The effect of transformational leadership on self-efficacy of employees : the influence of management control systems

(1)

The Effect of Transformational Leadership on

Self-Efficacy of Employees

- The Influence of Management Control Systems -

Ashna A. Birtantie (11160640) Faculty of Economics and Business

Executive Programme in Management Studies – Strategy track University of Amsterdam/ Amsterdam Business School Supervisor: Prof. dr. Frank Jan de Graaf

(2)

2

Statement of Originality

This document is written by Student Ashna Birtantie 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.

(3)

3

Preface

“There is simply no substitute for hard work when it comes to achieving success” Heather Bresh

With pride and hard work, you have before you the thesis “The Effect of Transformational leaders on the Self-efficacy of employees”. This is my final assignment in order to receive my master’s degree.

I started the Executive Master Programme at the University of Amsterdam because of my eagerness to enhance my academical knowledge. The past two and a half years at the University have been the most memorable times in my life. Even though challenging at times but with perseverance I made it.

I would like to thank my supervisor Prof. dr. Frank Jan de Graaf for his guidance through this process. Furthermore I would like to thank my colleagues of Rabobank

Amsterdam, without their support this would not be possible. I would also like to thank my fellow students for making it such fun every Friday at the UvA, sharing laughs, stories but also working hard when needed.

A special thanks is in order for my dear friends who listened to my complaints and tried to encourage me in times I needed it. Furthermore, I would like to offer my sincere gratitude to my parents and boyfriend. Even though 10.000 km apart, you still managed to encourage me to stay positive and kept believing in me.

Enjoy reading!

Ashna Birtantie

(5)

5

Abstract

This paper uses a quantitave method to investigate the relationship between transformational leadership and self-efficacy through the use of the framework on the Levers of Control by Simons (1995). Literature findings imply a positive relation between transformational leadership and self-efficacy as well as a positive mediating relation when using positively oriented management control systems (like beliefs and interaction). Research amongst 109 employees of Rabobank Amsterdam gave interesting results. Initial findings show no relation between leadership and self-efficacy and that mcs does not contribute as a mediating variable. Because of this, further statistical analysis was done on the groups of respondents. The

respondents were split according to the segments they work in. These new findings seem to support the theory that there is a positive relation between leadership and self-efficacy. An explanation for the lack in the initial analysis can be that the Corporate division experiences more autonomy then the consumers group. Research also found that negative mcs play a positive mediating role between transformational leadership and self-efficacy. An explanation of this analysis can be that both groups have commercial functions and because of this

respond better to boundaries and control systems that report their performance (negative mcs) instead of positive mcs (diagnostic and interactive control systems). This paper contributes to the literature findings that mcs do play a role in the relationship between leadership and self-efficacy and that managers, in line with the theory of Simons (1995), use mcs effectively in order to balance control and empowerment. The nuance here is situationally they might choose for negative mcs instead of positive mcs to achieve self- efficacy.

Keywords Transformational leadership, management control systems, self-efficacy, Levers of

(6)

6

1. Introduction

In recent years, the Levers of Control by Simons have gotten more and more attention from scholars. In 2012, the number of citations on Google Scholar was a mere 800 (Tessier & Otley, 2012). This number has since quadrupled in the recent years to a mere 3.300 in 2017 (Google Scholar).

Simons (1995, p.5) describes management control systems as “formal information-based routines and procedures managers use to maintain and alter patterns in organizational activities”. The purpose of management control system (“mcs”) is to provide managers with useful information in their decision-making, planning and evaluation process (Merchant & Otley, 2006). Management control systems are considered tools managers can use in order to balance empowerment and control (Simons, 1995). As contemporary organizations are faced with rapidly changing technology, customers, environment, competitors and other external factors, it is important for them to renew themselves in order to survive and prosper

(Danneels, 2002). Henri (2006) describes the use of management control systems as a form of RBV (resource base view) which organizations can deploy effectively in order to gain

competitive advantage.

Dependent on strategic uncertainty and risk, the choice is made which control systems are used. This has an impact on the organization through organizational learning and efficient use of management attention (Simons, 2000). The choice which mcs are used is dependent on the type of leader1 (Kleine & Weibenberger, 2013). Research acknowledges that leadership style can influence a firm’s strategy and implementation of management control systems (Abernethy et al, 2010; Menguc et al, 2007).

