which is negative. As discussed earlier and outlined by Spector and Fox (2005), CWB is harmful for the company and needs to be eliminated from the work floor or reduced to the minimum level possible. In addition, Huo, Smith, Tyler and Lind (1996) argue that communicating identity and self-worth towards the employees is regarded to be an important task for group leaders. Especially whether or not employees feel respected members of the group.

Combining these findings by Cole et al. (2013), Colquitt et al. (2013), Huo et al. (1996), Lipponen et al. (2005), Priesemuth et al. (2013), Tepper (2000), and we can conclude that leaders should have the task of making the team members feel appreciated and treated fairly in order to keep them satisfied and productive. As explained by Tepper (2000), De Cremer (2006) and Nyberg et al. (2008) autocratic leadership is perceived as more negative because it restricts workers to give input during the decision making procedures and resource allocations. A leader who centralizes decision making

(11)

11 procedures and adheres a higher power distance is considered to be a more autocratic leader (Cole et al., 2013; De Cremer, 2006; Tepper, 2000). This is confirmed by the literature described by Cole et al. (2013) and Colquitt (2013) which states that when leaders are violating team members values and expectations this leads towards higher power distances which in turn leads to the feeling of being treated unfair. Therefore we hypothesize the following:

H1: Autocratic leader behavior is negatively related to perceived procedural fairness

Undermining, sabotaging the organization’s goals, tasks resources, and effectiveness, motivation, well-being or job satisfaction of team members are common outcomes of people who repeatedly feel treated unfairly (Aquino, Lewis & Bradfield, 1999; Douglas & Martinko, 2001; Einarsen, Aasland & Skogstad, 2007). These behaviors can be considered as being counterproductive and undermining organizational effectiveness in the broad sense. Adding to this, Gruys and Sackett (2003) describe these behaviors as counterproductive work behaviors. These behaviors are intentional and conscious by nature that might be harmful to the legitimate interests of the particular organization or team. A legitimate interest is a legal standard by which a person may determine that one party or itself has a tangible stake in the question to be decided, i.e. does this person have the right or the feeling that he or she has the right to intervene with the leading figure. These behaviors are a set of various and distinct acts that share characteristics of intended violation, intentions to harm the organization and its stakeholders such as its clients, coworkers and their superiors and (Spector & Fox, 2005).

In a study on organizational justice by Cohen-Charash and Spector (2001) there is empirical evidence found that job performance and counterproductive work behavior are outcomes of the perceived justice by the team members. Most often this is related to procedural justice which is the same as procedural fairness in this case. To the extent employees perceive their organization to be unfair because it facilitates unfair procedures for decision making and resource allocations, team members will develop negative attitudes toward the organization such as anger, lower trust and commitment (Skarlicki & Folger, 1997; Skarlicki, Folger & Tesluk, 1999). Negative attitudes and emotions lead to employees

(12)

12 not having incentives to work in favor of the organization. Moreover, they might lead employees to act against the organization (Cohen-Charash & Spector, 2001; Fox et al. 2001; Fox & Spector, 1999).

As explained by Cole et al. (2013) and Colquitt et al. (2013) CWB is an reciprocal outcome which emerges from the feeling of being treated unfair by its leading figure. The feeling of being treated unfair can be the result of antecedents like high power distance, centralized decision making and overly criticizing poor work. Following the literature concerning the negativity of being treated unfair and its subsequent outcomes which we mentioned earlier, we hypothesize the following:

H2: Procedural fairness is negatively related to counterproductive work behavior

Social identity theory and the moderating role of leader group prototypicality

Social identity captures the group membership aspect of identity. This refers to that part of an individual's sense of self, which is rooted in the individual's group membership. The part that gives people a sense of “we” instead of “I” includes other members of the group in one's sense of self. This will create a true feeling of belonging to the group and will make workers act like a team (Van Knippenberg, 2011).

Core to the social identity theory of leadership is the notion that group leaders, like all members of a group, differ to the extent to which they are perceived to represent or embody the group prototype or seen as group prototypical (Hogg, 2001). This social identity is one’s self-definition, and the perception of group prototypicality thus has its roots in individual cognition, even when such perceptions often are socially shared. A leader is perceived to be group prototypical when he or she embodies the group (team, organizational) identity, to embody “who we are”. The leader derives influence from the implicit perception that he or she represents what is group-normative (Giessner et al., 2009). In addition, group leader prototypicality speaks to subordinates' motivation to pursue the group's interest. Therefore, group prototypical leaders are perceived to embody the group’s identity, norms, values and whatever the team considers important (Mayer, 1995; Hogg, 2001; van Knippenberg 2011).

(13)

13 Group leader prototypicality induces trust in the leader's motivation to pursue the group's best interest and therefore followers are more open to the influence of the prototypical leader. The more team members perceive the leader to be group prototypical the more effective the leader should be even when this leader makes wrong decisions once in a while (Giessner & van Knippenberg, 2008). Giessner, van Knippenberg and Sleebos (2009) show that group members’ benevolence of trustworthiness increases when a leader embodies group prototypes, values and believes. According to Priesemuth et al. (2013) this increased benevolence of trustworthiness is likely to increase the amount of procedural fairness perceived by the group members due to the leader, which has the trust and the commitment of the team members (Colquitt et al., 2013). This also lowers the power distance between the leader and his followers which creates a feeling of equality in the sense of values and believes. Therefore the leader is less likely to violate values expectations from team members (Cole et al., 2013; Colquitt et al., 2002).

