Enhancing Job Autonomy and Job
Satisfaction in the Public Sector:
The Role of Leadership
Master Thesis Executive Programme in Management Studies
Leadership and Management
June 30, 2016
Student name: Danny J. Loots Student number: 10730702
Statement of originality
This document is written by Danny Loots who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Preface
With great pleasure and pride, I present to you my thesis that I have written for my Master of Science in Management Studies at the University of Amsterdam. I started this Master study in February 2014, and after 2.5 years with learning, fun and sometimes tough moments, I am finishing now. As operational manager at Customs, I am dealing with the topics Leadership and Management on a daily basis and this study provided me with lots of new insights and encouraged me to highlight business issues from different perspectives. I have enjoyed working on this thesis and I am very proud of the end result.
I would like to thank everyone who helped and supported me in carrying out this study. I am especially grateful to my girlfriend Vera Senden for her patience, understanding, and above all, her helping hand in sometimes difficult moments. A special thanks to my supervisor, Claudia Buengeler, who helped me to get through this intensive thesis process and provided me with extensive feedback. Also, a word of thanks to my managers Robin Burgering and Els van der Wielen, who have enabled me to carry out this study within the Customs
organization.
Danny Loots
Abstract
This study aimed to examine the effects of task-oriented leadership, relation-oriented
leadership, change-oriented leadership, ethical leadership and passive leadership on
employees’ perceived degree of job autonomy, and whether and how the effect of employees’ perceived degree of job autonomy on job satisfaction is moderated by these leadership
behaviors. By using two surveys with a time lag of four weeks, data was collected from 924 officials working at Customs, a public enforcement organization that is part of the tax authority in the Netherlands. Results showed that ethical leadership, change-oriented
leadership and relation-oriented leadership are more effective leadership behaviors to improve employees’ sense of autonomy than task-oriented and passive leadership behaviors.
Furthermore, employees’ degree of autonomy is positively related to job satisfaction. However, no moderating role is found for management by exception-active, contingent reward, initiating structure, consideration leadership, transformational leadership and ethical leadership on the relationship between employees’ perceived degree of job autonomy and job satisfaction. Only passive leadership influences the relationship between job autonomy and job satisfaction at the .10 level of significance such that this relationship is stronger when passive leadership is high as compared to low. This study revealed that managers should be aware that their behavior matters when it comes to predicting employees’ sense of job autonomy and job satisfaction.
Key words: job satisfaction; job autonomy; task-oriented leadership; relation-oriented
Table of Content
STATEMENT OF ORIGINALITY ... 2 PREFACE ... 3 ABSTRACT ... 4 TABLE OF CONTENT ... 5 1 INTRODUCTION ... 72 THEORETICAL BACKGROUND AND HYPOTHESES ... 10
2.1LEADERSHIP AND JOB AUTONOMY ... 10
2.1.1 Task-oriented leadership and job autonomy ... 12
2.1.3 Relation-oriented leadership and job autonomy ... 13
2.1.4 Change-oriented leadership and job autonomy ... 15
2.1.5 Ethical leadership and job autonomy ... 17
2.1.6 Passive leadership and job autonomy ... 17
2.1.7 Relative effectiveness of leadership behaviors on job autonomy ... 19
2.2JOB AUTONOMY AND JOB SATISFACTION ... 20
2.3INFLUENCE OF LEADERSHIP ON LINK BETWEEN AUTONOMY AND SATISFACTION ... 22
3 METHOD ... 27
3.1QUANTITATIVE RESEARCH ... 27
3.2PROCEDURE AND SAMPLE ... 27
3.3MEASUREMENT OF VARIABLES ... 29
4 RESULTS ... 33
4.1TESTS OF VALIDITY AND RELIABILITY ... 33
4.2.1 Hypothesis 1 ... 35
4.2.2 Hypothesis 2 ... 40
4.2.3 Hypothesis 3 ... 41
4.2.4. Additional analyses ... 46
5 DISCUSSION ... 48
5.1.THEORETICAL AND PRACTICAL IMPLICATIONS ... 49
5.2.LIMITATIONS AND FUTURE RESEARCH DIRECTIONS ... 54
5.3.CONCLUSIONS ... 56
6 REFERENCES ... 57
7 APPENDICES ... 62
7.1.APPENDIX A.PITCH TO THE MANAGING DIRECTOR OF CUSTOMS ... 62
7.2.APPENDIX B.ANNOUNCEMENT ON INTERNAL WEBSITE DOUANENET ... 63
7.3.APPENDIX C.MEASUREMENT OF VARIABLES. ... 63
1 Introduction
A Staff Monitor (Effectory, 2015) and a Life Phase Staff Monitor (LPMO, 2015) has been conducted among employees within Customs, a public enforcement organization and part of the Tax authority in the Netherlands. The results of the Effectory are evaluated with other organizations in the Dutch public sector (benchmark) and show that Customs officials in general experience little autonomy in their work (5.7 on a scale of 10 with respect to 7.8 in the benchmark). Contentment with the role of the manager scored 6.1 compared to 7.5 in the benchmark. Regarding job satisfaction, Customs scored substantially lower (6.6) than the benchmark (7.8). The results of the LPMO show a similar picture and lead to concern of management with respect to the motivation and inspiration of Customs officials. By dealing with these negative points maybe the sense of job autonomy and job satisfaction can be improved (Effectory, 2015).
When employees have autonomy in their work, they may perceive a greater sense of personal responsibility for their work (Piccolo, Greenbaum, Den Hartog & Folger, 2010) and feel trusted in performing their job. Employees in autonomous jobs may experience more flexibility and meaningfulness in carrying out their activities and are therefore more satisfied (Fried & Ferris, 1987; cf. Piccolo & Colquitt, 2006). Job satisfaction in turn is positively related to outcomes such as job performance (Judge, Thoresen, Bono & Patton, 2001). So, employees’ job satisfaction and perceived job autonomy are important factors for
organizations to consider.
Leadership has been proposed as one of the most important factors contributing to
employee perceptions in the workplace (Bass, Avolio, Jung, & Berson, 2003). Following
Piccolo, Greenbaum, Den Hartog and Folger (2010) employees’ perceived degree of autonomy is partly based on the behavior of the leader. Leaders could be able to affect
followers’ sense of autonomy, because they have influence on both the objective and subjective nature of the job.
The literature has paid attention to the relationship among leadership behavior, job autonomy and job satisfaction (e.g. Fernandez, 2008; Hackman & Oldham, 1976; Judge & Piccolo, 2004; Piccolo & Colquitt, 2006; Piccolo, Greenbaum, Den Hartog & Folger, 2010). However, systematic studies examining the relative effectiveness of different leadership behaviors in shaping job autonomy are missing. Also, we lack knowledge on whether and how leadership behaviors influence the effectiveness of job autonomy in stimulating job
satisfaction. This is necessary to better understand whether this relationship that has been
established in prior research (e.g. Fried & Ferris, 1987; Piccolo & Colquitt, 2006) varies depending upon levels of different leadership behaviors. Piccolo, Bono, Heinitz, Rowold, Duehr and Judge (2012) suggest examining differences in organizational outcomes such as job satisfaction and leader effectiveness based on the level of the leader.