(7)

7

A way to align behavior with organizational goals is to implement systems like performance management and reward schemes (Ngyen et al, 2016; Widener, 2007). This implies that employees are only motivated through compensation. However Olafsen et al (2015, p.9) states that “employees are motivated to perform better when management provides positive feedback, acknowledging the employees’ skills and efforts, offering opportunities to learn new things, asking open-ended questions, listening actively to the employees, inviting participation in solving important problems, acknowledging the employees’ perspective, and offering choices within the structure of the organization”.

According to Rothfelder et al (2013) different leadership styles show a correlation with job satisfaction. Transformational leadership was positively related to job satisfaction according to the respondents. Whereas transactional leadership showed less job satisfaction because employees disliked the notion of management attention to failures.

According to Simons the four levers together generate opposing forces. Two of these levers (beliefs and interactive control) generate positive and inspirational forces whereas the other two levers (boundary and diagnostic control) create coercion and ensures compliance with orders (Simons, 1995). Selecting which levers to use and using them properly is a determinative decision for managers. “The choice in the type of levers used by manager reflect their own personal values, reveal opinions of subordinates, affect the probability of goal achievement, and influence the organization’s long term ability to adapt and prosper” (Simons, 1995 p.8).

According to the theory of Bass (1991) transformational leaders possess characteristics such as personal attention towards subordinates, stimulate creativity and inventiveness as well as ensuring that employees are connected to organization’s values. Based on aforementioned findings of Simons and Bass it is likely that transformational leaders show a preference for

(8)

8

using management control systems that are in line with their characteristics. The choice of managers which management control system are deployed makes it clear that these control systems cannot be influenced but can rather be seen as tools managers use effectively.

Management control systems are used in this study as a mediating variable. Mediation occurs when the relationship between a dependent variable (transformational leadership) and an independent variable (self-efficacy) is explained by their relationship to a third variable (management control systems) (Field, 2013). Mcs in this study are treated as instruments at the disposal of the transformational leader to apply to empower and “transform” the

employees.

This study aims to determine the relation between transformational leadership and self-efficacy and find out if this relation is explained by the certain types of management control systems managers can use in organizations. Based on this the following research question is formulated:

What is the effect of transformational leadership on the self-efficacy (SE) of employees through the use of management control systems (mcs)?

Relevance

There are numerous studies on the levers of control as to how and when to use these levers (Simons, 1995; Widener, 2007; 2008; 2009; 2013) and even how these levers relate to leadership (Ngygen et al, 2016). There are many studies about transformational leadership and how transformational style influences performance (Xiong & Fang, 2014), job

satisfaction (Rothfelder et al, 2013), and even employee engagement (Prochazka et al, 2017; Strom et al, 2014; Khuong & Yen, 2014). However, little is known about transformational leaders and how this type of leadership style contributes to the self-efficacy of employees and to what extent the Levers of Control play a part in their effectiveness. By means of qualitative

(9)

9

research this gap is further examined. The results of this research is not only of importance to science but also to Rabobank Amsterdam, where this study is conducted. Based on the findings of this study managers in the organization can deploy mcs effectively in order to enhance self-efficacy of employees.

Structure thesis

This study is structured as follows: chapter two is an literary analysis of the individual variables. This chapter includes an analysis of the relationship between these variables and ends with the derived hypotheses. The third chapter describes the methodology used in this study, collecting relevant data and how this data is analyzed. Chapter four presents the results of the statistical analysis including the regression and mediation analysis of the data collected. Chapter five consists of the discussion part of this study. The last chapter provides an overall conclusion.

(10)

10

2. Literature review

In order to answer the stated research question, it is important to take a closer look at the subjects that will be researched. This chapter discusses the setting where the study is conducted. It will then follow with current literature findings on leadership, the subject of management control systems and self-efficacy. First starting with the theoretical foundation on which leadership and mcs are built and followed by the relation between mcs, leadership and the self-efficacy of employees. To conclude the chapter outlines the hypothesis on the workings of these variables.

2.1 Setting

The Rabobank is a Dutch cooperative bank with its headquarters located in Utrecht. It is one of the larger financial services companies in the Netherlands. Employees serve retail and business customers on a daily basis offering a wide range of financial products and services. The company has more than a hundred local banks each operating to a certain degree of independence from the headquarters.