Workers might perceive autocratic leader behavior as a bad thing because he or she might not embody the group values and believes i.e. the group prototypes (Skarlicki & Folger, 1997; Nyberg et al., 2008; Giessner et al., 2009). However, when group values and believes support autocratic leadership behaviors, there is a possibility that this group will accept or tolerate autocratic leadership behavior because the leader embodies the group prototype. If this is true, than the relationship between autocratic leadership and perceived procedural fairness should be less negative for leaders who are more group prototypical. Therefore we hypothesize the following:

H3: Leader group prototypicality moderates the relationship between autocratic leadership procedural fairness, such that autocratic leadership is less negatively related to procedural fairness when leader group prototypicality is higher.

The mediating role of procedural fairness

We propose that the interaction between autocratic leadership behavior and group leader prototypicality affects perceived procedural fairness, which in turn might affect the amount of CWB reported by the team members. When an autocratic leader does not embody group norms, values and

(14)

14 believes i.e. low group leader prototypicality as argued by Giessner and Van Knippenberg (2008), autocratic leadership is likely to create a sense of procedural unfairness for the team members. Managers heavily centralizing decision making procedures might result in team members showing resistance and negative attitudes towards this kind of control (Priesemuth et al., 2013). Without a sense of higher procedural fairness at the follower level, the team members are less likely to invest themselves at work, concentrating themselves on their jobs which in turn might lead to distractions like CWB (Cohen-Charash & Spector, 2001; Fox & Spector, 1999; Fox et al., 2001).

In contrast, when group leader prototypicality is high, there is a possibility that team members are more willingly accepting the autocratic behaviors of the leader because they are part of the groups’ norms, values and believes. This is likely to result in less resistance and more positive attitudes towards the decision making procedures and resource allocations which is outlined by Skarlicki and Folger (1997) and in Giessner et al. (2009), which in turn are associated with a higher amount of procedural fairness among the team members.

As explained by Giessner and van Knippenberg (2008), when team members perceive decision making procedures and resource allocations as fair, this is merely because the followers trust their leaders because the leader embodies the norms, values and believes of the group. This benevolence of trustworthiness is therefore likely to reduce the negative relationship between autocratic leadership and procedural fairness due to lower power distance and less centralized decision making procedures and resource allocations (Cole et al., 2013; De Cremer, 2006; Lipponen et al., 2005). Team members will show more positive attitudes towards the decision making procedures and therefore are likely to show lower amounts of CWB because there is no reason for resistance (Skarlicki & Folger, 1997; Cohen-Charash & Spector, 2001).

Thus, as reflected in our theoretical model illustrated in Figure 1 and according to the literature on autocratic leader behavior, group leader prototypicality, procedural fairness and CWB we propose that the interaction between autocratic leadership with leader group prototypicality indirectly relates to CWB through its relationship with perceived procedural fairness. Therefore we hypothesize the following:

(15)

15 H4: Autocratic leadership is related to CWB via conditional indirect effects, such that the interaction between autocratic leadership and leader group prototypicality is less negatively related to perceived procedural fairness, which in turn is negatively related to CWB.

Research design

Method

This is a cross-sectional study because we took a sample from the Dutch working society. In this study we want to test certain hypotheses about the probability of different relationships between the given variables. We want this study to help us have a better understanding in the mechanisms between autocratic leadership, procedural fairness, CWB and the role of group leader prototypicality within an organization. The research on this topic is done through quantitative data analysis on data obtained by using a cross-sectional survey design. The questionnaire which consists of validated scales, is translated into Dutch especially for the Dutch working society. We choose to translate the questionnaire scales because of the reduction of bias possibly produced by people which would not be able to understand everything in the questionnaire. The questionnaire is solely addressed to the team members which are working under the guidance of a manager/leader or need to report to a superior or supervisor. The questionnaire was distributed digitally via the company and personal networks via an anonymous link which redirects the respondent to the digital survey on the Qualtrics website. This website provides additional help and a well-defined tutorial for designing a proper questionnaire. The website offers a structured template for an adequate survey. Team members need to answer questions about the behavior of their superiors within company teams, their personal and group level values among the group, questions regarding fairness of decision making procedures and questions regarding CWB. Questionnaires containing missing data are excluded from the dataset. After the collection of the data it was digitally analyzed with IBM SPSS 23 to test the hypotheses for significant correlations, mediation and moderation effects.

(16)

16

Sample and procedure

The population in this study consists of a cross-sectional sample of 266 respondents being part of a corporate team or working under supervision in the Dutch working society. The data is gathered through purposive heterogeneous sampling because this population is rather large and diverse. However the research will be conducted through a non-probability convenience sampling because total randomness is tough to thrive for. Besides questions concerning the content of the study and its main variables, questions about age, gender, tenure and educational background are asked to check the demographics which in turn, are used as control variables.

Survey packets were introduced via a digital link that was spread throughout the different personal networks such as Facebook, Twitter, WhatsApp, LinkedIn, family, friends, colleagues and acquaintances. Out of the 454 returned surveys, 266 surveys were filled out completely (59% response rate). The minimum amount of respondents for a survey, consisting of quantitative numerical data, to be reliable should be approximately 200, which in this case is sufficient. The average age of the respondents was 35 years (ranging from 18 to 62), 48% of the respondents were male and 52% were female and the average tenure was 6 years . Four percent of the respondents are working or have been working less than a year with their leader figure and 20% of the people are working or have been working 10 or more years with their leader figure.