Most previous studies have focused on one model of leadership behavior only or dealt with one dimension of such a model, without carefully evaluating the relative effects of various leadership models simultaneously (Hunter et al., 2007; cf. Piccolo, Bono, Heinitz, Rowold, Duehr & Judge, 2012). When researchers compare and assess multiple leadership behavior dimensions they have often focused on a particular theory, instead of making comparisons between different models (Derue, Nahrgang, Wellman & Humphrey, 2011). Therefore, this study examines the relative effects of task-oriented leadership, relational-oriented leadership, change-oriented leadership and passive leadership following Derue et al. (2011). Hence, to strengthen my claim of investigating the effect of different types of leadership behaviors, I also add ethical leadership (Brown, Trevino & Harrison, 2005), a more recent theory that has been the subject of several studies (e.g. Detert, Trevino, Burris & Andiappan, 2007; cf. Piccolo, Greenbaum, Den Hartog & Folger, 2010).
Fernandez (2008) calls for empirical research on leadership styles in public organizations, because most leadership theories have not been examined in a public environment. Trottier, van Wart and Wang (2008) also mentioned the need for better theorizing on and empirically testing of leadership effects in public settings, because only a fraction of research on
leadership has been conducted in the public sector. Furthermore, Van Wart (2003) argued that there is a need for a comprehensive leadership model in the public leadership literature that integrates more leadership elements.
To increase understanding regarding how different leadership behaviors influence employees’ perceived degree of job autonomy and satisfaction, I will examine whether and how leadership behaviors are associated with employees’ perceived degree of autonomy at work and how this in turn affects job satisfaction of employees. I focus on the Dutch Customs organization to expand knowledge on the effect and influence of different leadership
behaviors in a public setting. In addition, I propose that certain leadership behaviors will reinforce the effect of the degree of job autonomy on employees’ job satisfaction. My research question is as follows: “Which leadership behavior should managers in a public
organization apply to enhance employees’ perceived degree of job autonomy, and to strengthen the link between job autonomy and job satisfaction?”
The result of this study should give a better insight into the leadership behaviors experienced by and effective for public state officers. Moreover, the results provide better understanding of the effects of leadership behavior on employees’ perceived degree of job autonomy and job satisfaction. It may contribute to determine which leadership behavior can best be applied in a public organization where employees typically experience relatively little autonomy and are oftentimes low on job satisfaction (Effectory, 2015; LPMO, 2015). This can help pinpoint better strategies for recruiting, promotion and training of managers within public organizations.
2 Theoretical background and hypotheses
In the following section, I provide a brief summary of relevant literature on leadership behaviors, job autonomy and job satisfaction, and introduce several hypotheses regarding the relationships among leadership behaviors, job autonomy and job satisfaction.
2.1 Leadership and job autonomy
Job autonomy is - like variety, identity, significance and feedback - a job characteristic that is presented by Hackman and Oldham (1976) and describes terms in which employees would be intrinsically motivated when performing a job (Piccolo & Colquitt, 2006). Hackman and Oldham define job autonomy as the degree to which an employee has “the freedom,
independence, or discretion to schedule work, make decisions, or select the methods used to perform work tasks” (cf. Piccolo, Greenbaum, Den Hartog & Folger, 2010, p. 262).
According to Piccolo, Greenbaum, Den Hartog & Folger (2010), autonomy can be seen from two complementary perspectives; an objective and subjective point of view. By changing the demands and responsibilities related to a specific job, the objective feature of a job can be influenced. From the subjective point of view, autonomy can be manipulated through environmental cues such as social learning and social influence.
Leadership is important because it helps employees to cope with change by creating a direction and developing a vision for the future, aligning employees through communicating this direction, and motivating and inspiring employees, and keeps moving them in the right direction (Kotter, 1990). Leadership is the process of influencing employees to achieve organizational goals (Trottier, van Wart & Wang (2008).
Researchers use a range of conceptualizations of leadership behaviors and there is a lack of conformity to which conceptualizations and behaviors are appropriate and useful for leaders (Yukl, Gordon & Taber, 2002). The two most studied theories of leadership behavior are the
Ohio State model, which consists of initiating structure and consideration (e.g. Fleishman, 1953; Stogdill, 1963; cf. Derue, Nahrgang, Wellman & Humphrey, 2011) and the model of Bass, which consists of transactional leadership and transformational leadership (Bass, 1985; cf. Trottier, van Wart & Wang, 2008). Derue et al. (2011) have clustered the behaviors of these two models into a framework with four categories: (1) task-oriented behaviors, (2) relational-oriented behaviors, (3) change-oriented behaviors, and (4) passive behaviors. Ekvall and Arvonen (1991) also suggested grouping leadership behaviors in task-oriented, relation-oriented, and change-oriented behaviors (cf. Fernandez, 2008). Likewise, Yukl, et al. (2002) claimed leadership behaviors could be clustered in task, relation and change behaviors. Task-oriented leaders use the resources and personnel of the organization efficiently with a high reliability of operations, products and services (Yukl et al., 2002). Relation-oriented leaders are focused on holding strong social relationships with their employees; realize mutual respect and trust, and try to satisfy their needs (Burke, Stagl, Klein, Goodwin, Salas & Halpin, 2006). Such leaders try to achieve a strong commitment to the organization and the mission with a high level of collaboration among employees (Yukl et al., 2002). Change-oriented leaders coach and teach employees (Piccolo & Colquitt, 2006), develop and communicate a vision for change, and encourage innovative thinking (Derue et al., 2011). Furthermore,
several studies have found ‘passive’ leadership as a separate category (Yukl, 1999), whereby
leaders wait until the employee has created problems before they take action or take no action at all (Judge & Piccolo, 2004). In this study, I also focus on ethical leadership because ethical leaders might model the job experience of employees, such as the perceived degree of job autonomy, by sharing power and involving employees in the decision-making process (Piccolo, Greenbaum, Den Hartog & Folger, 2010).
Following Piccolo, Greenbaum, Den Hartog and Folger (2010) leaders have influence on the objective and socially constructed character of the job. Leaders do not only influence
whether employees have some autonomy, but they also influence whether employees perceive to have autonomy. By making changes to objective elements of a job, they might affect employees’ perceptions of autonomy. Employees ascribe the significance and personal responsibility of their jobs partly on behaviors of their leaders (Piccolo et al., 2010). Thus, through designing work and behave in such a way that employees identify themselves with the task, find it meaningful and experience autonomy (Burke, Stagl, Klein, Goodwin, Salas & Halpin, 2006), leaders may have a considerably impact on employees’ perception of job autonomy.
2.1.1 Task-oriented leadership and job autonomy
Here, task-oriented leadership behavior is conceptualized by three behaviors, namely
management by exception-active, contingent reward, and initiating structure. A leader, which is characterized by management by exception-active, observes continually the performance of employees and engage with corrective actions when deviations take place (Trottier, van Wart & Wang, 2008). Following Howell and Avolio (1993), a leader that is defined by an active form of management by exception continually checks employees’ performance and maintains control so he or she can anticipate on mistakes before they become a serious problem. They undertake corrective actions when required. Contingent reward is behavior where a leader makes clear which work should be done, how the work should be performed and which rewards will be provided when the services and tasks are rendered (Trottier, van Wart & Wang, 2008). According to Howell and Avolio (1993), rewards may involve recognition from a leader for realized work and bonuses. Contingent reward is viewed as an active and positive exchange between leaders and employees. Initiating structure is behavior whereby the leader explains his or her role, defines the role and function of the employee, coordinates the actions of employees, determines the standards to which employees must meet the task execution,
and ensures that employees meet these standards (Derue, Nahrgang, Wellman & Humphrey, 2011).