According to Mintzberg (1979) looking at the organizational structure of Rabobank Group it is a typical divisionalized organization. The characteristics of a divisionalized firm are that these firms are relatively old and very large, they operate in a relatively simple/stable

environment and have a high amount of middle line control. Rabobank is not an integrated firm as it is rather more a set of independent entities (local Rabobanks) grouped together by a loose administrative structure (Mintzberg, 1979). Mintzberg states that firms divisionalize because of the different products and services they provide and to enhance the scope of control. By placing responsibility to the individual divisions, the headquarter is less focused

(11)

11

on control. Thus, increasing the complexity of mcs (as administrative controls) in such large organizations (Chenhall, 2003).

The headquarters relies on standardization of output through performance control systems. Meaning that they solely look at the divisional results instead of what is done in order to achieve these results. Also headquarters includes several central supporting departments for local banks in order to facilitate an optimal work flow. Some examples of these departments are legal-, technical-, marketing-teams. Looking at the role of headquarters it has two functions for the local banks, one being an entrepreneurial role, which is creating value and the other being an administrative one, which is necessary for loss prevention in order to effectively use and allocate the firm’s capabilities (Chandler,1991).

Although the overall structure of Rabobank Group is that of a divisionalized organization, when zooming in on the local banks one sees a structure that is more of a machine

bureaucracy. According to Mintzberg (1979) this type of configuration features environments which are simple and stable and where work processes are standardized. Meaning that the specification of the content of work needs to be followed directly. Usually these types of organizations are found in the more mature organizations and where decision making follows a formal line of authority (Mintzberg, 1979). In these types of firms mcs are focused on compliance through the formal line of hierarchy (Chenhall, 2003). Because of legislation compliance is an important focusing point for Rabobank. Because of this formal line of hierarchy more boundary and diagnostic control mechanisms are used in order to shift management attention in more productive ways (Simons, 1995). These types of situations however demand more transactional leaders (Bass, 1991). This type of leadership is

characterized as having a contingent reward system and manage followers on exception based occurrences (Bass, 1991). Meaning that followers are motivated through promises, praises

(12)

12

and rewards and the task of managers is to monitor performance and correct mistakes if necessary (Bass & Steidlmeier, 1999).

Rabobank on the other hand seems to encourage transformational leadership styles. Managers have excessive on the job training and coaching in not only getting the job done but inspire, motivate and stimulate their subordinates in not only achieving organizational goals but personal growth as well. Because of changing and challenging external factors it is necessary for Rabobank to adapt to these circumstances by shifting focus from

standardization, top-down strategy, keeping things on track to being more customer oriented, focus on customization and meeting customer needs. To shift focus also requires a different set of management skills.

This study focuses on enhancing self-efficacy of employees at Rabobank Amsterdam using transformational leadership despite the fact that the organization’s configuration and mcs are tempting management into a more transactional leadership style.

2.2 Transformational Leadership

The last four decades the concept of leadership has gotten a lot of attention from scientists. Burns (1978, p. 425) defines leadership as “the reciprocal process of mobilizing by persons with certain motives and values, various economic, political, and other resources, in context of competition and conflict, in order to realize certain goals independently or mutually held by leaders and followers”. Bass & Avolio (1994) describe specific leadership styles in terms of effectiveness, degree of activity and frequency of utilization. Using the full range of

leadership development model they distinguish five specific leadership styles, these are: laissez-faire, management-by-exception-passive, management-by-exception-active,

constructive transaction and transformational leadership. In this research, the main focus will be on the transformational leadership style.

(13)

13

As the name implies, transformational leaders are concerned with the transformation or change of followers’ fundamental values, goals and aspirations. The basic difference with other types of leadership styles is that followers, under a transformational leader, share the organization’s values and are committed to the organizational goals (Rothfelder et al, 2013). Yukl (1989) refers to transformational leadership as ‘the process of influencing major changes in the attitudes and assumptions of organization members and building commitment for the organization’s mission or objectives’. Northouse (2007) concludes that transformational leaders assess followers’ motives, satisfy their needs and treat them as important human beings. This treatment makes followers accomplish more than what they are expected to.

According to Bass (1985) transformational leaders have three main factors that characterize their behavior. These are individualized consideration, intellectual stimulation, and charisma/inspirational motivation. These characteristics are used in this study and the following subparagraphs will further elaborate on its content.

Individualized consideration

Transformational leaders are individually considerate to each of their followers. According to Bass (1991) they pay close attention to the differences among their followers. Aiming to look at the individual who needs guidance in order to develop and grow to their full potential. Transformational leaders can achieve development and growth through delegation of projects “in order to provide opportunities for each follower to self-actualize and to attain higher standards of moral development” (Bass & Avolio, 1993ª). According to Nguyen et al (2016, p. 2014) individualized consideration is “a method of communicating timely information to followers through coaching and mentoring which provides for constant follow-up and feedback”.