Measurements and scales

All variables are numerical except for gender and educational background and other demographics which are categorical. The variables age, gender and tenure needed to be recoded in such way that they were solely numerical because many respondents added the word “year” after the number. This was a coding mistake but this was recoded after data collection during the data cleaning procedure. The questions used for each construct were directly translated to Dutch and checked for same meaning and conservation of Cronbach’s alpha. The actual translations and the scale items can be found in appendix 1.

(17)

17 Autocratic Leader Behavior

Autocratic leader behaviors were measured with five items each from the validated Dutch Charismatic Leadership in Organizations questionnaire (CLIO; De Hoogh, Den Hartog, & Koopman, 2004). The items have a 7-point response scale, ranging from 1 (strongly disagree) to 7 (strongly agree). Only the items concerning autocratic leadership behavior were used. Autocratic Leader Behavior Scale includes dominant behavior and leaders protecting their own position, such as making decisions alone without asking for the suggestions and harshly telling subordinates what to do. Examples of items are “My leader is bossy and orders subordinates around” and “My leader makes sure that his/her own interests are always met.” Coefficient alpha in both samples was .70.

Group Leader Prototypicality

For the measurement of group leader prototypicality the participants responded to the following 3 items derived from van Knippenberg and van Knippenberg (2005): “My supervisor is a good reflection of the people working in our team”; “My supervisor represents what characterizes our team”; “My supervisor has a lot in common with the team”. Participants responses are recorded on a 7-point scale ranging from 1 (absolutely not agree) to 7 (absolutely agree). A composite leader group prototypicality score was computed by summing across responses to each item (Cronbach’s Alpha was .93).

Procedural Fairness

For the measurement of procedural fairness climate at the workplace a 7 item questionnaire was used with a 7-point scale, ranging from 1 (to a small extent) to 7 (to a large extent). The seven items are part of the justice measurement questionnaire used in the validation study by Colquitt (2001). Question 1 and 2 are developed by Thibaut & Walker (1975) and questions 3 to 7 are developed by Leventhal (1980). The alpha reliability was (.95) for the procedural justice scale based on department-level data.

(18)

18 Counterproductive Work Behavior

For counterproductive work behavior a 10-item version (5 items targeting the organization and 5 targeting people) of the Counterproductive Work Behavior Checklist (CWB-C; Spector et al., 2006) measured CWB. The items have a 5-point response scale, ranging from 1 (never) to 5 (every day). A total score was computed for our purposes. The internal consistency (coefficient alpha) of the CWB-C in the current sample averaged .78 for the two employee forms.

Because the questions concerning CWB were counterintuitive they needed to be recoded into a new variables which has been mirrored before centering them and recoding them into the new variable. The variable names must consist of not more than 8 characters because PROCESS macro’s in SPSS cannot run variables that exceed the length of 8 characters (Preacher & Hayes, 2004).

Control Variables

The socio-demographics age, gender and tenure have been known moderators of work-related stressors and job performance (Shirom, Gilboa, Fried & Cooper, 2008). Due to the high diversity of the working society today these demographics are able to influence many other work related social dimensions like separation, variety and disparity and therefore business outcomes (Harrison & Klein, 2007). In our case we need to rule out the possibility of multicollinearity between the predictor variables, dependent variables and the covariates age, gender and tenure. Therefore we added them as control variables.

For the collection of these general demographics the questions which are used to gather information about age, gender, tenure, education and the industry in which someone is working can be found in appendix 1. These variables are being used as control variables to rule out the possibility of multicollinearity such as excessive correlations.

(19)

19

Results

Analysis

The correlations and descriptive statistics can be found in Table 1. Prior to analysis all variables were mean centered and recoded into different variables (Aiken & West, 1991). To test our hypotheses we will use a procedure which is outlined by Preacher et al. (2007) for examining a set of relationships which are proposed in Hypotheses 3 and 4.

By means of a two-step regression analysis and controlling for age, gender and tenure, we will determine if the control variables significantly alter the other variables and relationships. Once age, gender and tenure have been excluded from our analysis due to the non-significant interference with the other variables and relationships we can check if conditions for hypothesis 1 and 2 are met or rejected. When the control variables do not significantly impact the other variables and relationships, the correlations of hypothesis 1 and 2 are the same as in the regression analysis (Becker, 2005). In the correlation matrix which is depicted in Table 1, we find the significance levels for all variables.

Then first, a mediation model is used to check for mediation between the relationship of autocratic leader behavior and counterproductive work behavior via procedural fairness. Secondly, we check the conceptual model for possible moderation by group leader prototypicality between autocratic leader behavior and procedural fairness, as shown in Figure 1. To be more specific, we use the MED macro model for simple mediation (Model 4, Preacher & Hayes, 2004) and the MODMED macro model for moderated mediation (Model 7, Preacher & Hayes, 2004) which is providing relevant and sufficient results for the analysis of our four hypotheses.

Correlations

Before we want to step towards the analysis we need to conclude if the control variables age, gender an tenure do not inflict too much with the other variables causing excessive bias and therefore unwanted results. As control variables in the main analyses, we checked whether age, gender and tenure had a significant impact on the results. Age, gender and tenure did not significantly alter the variables

(20)

20 or relationships in our study, so we removed it from subsequent analyses to conserve statistical power (e.g., Becker, 2005) and report the analyses without the control variables.