Leaders showing task-oriented behaviors often make decisions without consulting their staff, and provide clear expectations. For example, leaders using active management by exception check employees’ performance and undertake corrective actions when problems take place or standards are not realized. Contingent reward behavior clarifies what needs to be done and initiating structure determines the standards and coordinates employees’ actions. Employees may even experience task-oriented leaders as “detached, autocratic and punitive” (Fernandez, 2008, p. 181). Task-oriented leaders do not make employees feel trusted and capable of making own decisions. Also, employees will feel less responsible for their tasks and experience less flexibility in performing their activities. Thus, I posit that employees, who experience task-oriented behavior of their leader, actually do not feel much space to make their own decisions about work, do not experience many opportunities to provide input, and therefore feel less sense of control and autonomy in their job. Therefore, the first hypothesis is as follows:
Hypothesis 1a: Task-oriented leadership is negatively associated with employees’ perceived degree of job autonomy.
2.1.3 Relation-oriented leadership and job autonomy
A behavior reflective of relation-oriented leadership is consideration leadership (Derue, Nahrgang, Wellman & Humphrey, 2011). Consideration leadership (Fleishman, 1953) is behavior where a leader focuses on co-operation and solidarity with employees. Leaders employing consideration leadership are friendly and approachable who treat employees with care and respect. Such leaders emphasize open communication, are open to the contribution of employees, and treat everyone as equals (Derue et al., 2011). Leaders showing
consideration leadership make employees feel comfortable and confident in their own capacities and talents (Piccolo, Bono, Heinitz, Rowold, Duehr & Judge, 2012).
According to Yukl, Gordon and Taber (2002), relation-oriented leadership involve supporting, recognizing, developing, and empowering behaviors. Empowering leadership, participative leadership and democratic leadership are similar relation-oriented behaviors (Conger, 1989; Gastil, 1994; Kahai, Sosik, & Avolio, 1997; cf. Derue, Nahrgang, Wellman & Humphrey, 2011). Derue et al. (2011) argue that there is a common theme between these leadership behaviors and consideration leadership in such a way that the leader builds employees’ respect and focuses on welfare of the team. Because of these similarities, I focus in this study on consideration leadership as relation-oriented leadership behavior.
Cooper and Cartwright (1997) have found in their study that leaders have an important function in monitoring the amount of ownership individuals have over their role and that increased delegation and participation were related with increased levels of well-being (cf. Nielsen, Randall, Yarker & Brenner, 2008). Relation-oriented leaders are focused on realizing relationships, shaping mutual trust and satisfying employees’ needs. Such leaders try to make followers feel comfortable and confident in their own abilities. A meta-analysis from Judge, Piccolo, and Ilies (2004) showed that consideration leadership was positively related to employees’ job satisfaction, motivation and leader effectiveness. I argue that employees who have relation-oriented leaders do feel space to give input; feel trusted by their leader and therefore, will make more own decisions about work or tasks. This may result in a higher sense of control and autonomy in their job. The second hypothesis is:
Hypothesis 1b: Relation-oriented leadership is positively associated with employees’ perceived degree of job autonomy.
2.1.4 Change-oriented leadership and job autonomy
Following Derue, Nahrgang, Wellman and Humphrey (2011), change-oriented leadership includes, among other aspects, the dimensions of transformational leadership. These are individualized consideration, idealized influence, intellectual stimulation, and inspirational motivation (Bass & Avolio, 1995). According to Bass and Avolio (1995), leaders showing individualized consideration focus on the needs of employees, perform as a mentor or coach and recognize the concerns of employees. A leader employing idealized influence, behaves in a praiseworthy way, act as a role model and has the effect that employees try to identify with their leader. Intellectual stimulation is leader behavior that encourage innovative thinking of employees by actively seeking for their ideas and stimulate questioning the status quo.
Leaders showing inspirational motivation encourage and inspire employees with an appealing vision. Such a leader stimulates employees with “high standards, communicate optimism about future goal attainment, and provide meaning for the task at hand” (cf. Judge & Piccolo, 2004, p. 755).
According to Yukl, Gordon and Taber (2002), change-oriented leadership include behaviors such as monitoring the external environment, articulating an inspiring vision, encouraging innovative thinking and taking personal risks to implement change. Charismatic leadership (House, 1977) is another form of change-oriented leadership. However, there is a strong overlap between transformational leadership and charismatic leadership (Judge & Piccolo, 2004; Van Knippenberg & Sitkin, 2013). Because of this overlap, I focus in this study on the dimensions of transformational leadership as change-oriented leadership behavior. Some aspects of change-oriented leadership behaviors may also have an overlap with relation-oriented behaviors. An example is individualized consideration whereby the leader listens to employees’ concerns. However, transformational leadership is
organizations” (Derue, Nahrgang, Wellman & Humphrey, 2011, p.16) and is therefore conceptually distinct from relation-oriented leadership.
Transformational leaders provide more meaning and challenge to employees. They
stimulate employees to be more innovative and creative when dealing with complex situations (Bass, 1985; cf. Piccolo & Colquitt, 2006). Such behaviors may signal confidence, trust and belongingness among employees. Change-oriented leaders encourage innovative thinking and invite employees to come up with different perspectives and ideas (Derue, Nahrgang,
Wellman & Humphrey, 2011). Employees with change-oriented leaders are likely to feel more autonomy because their leader is supportive of their personal needs and the idea of innovative ways to perform their tasks. According to Piccolo and Colquitt (2006), leaders who use intellectual stimulation may enhance employees’ perceptions of autonomy, and leaders who use individualized consideration by coaching should have employees’ that experience more autonomy. These authors found empirical support that transformational leadership was positively related to each of the perceptions of core job characteristics including autonomy. Also, Purvanova, Bono and Dzieweczynski (2009) have showed a positive correlation between transformational leadership and the core characteristics as well as autonomy. I therefore argue that followers, who experience change-oriented leaders, do feel to be creative in their job, so, they can make own decisions about work or tasks and feel a high sense of control and autonomy in their job. This results in de next hypothesis:
Hypothesis 1c: Change-oriented leadership is positively associated with employees’ perceived degree of job autonomy.
2.1.5 Ethical leadership and job autonomy
De Hoogh and Den Hartog (2008) state that an ethical leader involves employees in making decisions and listens to their ideas and suggestions. Brown, Trevino, and Harrison (2005) define ethical leadership as: “the demonstration of normatively appropriate conduct through personal actions and interpersonal relationships, and the promotion of such conduct to followers through two-way communication, reinforcement, and decision-making” (p. 120). To behave in a normatively appropriate way means that leaders should behave uniformly with common expectations concerning how leaders should behave.