(14)

14

Intellectual stimulation

According to Bass (1991) intellectual stimulation is the willingness and ability of leaders to provide followers insights to novel ways of looking at old problems, to teach them to see problems as something to be solved, and suggest rational solutions. Nguyen et al (2016) argue that intellectual stimulation creates new ways of exploring an organization’s set of goals. Overall this characteristic of a transformational leader ensures that their followers are stimulated and encouraged to be creative and inventive (Rothfelder, 2013).

Charisma/Inspirational motivation

According to Bass (1991) transformational leaders should attain charisma in order to succeed as a leader. Charismatic leaders have a great deal of power and influence on their followers. Because of this the followers’ level of trust and confidence increases. Bass (1991) explains that charismatic leaders inspire and excite followers into thinking that they can solve problems/difficulties with extra effort. Rothfelder (2013, p. 204) argues that inspirational leadership “demonstrates commitment to organizational goals, enhances team spirit and clearly articulates high expectations.” He also states that these types of leaders, in addition to the aforementioned, also provide followers with not just an optimistic and attractive vision of the future but an achievable one as well.

2.3 Levers of Control

Simons (1995, p.9) describes management control systems (mcs) as “formal information-based routines and procedures managers use to maintain and alter patterns in organizational activities”. The reason for using management control system (mcs) is to provide managers with useful information in their decision-making, planning and evaluation process (Merchant & Otley, 2006). Robert Simons was the first to introduce the Levers of Control (“LOC”) framework in his book in 1995.

(15)

15

The LOC framework indicates “that tensions are managed between freedom and constraint, between empowerment and accountability, between top-down direction and bottom-up creativity, between experimentation and efficiency” (Simons, 1995, p.4). Simon distinguishes four types of control systems. These are the belief systems, boundary systems, interactive control systems and diagnostic control system. According to Simons these four control systems work together to benefit the firm. The power of these four levers does not lie in their individual use but how they work, complement and balance each other (Simons, 1995). The levers together generate positive and negative forces that jointly create dynamic tension between innovation and strategic renewal on one hand and predictable goal

achievement on the other in order to secure the long-term success of the firm (Raisch & Birkenshaw, 2008). The beliefs system and interactive controls are considered positive control systems whereas the boundary system and diagnostic controls are considered a negative control systems (Widener, 2007).

(16)

16

Beliefs system

“The beliefs system is the explicit set of organizational definitions that senior managers communicate formally and reinforce systematically to provide basic values, purpose, and direction for the organization” (Simons, 1995, p. 34). According to Widener (2007) through the beliefs system core values are communicated to inspire and motivate employees to search, explore, create and expand effort engaging in appropriate actions. For Rabobank an example of beliefs system is the intranet that all employees working at Rabobank use. The intranet is used to inform all employees on all the different topics and activities which employees deal with on a daily basis. It also shows inspiring vlogs and blogs about what Rabobank is about and how projects which Rabobank participates in are connected to the core values of the firm.

Boundary system

“The boundary system delineates the acceptable domain of strategic activity for

organizational participants” (Simons, 1995, p. 39). The boundary system emphasizes the actions employees should avoid. Both the beliefs and boundary system are similar because both control systems motivate employees. Where the beliefs system encourages positive behavior the boundary system prohibits certain actions of employees. For Rabobank the code of conduct is very important. In order to compel employees to keep to the code of conduct a ‘bankeroath” was initiated. This means that Rabobank is not solely responsible if and when delicate information of customers is misused but the employee in question is as well. This oath is instilled so that employees are more aware of the consequences when misusing data/information.

Diagnostic control system

The diagnostic system is intended to motivate employees to perform and align behavior with the firm’s objectives (Widener, 2007). This control system makes it possible for managers to get information on critical success factors to focus their attention on organizational drivers

(17)

17

that must be monitored in order to realize the firm’s strategy (Widener, 2007). Rabobank uses Siebel as tool in order to get information on critical success factors. Siebel is a Customer Relationship Management tool where client data and requests are stored. This tool enables management to monitor workflow and performance of employees. Diagnostic control systems are the formal information systems which managers use to track organizational goals and correct mistakes from standard work processes. According to Simons diagnostics control systems have three characteristics. First, it makes measurement of outputs of processes

possible, it also has the ability to compare actual results to predetermined standards and last, it has the ability to correct deviations.