If we take a look at Table 1, the correlation matrix, we can see that there is a significant negative correlation (B = -.145, r = -.15, p <.05) between autocratic leader behavior and procedural fairness, this means that Hypothesis 1 is accepted. This negative relationship means that a high score on autocratic leader behavior is significantly related to a low score on procedural fairness. We cannot make any causal claims on the direction of the relationships because we did a cross-sectional study. The relationship between procedural fairness and counterproductive work behavior is also significantly negative at the two tailed level (B = -.126, r = -.30, p <.01). Conditions for hypothesis 2 are therefore met and thus accepted. This means that workers who feel treated unfairly during decision making procedures report showing significantly more counterproductive work behavior, at least at a 95% certainty level.

Table 1: Means, Standard deviations, Correlations and Reliabilities between parentheses.

Variables M SD 1 2 3 4 5 6 7

1. Age 34,8 10,98 -

2. Gender 1,52 0,5 0,04 -

3. Tenure 5,91 6,05 ,61** 0,03 -

4. Autocratic leader behavior 4,45 1,33 0,03 -0,05 -0,03 _(,73)

5. Procedural fairness 4,67 1,28 -0,1 0,07 0,02 _{-,15* (,85)}

6. Counterproductive work behavior 4,36 0,54 0,06 -0,11 0,1 -0,09 -,3** _(,86) 7. Group leader prototypicality 4,54 1,84 -,15* -0,11 0,01 -0,16* ,57** -,24** (,93)

**. Correlation is significant at the p<0,01 level (2-tailed). *. Correlation is significant at the p<0,05 level (2-tailed).

Mediation by procedural fairness

In the first step we measured the direct effect between autocratic leader behavior and procedural fairness. In the second step we measured the effect between autocratic leader behavior and CWB by adding procedural fairness to the model as a mediator variable and CWB as the outcome variable. The results for the mediation analysis are presented in Table 3 and reveal a significant direct effect between autocratic leader behavior and procedural fairness (B = -.145, p = .013) in step one. After adding CWB

(21)

21 as the outcome variable and procedural fairness as a mediator variable in step two, the direct effect between autocratic leader behavior and CWB is non-significant (B = .020, p = .416) while the direct effect between procedural fairness and CWB is highly significant (B = -.126, p <.001). It seems that procedural fairness fully mediates the relationship between autocratic leader behavior and counterproductive work behavior because the direct effect (B = .145, p = .013) has become non-significant (B = .020, p = .416) after adding procedural fairness (B = -.123, p < .001) to the model.

The indirect effect of B = .020 means that two workers who differ by one unit in their reported autocratic leader behavior estimated to differ by .020 units in their reported counterproductive work behavior as a result from those who perceive less procedural fairness as a result of a higher autocratic leader behavior score, which in turn translates into higher CWB. This indirect effect is statistically different from zero, as revealed by a 95% BC bootstrap confidence interval that is entirely above zero (CI .002 to .039).

The direct effect of Autocratic leader behavior, c′ = .020, is the estimated difference CWB between two workers experiencing the same level of Procedural fairness but who differ by one unit in their reported autocratic leader behavior, meaning that the person experiencing more autocratic leader behavior who is equally committed is estimated to be .020 units higher in his/her reported CWB. This direct effect is not statistically different from zero, t = .814, p = .416, with a 95% confidence interval (CI -.028 to .067).

Because the direct effect is not significant and the indirect effect is significant, procedural fairness fully mediates the effect between autocratic leader behavior and counterproductive work behavior. This means that workers who experience higher autocratic leader behavior are likely to perceive lower fairness in decision making procedures which is likely to result in a higher reported CWB.

(22)

22 Table 3: Mediation (MED)

Consequent

M Procedural

Fairness) Y (CWB)

Antecedent Coeff. SE p Coeff. SE p

X (Autocratic Leader Behavior) a -.145 .058 .013 c' .020 .024 .416 M (Procedural Fairness) -- -- -- b -.123 .025 <.001 constant i1 5.316 .271 <.001 i2 2.127 .174 <.001 R2 _=.023 _R2 _=.091 F (1,264) = 6.197, p=.013 F (2,263) = 13.126, p < .001

Moderated mediation by group leader prototypicality

For the second analysis we added leader prototypicality to the mediation model as a proposed moderator variable on the effect between autocratic leader behavior and procedural fairness and ultimately on CWB. As we can see in the moderated mediation analysis depicted in Table 4 and the coefficients in Table 5, the results indicate a non-significant effect of experiencing autocratic leader behavior on counterproductive work behavior via procedural fairness to be contingent on the employees group leader prototypicality. This is evidenced by a statistically non-significant interaction between XW in the model of Y (c’1 = .20, p = .416). A closer inspection of the conditional effects indicates that the direct and indirect effect is not conditioned by leader prototypicality. First, the direct effect between autocratic leader behavior and CWB appears as non-significant (effect B = .020, SE = .024, CI: -.028 to .067). Secondly, the indirect effect between autocratic leader behavior and CWB via procedural fairness was not contingent on leader prototypicality thus being a non-significant indirect effect (B = .012, SE = .009, CI: -.003 to .035). From this data we can conclude that group leader prototypicality does not have a significant effect on the relationship between autocratic leader behavior and procedural fairness and ultimately on counterproductive work behavior.