Ethical leaders give employees the possibility to express themselves and provide them with high levels of autonomy (Brown, Trevino & Harrison, 2005). Karlshoven, Den Hartog and De Hoogh (2008) found a high correlation (r = .74) between power sharing and overall ethical leadership (cf. Piccolo, Greenbaum, Den Hartog & Foler, 2010), which implies a positive relation between ethical leadership and employee’s perceived degree of job autonomy. Furthermore, Piccolo et al. (2010) found a significant positively relationship between ethical leadership and autonomy. Based on this argument, I posit when a leader uses ethical
leadership behavior their employees will experience more trust and space to make decisions by their own about their work and their methods to complete tasks. This results in de
following hypothesis:
Hypothesis 1d: Ethical leadership is positively associated with employees’ perceived degree of job autonomy.
2.1.6 Passive leadership and job autonomy
Kelloway, Mullen, and Francis (2006)define passive leadership as consisting of passive
exception-passive only intervenes when task-related problems or challenges occur, and employees do not realize the standards. A leader with passive management by exception behavior intervenes with criticism and reprimand only after the employee makes the mistake. He or she waits until the task is completed before ascertaining that a problem has occurred and then brings the problem to the consciousness of the employee. Leaders emphasized by laissez-faire leadership behavior “avoid making decisions, hesitate in taking action, and are absent when needed” (Judge & Piccolo, 2004, p. 756). Some authors have called this type of leadership non-leadership (Trottier, van Wart & Wang, 2008) or destructive leadership (Schyns & Schilling, 2013; Skogstad, 2007). According to Bass, Avolio, Jung and Berson (2003), passive leaders “avoid specifying agreements, clarifying expectations, and providing goals and standards to be achieved by employees” (p. 208).
Laissez-faire leaders ignore expectations from employees by being absent when they are needed and avoid making decisions, and do not give any feedback. Avolio (1999) mentioned that laissez-faire leadership is highly dissatisfying (cf. Judge & Piccolo, 2004). An empirical study from Schyns and Schilling (2013) have shown that destructive leadership behaviors are negatively associated with followers’ job satisfaction, organizational commitment, trust and well-being. However, when leaders are passive, employees may experience some freedom, and should make own decisions about their work and methods to accomplish their tasks. Therefore, I expect that passive leadership will be positively associated with employees’
perceived degree of autonomy. This results in the following hypothesis:
Hypothesis 1e: Passive leadership is positively associated with employees’ perceived degree of job autonomy.
2.1.7 Relative effectiveness of leadership behaviors on job autonomy
To study which role leadership has in relation to job autonomy, I examine the relative effectiveness of task-oriented leadership, relation-oriented leadership, change-oriented leadership, ethical leadership and passive leadership on employees’ perceived degree of job autonomy.
As mentioned in the sections above, I propose that employees who have relation-oriented leaders do feel space to give input and feel trusted by their leader. Therefore, employees will make more own decisions about work or tasks, which may result ultimately in a higher sense of control and autonomy in their job. Likewise, change-oriented leaders may enhance
employees’ degree of autonomy, because employees experience to be more creative and innovative in their job. They can make own decisions about work or tasks and feel a high sense of control. Also, when leaders use ethical leadership behavior, employees will experience more trust and space to make decisions by their own about their work and their methods to complete tasks. In contrast, employees with task-oriented leaders should feel less trusted and capable of making own decisions. They may feel less responsible for their tasks and experience less flexibility in performing their activities. As a result, they experience less sense of control and autonomy in their job. When leaders behave passive, employees may feel some freedom, and should make decisions about their work and methods to accomplish their tasks. However, they are not encouraged or challenged by their leader to be creative or innovative, or feel trusted in making the right decisions by their own. I therefore expect that passive leadership will be positively associated with employees’ perceived degree of
autonomy, but this relationship may not as strong compared to relation-oriented leadership, change-oriented leadership and ethical leadership. This results in the following hypothesis:
Hypothesis 1f: Relation-oriented leadership, change-oriented leadership, and ethical leadership will be stronger related to employees’ degree of job autonomy than passive leadership and task-oriented leadership.
2.2 Job autonomy and job satisfaction
An important claim of the job characteristics model (Hackman and Oldham, 1976) is that enhanced job characteristics, and job autonomy is one of these characteristics, is, among other things, related with higher levels of job satisfaction (Hackman & Oldham, 1976; cf. Piccolo, Greenbaum, Den Hartog, & Folger, 2010). Job satisfaction can be defined as a feeling that people have about their jobs and the organization in which they work (Rad &
Yarmohammadian, 2006). Kalleberg (1977) defines job satisfaction as an employees’ overall emotional position toward work roles, and their feelings about elements such as pay, benefits, career opportunities, job tasks, work conditions, relationship with fellow workers, and their leader.
Autonomy gives employees the feeling of space and freedom to make better use of their knowledge, experience and skills. This in turn has a motivating effect on employees to develop new knowledge and skills (Parker, Wall, & Jackson, 1997; cf. Piccolo et al., 2010) and when employees feel they are improving and getting better in their job over time, they also feel more satisfied (Hackman & Oldham, 1980; cf. Derue, Nahrgang, Wellman, & Humphrey, 2011). Designing jobs in such a way that employees experience a sense of autonomy helps providing a structure that empowers team effectiveness by increasing employees’ motivation (Hackman & Oldham, 1980; cf. Burke, Stagl, Klein, Goodwin, Salas & Halpin, 2006). A sense of autonomy motivates employees intrinsically, because it fulfills an employees’ desire for a sense of control, and responsibility (Fuller, Marler & Hester, 2006;
cf. Piccolo, Greenbaum, Den Hartog, & Folger, 2010). These authors state that job autonomy admits employees to expand the range of responsibility and enlarge the view of their own work roles. An increase in autonomy will furthermore permit employees to be more flexible in how they describe their role, because they will have greater freedom in deciding how to perform their task (Fried, Hollenbeck, Slowik, Tiegs & Ben-David, 1999; cf. Morgeson, Delaney-Klinger & Hemingway, 2005).
Parker (1998) found that more autonomy results in an increase in the sense of ownership of problems. Furthermore, employees will identify a wider range of skills and knowledge as important for their roles. More control over the work environment motivates employees to try out and learn new tasks, which implies when given more autonomy to employees, they are likely to combine more tasks into their main role. Employees who have the opportunity and capability to do more will do more in their jobs (cf. Morgeson, Delaney-Klinger &
Hemingway, 2005). In sum, it could be argued that job autonomy enhances trust by staff in performing their job and employees feel more personally responsible for their work. Employees in autonomous jobs experience more flexibility in performing their tasks.
Furthermore, employees feel more sense of meaningfulness in their job and more ownership for problems at work when job autonomy is high. Concluding, it is thus likely that job autonomy results in higher job satisfaction.
Consistent with the prior, previous empirical studies have found a significantly positive relationship between job autonomy and job satisfaction. For example, research by Fried and Ferris (1987) has supported the belief when employees perceive a high degree of job
autonomy; they are more satisfied and intrinsically motivated with their jobs. A meta-analysis done by Loher, Noe, Moeller and Fitzgerald (1985) found a positive correlation between each of the job characteristics (r = .46 for autonomy) and job satisfaction. Thus, I assert that
employees’ perceived degree of job autonomy is positively related with employees’ job satisfaction, which results in the following hypothesis:
Hypothesis 2: Employees’ perceived degree of job autonomy is positively related with job satisfaction.