Interactive control system

The interactive system is characterized by active and frequent dialogue. Managers use information generated from systems in order to frequently and regularly initiate face to face meetings in order to interpret and discuss data. Simons (1995) states that the interactive control system has four characteristics that sets it apart from the diagnostics control system. First, it keeps focus on changing information that top level has deemed as potentially

strategic. Second, because of the significance of the information frequent attention is needed from managers throughout all levels of the organization. Third, this information is best interpreted and discussed face to face with all employees. The last characteristic of the

interactive control system is the continual interaction about data, assumptions and action plans (Simons, 1995). According to Simons the interactive control system stimulates learning and search which in return allows for new strategies to emerge. Within Rabobank frequent meetings are part of working at Rabobank. Usually three to four hours in the week are spend in meetings where data is discussed and interpreted, and progress is monitored.

(18)

18

The main difference between diagnostic and interactive control system lies in the fact that interactive control systems require management attention constantly whereas diagnostic control systems are used as an ex post monitoring system (Simons, 1995).

As markets become more dynamic and intricate, organizations are faced with

opportunity on one hand and competitive environments on the other (Simons, 1995). The four levers of control set in motion powerful forces that amplify each other to face opportunity and competition. By using these four levers effectively managers can focus on innovation and creativity without sacrificing control (Simons, 1995).

2.4 Self-efficacy

Bandura (1994) defines self-efficacy as an individual’s persuasion of their ability to reach goals and to feel in control of events that influence them on a daily basis. Self-efficacy influences how people feel, think and behave and what motivates them (Bandura, 1994). Bandura (1994) argues that a high self-efficacy increases human accomplishment and personal well-being. People tend to approach difficult tasks as a challenge to overcome instead of a hurdle to be avoided.

In his study, Bandura mentions four main forms of influence as to increase a strong sense of efficacy. The first influence is mastery experiences. Bandura (1994) implies that in order to manage life changing circumstances it involves acquiring cognitive, behavioral, and self-regulatory tools for creating and executing appropriate courses of action.

The second influence to enhance self-efficacy is vicarious experiences through social models. This indicate that people seek proficient role-models that possess competencies to which they aspire. Bandura (1994) argues that these so called social models transmit knowledge and teach observers skills and strategies for managing environmental demands.

(19)

19

The third influence Bandura mentions is social persuasion. This is a way of strengthening a person’s believe into thinking they have what it takes to be successful. Bandura (1994) argues that people who are verbally persuaded that they possess the competencies to master given activities are likely to mobilize greater effort and sustain it than if they harbor self-doubt and dwell on personal deficiencies when problems arise.

The fourth and last influence is emotional arousal. Bandura (1994) explains that people with high self-efficacy are likely to view there affectiveness as an energizing facilitator of performance whereas people with an over amount of self-doubt regard their arousal as a debilitator.

2.5 Relationship between variables and hypotheses

This paragraph tries to explain the relationship between the subjects researched on in this study and how based on those findings the different hypotheses are formulated. First the relationship between leadership and self-efficacy is explained. Followed by the explanation between leadership and mcs and lastly the indirect effect between mcs and self-efficacy.

2.5.1 Transformational leadership and self-efficacy

Research has found that self-efficacy positively relates to job attitudes (Saks, 1995), job performance (Stajkovic & Luthans, 1998), and training proficiency (Martocchio & Judge, 1997) and helps reduce stress factors on employee psychological well-being (Jex & Bliese, 1999).

According to the study by Prozacha et al (2017) it is argued that self-efficacy and transformational leadership are related. They argue that transformational leadership enhances self-efficacy through at least three of the four influences mentioned by Bandura (1994). One of the first influences Bandura mentions is mastership experience, which refers to a person’s ability to be skillful in a certain area. Transformational leaders possess the ability to provide

(20)

20

positive feedback to their followers on what they accomplished (Bass, 1991), this increases the confidence level and so does mastership experience (Prozacha et al, 2017). Because transformational leaders are inspirational motivators they also possess role-modeling behavior this can increase vicarious experience. Through intellectual stimulation transformational leaders share their vision with their followers (Prozacha et al, 2017). Thus, trying to share followers’ ideas and use skills and abilities to solve problems therefore enhancing social persuading. In the study, self-efficacy was a moderator for the relationship between transformational leadership and engagement. The outcome was that self-efficacy has a positive (although weak) relation with transformational leadership.