(23)

23 If we look at the correlation depicted in Table 1 we can find a significant negative relationship between autocratic leader behavior and procedural fairness (r = -.151, B = .145, p = .013) which is in line with our prediction. This means Hypothesis 1 is accepted. The negative relationship (r = -.3, B = .126, p < .001) between procedural fairness and CWB is resulting into the acceptance of Hypothesis 2. Hypothesis 3 is rejected because the interaction between MV (leader prototypicality) and IV (autocratic leader behavior) is non-significant (B = .021, p = .378). Hypothesis 4 predicted that conditional indirect effects by leader prototypicality would have a negative interaction effect on the relationship between autocratic leader behavior an procedural fairness and subsequently on CWB. This hypothesis was rejected because the interaction effect was also non-significant (B = .020, p = .416).

Table 4: Moderated Mediation (MODMED)

Consequent

M (Procedural

fairness) Y (CWB)

X (Autocratic Leader Behavior) a1 -.056 .049 .254 c'1 .020 .024 .416 M (Procedural Fairness) -- -- -- b1 -.123 .025 <.001 W (Leader Prototypicality) a2 .392 .035 <.001 c’2 .034 .058 .552 XW (Intraction) a3 .021 .024 .378 c’3 -.015 .013 .262 constant i1 4.678 .065 <.001 i2 2.214 .222 <.001 R2_=.336 _R2_=.091 F (3,262) = 44.09, p < .001 F (2,263) = 13.125, p < .001

(24)

24 Table 5: Estimated coefficients of the moderated mediation model

Predictor B SE F R2

Constant 4.332** .299

Age .002 .008

Gender .177 .177

Tenure .001 .014

Autocratic Leader Behavior -.052 .049

Group Leader prototypicality .395** .036

Autocratic Leader Behavior x Group Leader

Prototypicality .022 .024 22.318 .341 Constant 2.224** .189 Age .003 .004 Gender -.103 .064 Tenure .004 .007

Autocratic Leader Behavior .018 .024

Procedural Fairness -.119** .025 6.209 .107

Note. N = 266 Respondents * p < .05 ** p < .01 Dependent variable: CWB

Exploratory results

If we go back to the correlation matrix shown in Table 1, we can see a rather strong and significant positive relationship between group leader prototypicality and procedural fairness (r = .57, B = .399 p < .001). It also has a strong relationship with the other variables autocratic leader behavior (r = -.16, B= -.217, p < .01) and counterproductive work behavior (r = -.24, B = .817, p < .001). So there a possibility that group leader prototypicality is not a moderator but one of the main effects itself. After running another macro model 4 as used in Preacher et al. (2004) to test for mediation of group leader prototypicality between autocratic leader behavior and CWB we found a significant effect (B = -0.204, p = .015), for the mediation of leader prototypicality between autocratic leader behavior and CWB. These findings are shown in Table 6 Group leader prototypicality seems to fully mediate the relationship between the IV and the DV. This might mean that when team members score their leader as more autocratic this is likely to lead towards higher levels of CWB which is explained by the similarity of the leaders and their teams members in terms of group norms, values and beliefs. The direct effect between autocratic leader behavior and CWB drops significantly when group leader prototypicality was added to the model as a mediator (B = .021, p = .398). This might mean that when

(25)

25 team members score their leader as more autocratic, they generally perceive their leader as less prototypical and this might lead to higher levels of CWB which is likely to be explained by the dissimilarity of the leaders and their teams members in terms of group norms, values and beliefs.

Table 6: Mediation Group Leader prototypicality (MED)

Consequent

M (Procedural

Fairness) Y (CWB)

X (Autocratic Leadership Behavior) a -.204 .083 .015 c' .021 .025 .398 M (Leader Prototypicality) -- -- -- b -.067 .018 <.001 constant i1 6.614 .627 <.001 i2 1.953 .218 <.001 R2 =.059 R2 =.079 F (4,261) = 4.066, p = .003 F (5,260) = 4.484, p < .001

(26)

26

Discussion

Theoretical Implications

According to our predictions we proposed that autocratic leadership can be less negative for the leader of that specific group hence the leader prototypicality within that group is high. Besides this concept we also proposed that this effect would be facilitated by the feeling of being treated fair during decisions making procedures and resource allocations (procedural fairness). This would be ultimately measured by the report of counterproductive work behavior by the team members. Our study confirms the idea of the negative effect of the presence of autocratic leadership behavior (i.e. low procedural fairness and high CWB). When the team members report their leader to be more autocratic they report a lower procedural fairness and a generally higher CWB.

However, when we add group leader prototypicality to the model, something unsuspected happened. The idea that if the team leader embodied group norms and values and therefore scored high on group leader prototypicality, this should result in higher reported procedural fairness and consequently lower reported CWB. This statement was unfortunately not supported. Group leader prototypicality was not moderating the relationship between de IV and the MV and consequently the DV.

Reviewing the literature about this phenomenon seems to direct us towards the idea that group leader prototypicality should moderate the relationship between autocratic leader behavior and CWB but our study did not provide evidence that this is the case (Giessner & Van Knippenberg, 2008; Giessner, Van Knippenberg & Sleebos, 2009).

The unsuspected exploratory results of the mediation by group leader prototypicality also was a surprise because, based on the literature provided by Giessner et al., 2009; Hogg, 2001; Mayer, 1995; Van Knippenberg, 2011; Knippenberg and Sleebos, 2009, we predicted that leader prototypicality would moderate the effects between autocratic leadership behavior and procedural fairness. The finding that group leader prototypicality was actually mediating the effect between autocratic leader behavior and

(27)

27 CWB was not suspected. But when group leader prototypicality was added to the model as a mediator variable the direct effect between the IV and the DV dropped significantly revealing a full mediation.