2.3 Influence of leadership on link between autonomy and satisfaction
In this study, I examine whether and how the positive relationship between employees’ perceived degree of job autonomy and job satisfaction is moderated by leadership behavior. Howell, Dorfman and Kerr (1986) define a moderator as a variable that “affects the nature of the relationship between two other variables, without necessarily being correlated with either
of them” (p. 89). The literature is replete with studies reporting the direct effects of leadership
on outcomes such as follower motivation, satisfaction, and performance. However, how does leadership affect the relationship between job autonomy and job satisfaction? This
relationship may vary depending upon levels of leadership, because certain leadership behaviors may provide a context that maximizes the positive effects of job autonomy on job satisfaction and other leadership behaviors may minimize these positive effects. Thus, this study addresses how leaders may contribute to the perceived degree of job autonomy by employees and considers how their leadership behavior strengthens or weakens the impact of job autonomy on job satisfaction.
Change-oriented leaders do not only influence employee’s attitude and behaviors but also inspire followers to perform beyond previous standards (Piccolo, Bono, Heinitz, Rowold, Duehr & Judge, 2012). Shamir, House and Arthur (1993) argue that leaders who appeal to ideological values and use intellectual stimulation add meaningfulness into their employees’ work (cf. Piccolo & Colquitt, 2006). It is therefore reasonable to expect that employees who
perceive a high degree of job autonomy will be more satisfied when they have a change-oriented leader who encourages employees to come up with innovative and creative ideas, provide support when this is needed, and inspires them to perform beyond standards. Such leaders create a setting whereby employees can positively influence their work, experience meaningfulness in their job and provide space to make own decisions. Employees should than be more satisfied as compared with a situation where this setting is missing. Therefore, I posit that change-oriented leaders strengthen the positive relationship between job autonomy and job satisfaction such that this relationship is stronger when change-oriented leadership is high as compared to low.
Relation-oriented leaders are focused on realizing and shaping relationships, mutual trust and satisfying employees’ needs. Such leaders try to make followers feel comfortable and confident in their own abilities. In such a way, considerate leaders provide more trust among employees in performing their work and therefore it is reasonable to argue that employees who perceive a high degree of job autonomy will be more satisfied when they have a leader who displays trust and confidence in own competences and talents. Employees are more satisfied because their leader challenges them to be autonomous and creates a setting where they feel confident in their own abilities. Employees are than more likely to show initiatives and own decisions in their work. Employees should be more satisfied as compared with a setting where a leader does not show confidence in employees’ knowledge, skills and talents and displays trust. I therefore argue that relation-oriented leaders strengthen the positive relationship between job autonomy and job satisfaction such that this relationship is stronger when relation-oriented leadership is high as compared to low.
Ethical leaders are trustworthy and caring, and offer employees autonomy in their work. Such leaders provide employees with opportunity to participate in decision-making and in assignments they feel are meaningful (Piccolo, Greenbaum, Den Hartog, & Folger, 2010).
Through clarifying how performing the tasks contribute to the achievement of goals, and make noticeable what the purpose and morality of the work is, this will probably foster the significance of employees’ tasks. Ethical leaders provide meaning to the work of employees and is it reasonable to expect that employees who perceive a high degree of job autonomy will be more satisfied when they have an ethical leader who displays fair behaviors and intensifies feelings of moral devotions. Such leaders create an environment for employees where they can participate in the decision making process and challenge employees to exert influence on their work. This translates into more satisfaction as compared with a situation where such an environment is missing. Therefore, I argue that ethical leaders strengthen the positive
relationship between job autonomy and job satisfaction such that this relationship is stronger when ethical leadership is high as compared to low. Therefore, the following hypothesis is stated:
Hypothesis 3a: Change-oriented leadership, relation-oriented leadership and ethical leadership moderate the relationship between job autonomy and job satisfaction such that this relationship is stronger when change-oriented leadership, relation-oriented leadership, and ethical leadership are high as compared to low.
Task-oriented leaders make decisions without seeking advice from employees, give clear targets and standards and continually check employee’s performance so that he or she can anticipate mistakes (Derue, Nahrgang, Wellman & Humphrey, 2011; Howell & Avolio, 1993). It is therefore reasonable to expect that employees who perceive a high degree of job autonomy will be less satisfied when they have a task-oriented leader who displays behavior that is focused on realizing targets and continually observes their work as compared with a
setting where a leader does not show such controlling and coordinating behavior. Such leaders create an environment that is not motivating for employees to provide ownership and were employees experience less confidence to come up with own ideas. A leader restricts
employees in their autonomy, because the work is tightly coordinated and decisions are made without input from employees. Thus, I argue that task-oriented leadership weakens the positive relationship between job autonomy and job satisfaction such that this relationship is weaker when task-oriented leadership is high as compared to low.
Kelloway, Sivanathan, Francis and Barling (2005) noted that passive leadership might cause employees to experience workplace stressors such as role conflict, role ambiguity, and the perceptions of low-quality interpersonal treatment by their leader (cf. Skogstad, Einarsen, Torsheim, Aasland & Hetland, 2007). So, passive leaders do not contribute to the clarity of work and do not help when people experience problems during the execution of their duties. In addition, employees experience no connection with their supervisor. Such a setting is not beneficial to employees’ job satisfaction as compared to a setting where a leader provides clarity and a helping hand when employees are struggling in their work. It is therefore reasonable that when employees are led by passive leaders, their perceived autonomy makes them less strongly satisfied than when their leader is low in passive leadership. I argue that passive leaders slack the positive relationship between job autonomy and job satisfaction such that this relationship is weaker when passive leadership is high as compared to low. This results in the following hypothesis:
Hypothesis 3b: Task-oriented leadership and passive leadership moderate the
relationship between job autonomy and job satisfaction such that this relationship is weaker when task-oriented leadership and passive leadership are high as compared to low.
The hypotheses established in this thesis are summarized in Figure 1. Hypotheses 1 a, b, c, d, e and f are depicted as Model 1a and refer to the main-effect relationships between the different leadership behaviors and job autonomy. Hypothesis 2 and Hypotheses 3a and b represent the second model of this study (Figure 1, Model 1b) and refer to the main-effect relationship between job autonomy and job satisfaction, and the moderating influence of the leadership behaviors on the relationship between job autonomy and job satisfaction.
3 Method
In the following section, I provide a complete description of the way the data is obtained, the operationalization of the constructs that are used, and the types of analyses that are conducted.
3.1 Quantitative research
To examine the hypotheses and finally answer the research question in this paper, I will apply a quantitative research. Quantitative research is usually related with a deductive approach that is used when a clear theoretical position is adopted and will be tested using hypothesis testing on the basis of collected data. The focus in this paper is on using data to test theory, so a deductive approach is used (Saunders, Lewis & Thornhill, 2012). With a quantitative research method, the relationships between variables are examined. These relationships are measured numerically and by using a range of statistical techniques these relationships are analyzed. The data is collected with surveys with a small time lag between the measurement points, which is a popular and often used strategy in business and management research (Saunders et al., 2012).