Because our study is cross-sectional we cannot make any causal indications but the mediation of leader prototypicality does say something. Group leader prototypicality shows the differences between the team members and their leaders in terms of group norms, values, beliefs and terms like morality (Mayer, 1995; Hogg, 2001; van Knippenberg 2011). Because group leader prototypicality is mediating the effect between the IV and the DV we can say that these differences between the team members and their leaders do have a significant effect on the reports of CWB. Since this behavior is harmful for the company, this should be kept as low as possible (Sackett & DeVore, 2002; Spector & Fox, 2005).

By finding significant mediation effects of both procedural fairness and group leader prototypicality we add to the literature in such way that these constructs have not been yet studied in the context of our theoretical model and framework, such as shown in Figure 1. Both variables group leader prototypicality and procedural fairness seem to be strongly related to counterproductive work behavior. Especially if this is linked to situations where team members share an autocratic leader.

As explained in the literature review there were numerous accreditations and loads of empirical evidence which direct us towards the belief that there is a possibility for a negative relationship between autocratic leader behavior and procedural fairness (De Cremer, 2006; Lind, 2001; Priesemuth et al., 2013; Tepper, 2000). These two constructs have not been formulated in a model similarly like our model, which is showed in Figure 1. Our finding that autocratic leader behavior does have a significant negative relationship with procedural fairness adds to the existing literature on autocratic leader behavior by Bass and Bass, 2008; De Cremer, 2006; Nyberg et al., 2008; Pearce et al.,2003 and Tepper, 2000 as well as on the literature on procedural fairness (Cole et al., 2012; Colquitt et al., 2013; Colquitt et al., 2002; Lipponen et al., 2005).

The finding that procedural fairness does indeed fully mediate the relationship between autocratic behavior and CWB does support the predictions made in our theoretical framework. Many scholars made the suggestion that the feeling of being related unfair during decision making procedures and resource allocations would have a negative effect on behavioral outcomes (Cole et al., 2013;

(28)

28 Colquitt et al., 2012; Colquitt, Noe and Jackson, 2002; Huo et al., 1996; Lipponen et al., 2005). This idea is supported by our findings. By presenting the model whereas procedural fairness is depicted as a full mediator between autocratic leader behavior and CWB is not yet been researched yet. Thereby we contribute to the literature concerning procedural fairness in a way that we a found a significant effect for the full mediation of procedural fairness.

As for CWB, the literature guided us towards the idea that the feeling of being treated fairly during decision making procedures would have a strong effect on the behavioral outcomes of a person under the supervision of an autocratic leader (Cohen-Charash and Spector 2001; Folger, 1997; Fox et al. 2001; Fox & Spector, 1999; Skarlicky et al., 1999; Spector & Fox, 2005). As predicted we found a significant negative effect between procedural fairness and CWB. This adds to the existing literature by proving that there indeed is a significant negative relationship between procedural fairness and CWB.

The literature on group leader prototypicality suggested that there would be situations possible were autocratic leaders would be seen as less negative by their team members and therefore would report higher procedural fairness which ultimately results in lower reported CWB (Giessner et al., 2009; Hogg, 2001; Mayer, 1995; Van Knippenberg, 2011; Knippenberg & Sleebos, 2009). Unfortunately we were not able to find evidence for the moderating effect of group leader prototypicality. But we did find evidence for the mediating effect of group leader prototypicality on the relationship between autocratic leader behavior and CWB. We did not found much literature which supports the idea that the amount of group leader prototypicality explains the amount of CWB that is reported by team members with an autocratic leader. By finding significant evidence for a fully mediated relationship between autocratic leader behavior and CWB through group leader prototypicality we add to the existing literature concerning autocratic leader behavior, CWB and group leader prototypicality.

Managerial Implications

Our study sheds light on the importance of the situation in which team members need to work under the guidance of a superior. Specifically, the importance of the feeling of being treated fairly in a situation where decisions are made and resources are allocated. Besides this, the importance of group

(29)

29 leader prototypicality, which is the similarity between leaders and team members concerning the group norms and values, is shown. The two constructs consisting of procedural fairness and group leader prototypicality both show to be a significant social aspect which in terms can result in lower or higher CWB. As argued by Cohen-Charash and Spector (2001) this behavior is harmful for the company and the people working there. Businesses should invest as much effort as possible to reduce the amount of CWB. Due to the demanding job climates today, which seem to be feeding grounds for autocratic leadership styles, businesses should invest in ways to maximize the feeling of procedural fairness and try to keep the similarity of team members and their leaders as high as possible concerning their group norms and values (Van de Vliert, 2006). Strategies which embrace these changes should yield situations where CWB is uncommon an unwanted by all members of the team and this situation produce fertile grounds for organizational successes and ultimately profitability and efficiency.

Limitations and implications for future research

Because the exploratory findings on the mediation of group leader prototypicality exceeds the scope of our study this is something we highlight for future research as we wonder how the causal relationships and mechanisms work. Particularly between both group leader prototypicality and procedural fairness and their interconnectedness with the other constructs. Despite our findings we cannot make any causal indications on the direction of the relationships we found in our data analysis as a result of doing a cross-sectional study.