3.2 Procedure and sample
This study is conducted within Customs, a public organization that is part of the tax authority in the Netherlands. Customs employs 4,784 staff working in various functions and job levels. The officials are generally performing highly standardized tasks such as physical checks, administrative work, declaration processing, customer relationship management, business management, and management. Customs consists of nine regional offices across the country, and a main office in Rotterdam. Each official has a formal leader with managerial
officials.
The managing director of Customs has given permission to conduct this study within the Customs organization. First, on the internal website of Customs (Douanenet) an
announcement of the study is posted with a brief introduction, the purpose of the study and invitation to complete the surveys. Then, all 4,784 officials were invited through a business email with a covering letter stressing confidentiality and voluntary participation, as well as the importance of the study for all parties involved, and an individual link to a digital
questionnaire. In the questionnaire the officials were asked to rate their immediate manager on a variety of leadership behaviors and they were asked questions about their task
complexity, task interdependency, and perceived work characteristics like variety, identity, feedback, and autonomy. The officials had three weeks to fill in the questionnaire. After two weeks a reminder was sent by email to the officials who had not yet participated or completed the questionnaire using the Qualtrics participant administration tool. 1131 officials completed the questionnaire. The response rate was 23.6%. 331 respondents were female (29.3%) and the average age was 50.35 years (SD = 10.39).
Four weeks later, a very short second questionnaire was sent to the 1131 officials who participated in the first questionnaire. The participants were asked questions about their job satisfaction, absenteeism, productivity, and innovative behavior. As announced in the
covering letter, among participants who completed both questionnaires 10 bol.com gift cards of 25 euros were raffled. To make this claim they could enter their email address in the survey, which was the only identifier included with their data. After one week and again after two weeks a reminder was sent by email to the officials who had not yet participated or completed the second questionnaire. 924 officials completed the second survey. The response rate was 81.7%. 268 were female (29.0%) and the average age was 50.86 years (SD = 9.25). To show that those who dropped out at the second measuring time point did not
systematically differ from those completing both parts of the survey, the demographic differences (gender, age) between group 1, the respondents who completed the first and second survey, and group 2, the respondents who only completed the first survey were
checked. An Independent Sample t-test was conducted to compare the age for the two groups. There was a significant difference in age for group 1 (M = 50.86, SD = 9.25) compared to group 2 (M = 48.11), SD = 10.39; t (1104) = 3.74, p = .000, two-tailed). A Chi-Square Test was conducted to exclude the possibility of gender differences in the two groups. There was no significant difference in gender (p = .674, two-tailed). The covering letter and
announcement are provided in the appendix.
3.3 Measurement of variables
All items used in the questionnaires were derived from articles published in English-language journals. Since all respondents to the surveys had Dutch as their first language, the original items were translated into Dutch. In order to assure the content of the items remained unchanged, the translated Dutch items were back translated into English by a third person with a very high command of both languages and content knowledge following the procedure outlined by Brislin (1970). A small number of discrepancies between the back translated and the original items were corrected in the final Dutch version of the questionnaires. Except for the variables job satisfaction and leader-member exchange (LMX), all variables were measured in survey 1. The questionnaires are provided in the appendix.
Task-oriented leadership
Task-oriented leadership consists of initiating structure, contingent reward and management by exception-active. Initiating structure is measured using ten items from the Leader Behavior Description Questionnaire (LBDQ-form XII; Stogdill, 1963) and is also used by Piccolo,
Bono, Heinitz, Rowold, Duehr and Judge (2012). Responses were evaluated on a scale by 1 (strongly disagree) to 7 (strongly agree) and sample items are: “My leader schedules the work to be done”, “My leader assigns group members to particular tasks”, and “My leader decides what shall be done and how it shall be done”. The coefficient alpha reliability for this scale was .91.
Contingent reward and management by exception-active are measured by using the eight items of the Multifactor Leadership Questionnaire described by Bass and Avolio (1989), four items for contingent reward and four items for management by exception-active. Sample items for contingent reward are: “My leader let me know what I get when my work meets the requirements”, and “My leader tells me what to do to be rewarded for my efforts. The
coefficient alpha reliability for contingent reward was .94. A sample item for management by exception-active is: “My leader directs attention toward failures to meet standards”. The coefficient alpha reliability for this scale was .84.
Relation-oriented leadership
Relation-oriented leadership was conceptualized as consideration in this study. Consideration leadership is measured using ten items from the Leader Behavior Description Questionnaire (LBDQ-form XII; Stogdill, 1963), which is used by Piccolo, Bono, Heinitz, Rowold, Duehr and Judge (2012). Responses were evaluated on a scale by 1 (strongly disagree) to 7 (strongly agree) and sample items are: “My leader does little things to make it pleasant to be a member of the group”, and “My leader puts suggestions made by the group into operation”. The coefficient alpha reliability for this scale was .92.
Change-oriented leadership
Change-oriented leadership was measured with eleven items from the CLIO (Charismatic Leadership in Organizations) as described by De Hoogh, Den Hartog and Koopman (2004). Responses were evaluated on a scale by 1 (strongly disagree) to 7 (strongly agree) and sample items are: “My leader has a clear vision and an image of the future” and “My leader
encourages subordinates to develop their full potential”. The coefficient alpha reliability for this scale was .96.
Ethical leadership
Ethical leadership is measured by using ten items based on Brown, Trevino and Harrison (2005). Responses were evaluated on a scale by 1 to 7 (strongly disagree - strongly agree) and a sample item is: “My leader disciplines employees who violate ethical standards”. The coefficient alpha reliability for this scale was .93.
Passive leadership
Passive leadership is measured by using six items that are used in the CLIO (Charismatic Leadership in Organizations) as described by De Hoogh, Den Hartog and Koopman (2004). Responses were evaluated on a scale by 1 to 7 (strongly disagree - strongly agree) and a sample item is: “My leader only acts if things go wrong”. The coefficient alpha reliability for this scale was .87.
Job autonomy
Job autonomy is measured based on two items from the revised form of the Job Diagnostic Survey (Idaszak & Drasgow, 1987), which are also used by Piccolo and Colquitt (2006). A response scale is used in which 1 was “strongly disagree” and 7 was “strongly agree”. A
sample item is: “The job gives me considerable opportunity for independence and freedom in how I do the work”. However, two items for measuring job autonomy is somewhat limited and therefore I have added two more items adapted from Hackman and Oldham (1980), which were also used by Morgeson, Delaney-Klinger and Hemingway (2005). These items are rated on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree) and a sample item is: “I can decide on my own how to go about doing my work”. The coefficient alpha reliability for this scale was .96.
Job satisfaction
Job satisfaction is measured based on five items from the Brayfield-Rothe measure of job satisfaction (Brayfield & Rothe, 1951), which are also used by Piccolo, Bono, Heinitz, Rowold, Duehr and Judge (2012). Responses were evaluated on a scale of 1 to 7 (strongly disagree - strongly agree). Sample items are: “I feel fairly satisfied with my present job”, and “I find real enjoyment in my work”. The coefficient alpha reliability for this scale was .85.
Control variables
Demographic characteristics were included in the last section of the first questionnaire as control variables. These demographic variables might have a confounding effect on the relationship between leadership, job autonomy and job satisfaction. These demographic characteristics and organizational variables include gender, age, job function, educational level, and job and team tenure.
Additionally, consistent with prior research (e.g., Fernandez, 2008; Piccolo, Greenbaum, Den Hartog, & Folger, 2010) the variables task complexity, job characteristics (variety, significance, identity, feedback), and leader-member exchange (LMX) were included as controls to account for the potential effects of these variables on job autonomy and job
satisfaction. I controlled for task complexity since prior research of leadership (e.g. Barrow, 1976; Kerr, Schriesheim, Murphy & Stogdill, 1974) has shown that complexity of tasks could affect employees’ perceived degree of job autonomy and job satisfaction. Furthermore, Piccolo, Greenbaum, Den Hartog and Folger (2010) suggest the link between leadership and job behaviors may also depend on the nature of the relationship between a leader and the employee. The four job characteristics from the Job Characteristics Model were measured using eight items from Hackman and Oldham (1974). Leader-member exchange was measured with nine items based on Liden, Wayne and Stilwell (1992) and Bauer and Green (1996). Task complexity was measured using three items from the Work Design
Questionnaire (WDQ) based on Morgeson and Humphrey (2006).
4 Results
This section explains the collected data of this study. The data will be analyzed and the hypotheses will be tested. The results are presented in two parts: tests of the validity and reliability of the variables, and tests of the hypotheses.
4.1 Tests of validity and reliability
To perform the statistical analyses, the Statistical Software Package for Social Sciences (SPSS) was used. The means, standard deviations, correlations and reliabilities of the variables are exhibited in Table 1.
First, a frequency test for all variables was conducted to examine if there were any errors or missing values in the data. There were no errors in the possible ranges of variables. Furthermore, only cases that had no missing data for all variables were analyzed (list wise deletion).
Table 1. Means, Standard Deviations, and correlations Variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 1. Gender 1,29 0,45 - 2. Age 50,86 9,25 -.27** - 3. Task complexity 2,74 1,47 .04 -26** - 4. Mbea 4,42 1,18 -.06 .07* .03 (.84) 5. Contingent reward 3,25 1,41 .01 -.06 -.01 .46** (.94) 6. Initiating structure 4,83 1,07 .00 .02 -.09** .56** .50** (.91) 7. Consideration leadership 4,99 1,13 .00 .01 -.18** .28** .45** .65** (.92) 8. Transformational leadership 4,90 1,29 .00 .06 -.18** .38** .53** .74** .81** (.96) 9. Ethical leadership 4,99 1,09 .00 .01 -.16** .34** .48** .71** .83** .81** (.93) 10. Passive leadership 3,42 1,27 .00 -.03 .17** -.30** -.34** -.55** -.57** -.57** -.53** (.87) 11. Job autonomy 5,56 1,36 .01 .16** -.42** .03 .23** .29** .46** .45** .48** -.23** (.96) 12. Job satisfaction 5,78 0,98 .01 .11** -.36** .09** .15** .25** ,33** .34** .36** -.24** .44** (.85)
Notes: N = 924. Correlations are reported below the diagonal. Coefficient alpha reliability estimates are on the diagonal in bold.
**. Correlation is significant at the 0.01 level (2-tailed). * . Correlation is significant at the 0.05 level (2-tailed).
Next, the counter-indicative items were recoded and descriptive statistics, skewness, kurtosis and normality tests were computed for the variables. Initiating structure leadership, consideration leadership, transformational leadership and ethical leadership had moderately negatively skewed distributions. Job autonomy and job satisfaction were substantially negatively skewed. The data was checked for potential outliers and no outliers were found.
To examine the consistency of the measurements, reliability checks were done for all variables. Cronbach’s alpha, that represents the estimator of the internal consistency, has been tested to verify if all items in one scale measure the same. All variables have a Cronbach’s alpha > .70. This indicates high levels of internal consistency which means all items have a good correlation with the total score of the scales (Saunders, Lewis & Thornhill, 2012).
Scale means were computed for all variables and the correlations between the variables were calculated. The correlation among the variables revealed the leadership behaviors were positively correlated with each other; only passive leadership was negatively related to the leadership behaviors. Contingent reward, initiating structure, consideration leadership, transformational leadership and ethical leadership were positively correlated with job
autonomy. Initiating structure, consideration leadership, transformational leadership and ethical leadership were positively related to job satisfaction. Finally, job autonomy and job satisfaction were positively related.
4.2 Testing the hypotheses
To test my hypotheses, hierarchical regression analyses were performed with standardized predictor variables (Aiken & West, 1991; cf. Buengeler & Den Hartog, 2015). In the first step of the regression analysis, I entered the control variables gender, age, job function and task complexity, followed by the respective predictor (leadership) in the second step.
4.2.1 Hypothesis 1
In hypothesis 1a, I predicted that task-oriented leadership was negatively associated with job autonomy. The results for task-oriented leadership are presented in Table 2. By entering management by exception-active in the second step of the regression, the total variance explained by the model as a whole was 26.3%; F (10, 890) = 31.79; p < .01, and an additional 0.4% of variance in job autonomy, after controlling for gender, age, job function and task
complexity, could be explained (R²change = .004; F change [1, 890] = 4.55; p < .05).
Management by exception-active was positively related to job autonomy, however only at the .10 level of significance (β = .06; p < .05). By entering contingent reward in the second step of the regression, the total variance explained by the model as a whole was 30.6%; F (10, 890) = 39.32; p < .01, and an additional 4.7% of variance in job autonomy after controlling
for gender, age, job function and task complexity, could be explained (R²change = .047; F
change [1, 890] = 60.35; p < .01). Contingent reward was significantly positively related to job autonomy (β = .22; p < .01). Entering initiating structure in the second step of the
regression, the total variance explained by the whole model was 32.4%; F (10, 890) = 42.580; p < .01, and an additional 6.4% of variance in job autonomy after controlling for gender, age,
job function and task complexity, could be explained (R²change = .064; (R²change = .064; F
change (1, 890) = 84.48; p < .01). Initiating structure was significantly positively related to job autonomy (β = .26; p < .01). In sum, the results showed a positive relation between task-oriented leadership behaviors and job autonomy. Therefore, hypothesis 1a was rejected. In hypothesis 1b, I predicted that relation-oriented leadership is positively associated with job autonomy. The results for relation-oriented leadership are presented in Table 2. After entering consideration leadership in the second step of the regression, the total variance explained by the model as a whole was 39.4%; F (10, 890) = 57.98; p < .01, and an additional 13.5% variance in job autonomy, after controlling for gender, age, job function and task
complexity, could be explained (R²change = .135; F change [1, 890] = 198.56; p < .01).
Consideration leadership was significantly positively related to job autonomy (β = .38; p < .01). Hence, hypothesis 1b was supported.