There is a possibility that other variables influence the relationships between the variables in our theoretical model. These are variables which are called omitted variables which can cause omitted variable bias, OVB for short. According to Clarke (2005) OVB occurs when a model is created incorrectly and leaves out one or more important constructs. The bias is created when the model tries to compensate for the missing construct by over- or underestimating the effect of one of the other constructs. For instance, social desirability can be a construct of attention and might be a omitted variable itself. As outlined by Thompson and Phua (2005) social desirability bias is a term that describes a type of response bias that illustrates the tendency of survey respondents to answer questions in a way

(30)

30 that will be viewed favorably by others. This can take the form of over-reporting preferred behavior or under-reporting undesirable behavior. The tendency poses a serious problem with conducting research with self-reports, especially questionnaires, like the survey we used in our study. This bias interferes with the interpretation of average tendencies as well as individual differences. Consequently, the need to control for social desirability response is important, although few management studies include the use of relevant scales commonly incorporated in psychology research (Thompson & Phua, 2005). In future research we suggest that we should control for social desirability to minimize bias.

Our theoretical model consists of four variables which we isolated for our research. As mentioned above, there is a good possibility that our respondents were influenced by other variables like social desirability or strong social in-group norms and values specific for one’s group which he or she belongs to. There is also a possibility that our respondents had a different view on the four constructs we presented them. The difference can lie in the way we presented our questions or the way we translated our questions, but this is a hard question to answer. The way our sample might form an isolated group compared to the other inhabitants of the Dutch working society can also be a form of bias to our data. Somehow, there might be specific characteristics e.g. the city they live or work in, which make the sample produce specific data. Which in this case might produce heterogeneous data which is better predictor for generalizing to the population. This is something we could address in a future study by more carefully select population samples which represent the Dutch working society in a more heterogeneous way.

Note that we used a sample of people which are closely related to each other instead of using a more random population sample in which nobody knows each other. According to Podsakoff, MacKenzie, Lee, and Podsakoff, (2003) the sample needs to be heterogeneous to reduce bias for analyzing and interpreting the data as well as generalizing to the population. Because our respondents were all friends, acquaintances or friend of friends, there is a high possibility that the sample is quite homogeneous which could influence the data.

Group norms and values of our sample can be a little skewed compared to a more heterogeneous sample of the Dutch working society which can result is bias. This is a result of what Podsakoff et al., (2003) call spurious/common source data variance. This is error variance that is attributable to

(31)

31 the measurement method, rather than to the constructs the measures are assumed to represent. This systematic error variance is shared among the measured variables and is introduced as a function of the same method or source and therefore contributes to the proposed bias. For instance, when someone wants to fill out the survey electronically, but this person is not used to computers and electronic gadgets. This is presumed to influence the data which is collected from this respondent in such way that the inter-correlations can be inflated or deflated depending on several factors. For example, mistakes can be made or the translation is not done correctly which in this case might lead to a respondent interpreting the question wrong. We directly translated the measurement scales from English into Dutch which can lead to a sample that is possibly not totally homogeneous because some questions might have a different undertone or context in different languages. This can be the cause of the rejection of our

3

rd_{and 4}th

hypothesis, but to be sure we need to rule out all the other possibilities in a follow up study with a large scale factor analysis to identify all other variables which correlate significantly with our constructs. This might have influenced our results in such a way that our model does not show all predicted significant moderation and mediation effects.

Collecting our data from the respondents took just over four days. There is a possibility that if we replicate the study with five times more respondents and select them more randomly from the Dutch working society population, in a larger time span, controlled for social desirability and rule out the possibility of OVB, there is a chance we find different results.

Conclusion

Although we might predict that there are situations where autocratic leadership is being perceived as less negative as opposed to what people think in general, we found that followers of autocratic leaders generally experience less fairness and perceive their leader to be less prototypical and in turn show more CWB.

(32)

32

References

Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz, Advances in experimental social psychology (pp. 267-299). New York: Academic Press.

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.

Anderson, C., & Brown, C. (2010). Functions And Disfunctions of Hierrchy. Research In Organizational Behavior, 30, 55-89.

Aquino, K., Lewis, M. U., & Bradfield, M. (1999). Justice constructs, negative affectivity, and employ a proposed model and empirical test. Journal of Organizational Behavior, 20, 1073-1091.

Bass, B. M., & Bass, R. (2008). The Bass Handbook Of Leadership: Theory, Research and Managerial Implecations. New York: Free Press.

Becker, T. E. (2005). Potential problems in the statistical control of variables in organizational research: A qualitative analysis with recommendations. Organzational Research Methods, 8(3), 274-289.

Clarke, K. A. (2005). The phantom menace: Omitted variable bias in econometric research. Conflict management and peace science, 22(4), 341-352.

Cohen-Charash, Y., & Spector, P. E. (2001). The role of justice in organizations: A meta-analysis. Organizational behavior and human decision processes, 86(2), 278-321.

Cole, M. S., Carter, M. Z., & Zang, Z. (2013). Leader–Team Congruence in Power Distance Values and Team Effectiveness: The Mediating Role of Procedural Justice Climate. Journal Of Applied Psychology, 98(6), 962-973.

Colquitt, J. A. (2001). On the dimentionality of organizational justice: A construct validation of measure. Journal of applied Psychology, 86(3), 386-400.

(33)

33 Colquitt, J. A., Noe, R. A., & Jackson, C. L. (2002). Justice in teams: Antecedents and consequences

of organizational citizenship justice. Personnel Psychology, 55, 83-109.