In hypothesis 1c, I stated that change-oriented leadership is positively associated with job autonomy. The results are presented in Table 2. After entering transformational leadership in the second step of the regression, the total variance explained by the model as a whole was 38.8%; F (10, 890) = 56.54; p < .01, and an additional 12.9% variance in job autonomy, after
controlling for age, gender, job function and task complexity, could be explained (R²change
= .129; F change [1, 890] = 187.86; p < .01). Transformational leadership was significantly positively related to job autonomy (β = .37; p < .01). Hypothesis 1c was thus supported. In hypothesis 1d, I expected that ethical leadership is positively associated with job autonomy. The results for ethical leadership are presented in Table 2. By entering ethical leadership in step two of the regression, the total variance explained by the model as a whole was 41.2%; F (10, 890) = 62.27; p < .01, and an additional 15.2% variance in job autonomy,
37
MBEA Contingent reward Initiating structure Consideration Transformational Ethical Passive Variable Step 1 Step 2 Step 1 Step 2 Step 1 Step 2 Step 1 Step 2 Step 1 Step 2 Step 1 Step 2 Step 1 Step 2
Step 1 Age .07* .06* .07* .08** .07* .07* .07* .08* .07* .06* .07* .08* .07* .07* Gender .07* .08* .07* .08** .07* .08** .07* .07** .07* .06* .07* .07* .07* .08* Task complexity -.30** -.30** -.30** -.29** -.30** -.27** -.30** -.23** -.30** -.24** -.30** -.24** -.30** -.28** Job function B .13** .13** .13** .12** .13** .10** .13** .10** .13** .10** .13** .09** .13** .13** Job function E .18** .19** .18** .18** .18** .19** .18** .16** .18** .18** .18** .18** .18** .18** Job function F .17** .18** .17** .17** .17** .17** .17** .16** .17** .16** .17** .16** .17** .17** Job function I .16** .16** .16** .17** .16** .16** .16** .15** .16** .15** .16** .16** .16** .16** Job function Management .17** .17** .17** .16** .17** .14** .17** .13** .17** .12** .17** .10** .17** .15** Job function Staff .21** .22** .21** .21** .21** .22** .21** .19** .21** .20** .21** .18** .21** .21** Step 2 MBEA .06* Contingent reward .22** Initiating structure .26** Consideration .38** Transformational .37** Ethical .40** Passive -.16** R² .26 .26 .26 .31 .26 .32 .26 .39 .26 .39 .26 .41 .26 .28 R² adjusted .25 .26 .25 .30 .25 .32 .25 .39 .25 .38 .25 .41 .25 .28 R² .004 .05 .06 .14 .13 .15 .02 F 34.68** 31.79** 34.68** 39.32 34.68** 42.58** 34.68** 57.98** 34.68** 56.54** 34.68** 62.27** 34.68** 35.24** F 4.55* 60.35** 84.48** 198.56** 187.86** 230.28** 30.11** Notes: N = 924. Standardized regression coefficients are reported.
after controlling for age, gender, job function and task complexity, could be explained (R² change = .152; F change [1, 890] = 230.28; p < .01). Ethical leadership was significantly positively related to job autonomy (β = .40; p < .01). Thus, hypothesis 1d was supported. In hypothesis 1e, I proposed that passive leadership is positively associated with job autonomy. The results for passive leadership are also presented in Table 2. Passive leadership was entered in step two of the regression and the total variance explained by the model as a whole was 28.4%; F (10, 890) = 35.24; p < .01, and an additional 2.4% variance in job autonomy, after controlling for age, gender, job function and task complexity, could be
explained (R²change = .024; F change [1, 890] = 30.11; p < .01). Passive leadership was
significantly negatively related to job autonomy (β = -.16; p < .01). Therefore, hypothesis 1e was rejected.
Next, I examined the relative effectiveness or importance of task-oriented leadership, relational-oriented leadership, change-oriented leadership, passive leadership, and ethical leadership in influencing job autonomy. The relative importance also used term relative weight can be defined as “the proportionate contribution each variable makes to R², considering both its unique contribution and its contribution when combined with other variables” (Johnson, 2000, p. 2). Following LeBreton, Binning, Adorno and Melcher (2004), a common index of importance is the squared standardized least-squares regression weight
(β2
j) provided for each predictor variable (leadership behavior). According to these authors,
squared Betas provide information about the relative importance of variables on what Budescu (1993) called total effects, and deliver estimates of relative importance, conditional on all of the other predictor variables. Results are presented in Table 3. A regression analysis with weighted least square is performed. By entering management by exception-active, contingent reward, initiating structure, consideration leadership, transformational leadership, ethical leadership and passive leadership in the second step of the regression, the total
Table 3. Results of weighted least square regression analyses with job autonomy as dependent variable.
variance explained by the model as a whole was 46.2%, F (16, 884) = 47.392; p < .01, and an additional 19,2% variance in job autonomy, after controlling for gender, age, job function and
task complexity, could be explained (R²change = .192; F change [7, 884] = 45.044; p < .01).
Five out of seven leadership variables were significantly positive, with ethical leadership recording the highest Beta value (β = .24; p < .01), followed by transformational leadership (β = .19; p < .01), consideration leadership (β = .13; p < .01), passive leadership (β = .09; p < .01) and contingent reward (β = .07; p < .05). Management by exception-active was significantly negative (β = -.09; p < .01) and initiating structure was negative, but
nonsignificant. In the third step, the standardized regression coefficients for the leadership
Job autonomy
Variable Step 1 Step 2 Step 3 Step 4
Step 1 Age .06 .09 Gender .08 .06 Task complexity -.30 -.23 Job function B .13 .09 Job function E .17 .15 Job function F .17 .14 Job function I .14 .13 Job function Management .16 .10 Job function Staff .20 .16 Step 2 Mbea -.09** .008 5.6% Contingent reward .07* .005 3.4% Initiating structure .05 .003 2,1% Consideration leadership .13** .02 13.9% Transformational leadership .19** .04 28% Ethical leadership .24** .06 41.7% Passive leadership .09** .008 5,6% Notes: N = 924. Standardized least-squares regression coefficients are reported. **. p < 0.01 level (2-tailed). *. p < 0.05 level (2-tailed).
behaviors were squared, which resulted in relative weights. In step 4, the percentages were calculated by dividing the individually squared regression coefficients by their sum and multiplied by 100. The most important leadership behavior predicting job autonomy is ethical leadership (41.7%), followed by change-oriented leadership (28%), relation-oriented
leadership (13.9%), passive leadership (5,6%), management by exception-active (5,6%), contingent reward (3,4%), and initiating structure (2.1%). Therefore, hypothesis 1f was supported.
4.2.2 Hypothesis 2
In hypothesis 2, I expected that employees’ perceived degree of job autonomy is positively related with job satisfaction. Results are presented in table 4. In the first step of the regression the four control variables were entered: gender, age, job function and task complexity. After entering job autonomy at step 2 the total variance explained by the model as a whole was 23.5%; F (10, 890) = 27.31; p < .01, and an additional 8.4% variance in job satisfaction, after
controlling for age, gender, job function and task complexity, could be explained (R²change
= .084; F [1, 890] = 97.35; p < .01. Job autonomy was significantly positively related to job satisfaction (β = .34; p < .01) and hypothesis 2 was thus supported.