Colquitt, J. A., Rodell, J. B., Zapata, C. B., Scott, B. A., Long, D. M., & Conlon, D. E. (2013). Justice at the Millennium, a Decade Later: A Meta-Analytic Test of Social Exchange and Affect-Based Perspectives. Journal of Applied Psychology, 98(2), 199-236.

De Cremer, D. (2006). Affective and moivational consequences of leadre self-sacrifice: The moderating effect of autoocratic leadership. The Leadership Quartely, 17(1), 79-93.

De Hoogh, A. H., & Den Hartog, D. N. (2009). Neuroticism and locus of control as moderators of the relationsships of charasmatic and autocratic leadership with burnout. Journal of applied psychology, 94(4), 1058.

De Hoogh, A. H., Den Hartog, D. N., Koopman, P., Thierry, H., Van den Berg, P., Van der Weide, J., & Wilderom, C. (2004). Charismatic leadership, environmental dynamism, and performance. European journal of work and organizational psychology, 13(4), 447-471.

De Hoogh, A. H., Greer, L. L., & Den Hartog, D. N. (2013). Diabolical dictators or capable commanders? An investigation of the differential effects of autocratic leadership on team performance. The Leadership Quartely, 26(5), 687-701.

Demerouti, E., Bakker, A. B., & Bulters, A. J. (2004). The loss spiral of work pressure, work-home interference and exhaustion: Reciprocal relations in a three-wave study. Journal of vocational behavior, 64(1), 131-149.

Douglas, S. C., & Martinko, M. J. (2001). Exploring the role of individual differences in workplace aggression. Journal of Applied Psychology, 86(4), 547-559.

Einarsen, S., Aasland, M. S., & Skogstad, A. (2007). Destructive leadershipbehavior: A defenition and conceptual model. The Leadership Quartely, 18(3), 207-216.

Fox, S., & Spector, P. E. (1999). A model of work frustration and agression. Journal of Organizaional Behavior, 20, 915-931.

(34)

34 Fox, S., Spector, P. E., & Miles, D. (2001). 2001). Counterproductive work behavior (CWB) in

response to job stressors and organizational justice: some mediator and moderator tests for autonomy and emotions. Journal Of Vacational Behavior, 59, 291-309.

Giessner, S. R., & van Knippenberg, D. (2008). “License to fail”: Goal definition, leader group

prototypicality, and perceptions of leadership effectiveness after leader failure. Organizational Behavior and Human Decision Processes, 105(1), 14-35.

Giessner, S. R., Van Knippenberg, D., & Sleebos, E. (2009). Licence to fail? How leader group prototypicality moderates the effects of leader performance on perceptions of leadership effectiveness. The leadership quarterly, 20(3), 434-451.

Giessner, S. R., Van Knippenberg, D., Van Ginkel, W., & Sleebos, E. (2013). Team-oriented leadership: the interactive effects leader group prototypicality, accountability, and team identification. Journal of applied psychology, 98(4), 658.

Gruys, M. L., & Sackett, P. R. (2003). Investigating the dimentionality of counterproductive work behavior. International journal of selection and assessment, 11(1), 30-42.

Harrison, D. A., & Klein, K. J. (2007). What's the Difference? Diversity Constructs as Separation, Variety, or Disparity in Organizations. The Academy of Management Review, 32(4), 1199-1228.

Hofstede, G. H. (2001). Culture’s consequences: Comparing values, behaviors, institutions, and organizations across nations (2nd ed.). Thousand Oaks, CA, USA: Sage.

Hogg, M. A. (2001). A Social Identity Theory of Leadership. Personality and social psychology review, 5(3), 184-200.

Huo, Y. J., Smith, H. J., Tyler , T. R., & Lind, E. A. (1996). Superordinate identification, subgroup identification, and justice concerns: Is seperatism the problem; Is assimilation the answer? Psychological Science, 40-45.

(35)

35 Jago, A. G. (1982). Leadership: perspectives in theory and research. Management Science, 28(3),

315-336.

Kelloway, E. K., Francis, L., Prosser, M., & Cameron, J. E. (2010). Counterproductive work behavior as protest. Human resource management review, 20(1), 18-25.

Koivisto, S., & Lipponen, J. (2015). A leaders' procedural justice, respect and extra-role behavior: The roles of leader in-group prototypicality and identification. Social justice research, 28(2), 187-206.

Leventhal, G. S. (1980). What should be done with equity theory? In Social exchange (pp. 27-55). Springer US.

Lewin, K., Lippitt, R., & White, R. K. (1939). Patterns of agressive behavior in experimetally created "social climates". The journal of social psychology, 10(2), 269-299.

Lind, E. A. (2001). Fairness heuristic theory: Justice judgements as pivotal cognitions in organizational relations. Advances in organization justice, 56(8).

Lind, E. A., & Tyler, T. R. (1988). The social psychology of procedural justice. New York: Springer Science & Business Media.

Lippitt, R. (1940). An experimental study of the effect of democratic and authoritarian group atmospheres. University of Iowa Studies: Child Welfare, 16(3), 43-195.

Lipponen, J., Koivisto, S., & Olkkonen, M. E. (2005). Procedural Justice and status jugements: The moderating role of leader ingroup prototypicality. The Leadership Quartely, 16(4), 517-528.

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of mangement review, 20(3), 709-734.

Muller, E. N. (1985). Income inequality, regime repressiveness, and political violence. American sociological review, 50, 47-61.

Will employees suffer from autocratic leadership behavior or is this behavior sometimes justified? : a closer look at group-leader prototypicality