The study by Kark et al (2003) supports the theory that transformational leadership has a positive correlation with empowerment through self-efficacy, group efficacy and OBSE (organization-based self-esteem). In their research combined referred to as social

identification. Chen and Bliese (2002) researched the mechanisms of self-efficacy. They found that work experience, role clarity and psychological strain are direct predictors for self-efficacy.

Consistent with previous research we can expect that transformational leadership has a positive effect on the self-efficacy of employees. Based on these findings the first hypothesis is developed.

Hypothesis 1. The use of transformational leadership will have a positive influence on

employee self-efficacy.

Based on the influences of self-efficacy of Bandura (1994) and the characteristics of

transformational leadership such as inspirational leadership and individualized consideration this study tends to find a positive relation between these two variables.

(21)

21 2.5.2 Transformational leadership and mcs

Back in the 1950s and 1960s when looking at the typical machinelike bureaucratic

organizations, managers executed control by literally telling people how to do their job and monitoring every move employees made (Simons, 1995). However, many managers who are facing markets that are dynamic and highly competitive, it is just not realistic to spend all of their time monitoring and instructing employees what to do and how to do their job. Simons (1995) expresses that these managers need to urge incumbents to initiate improvements and find new ways to responding to the needs of customers and changing environment. By adopting the four levers of control (boundary, diagnostic, interactive and beliefs systems) managers can increase employee creativity. Each of these levers have their specific mechanism that contributes to organizations goals.

Boundary systems enables individual creativity within established limits (code of conducts, procurement systems). Beliefs systems makes it possible for employees to have a clear and coherent understanding of core values of their organizations. Which in return can inspire employees to create new opportunities (Simons, 1995). Simons considers beliefs systems as the opposite of boundary systems.

Managers can use diagnostic control systems for several reasons, for instance to achieve an efficient allocation of scarce resources, to define goals, to motivate and make it possible for management to tend to other activities then monitoring employees2. The last lever, the interactive system, enables organization attention to strategic uncertainties and new initiatives and strategies thus responding to opportunities and threats (Simons, 1995).

Interactive control systems differ from diagnostic control system because of continual personal involvement from managers. Thus resulting in managers choosing where to allocate

(22)

22

attention in organizations and because of this signals where subordinates should allocate their attention to.

As mentioned before transformational leaders have several characteristics, these are charismatic/inspirational motivator, intellectual stimulation and individualized consideration. Because of these characteristics and how they work, it is likely that transformational leaders use of style can be related to the use of positive mcs. Interactive control system has

characteristics of face to face conversations/meetings with managers, employees and other subordinates. Because of this the characteristics of transformational leaders such as

intellectual stimulation and individualized considerations seems to overlap. The beliefs system and inspirational motivator have the same function, this being inspiring subordinates to believe in organization mission and inspire to search for new opportunities (Bass, 1991; Simons, 1995). Below in table 1 the characteristics of transformational leaders and the link to positive mcs is presented.

Table 1: TL and positive mcs

Transformational leadership Positive mcs

Charisma/inspirational motivator - Increases trust and confidence - Increases effort

- Enhances team spirit

- Commitment to organizational goals

Beliefs system

- Creates understanding to core values which in return - Inspires and increases effort

to search opportunities Individualized consideration

- Personal attention to skills and abilities - Coaching and mentoring

- Highly communicative on individual level

Interactive control system

- Continual interaction about information

(23)

23

- Frequent attention Intellectual stimulation

- Provides insights to solving problems - Stimulates creativity and inventiveness

Belief system and Interactive control system

- Increases effort in order to solve problems

- Frequent attention Bass (1991); Simons (1995)

Even though there are characteristics which seem to overlap use of mcs and transformational leadership style, Simons does not mention how mcs are deployed by managers. The theory of transformational leadership on the other hand does mention means by which managers can use their resources to their full potential. Hence, arguing that transformational leaders can use the available mcs or even further develop the use of available mcs in their organization.

Based on the above literature findings the second hypothesis is formulated.

Transformational leaders show a preference for positive mcs over negative mcs.

H2a. Transformational leaders will have a positive relation with the use of positive mcs

(beliefs system and interactive control system).

H2b. Transformational leaders will have a negative relation with the use negative mcs

(boundary system and diagnostic control system).

2.5.3 mcs and self-efficacy

Burney and Widener (2013) argue that strategic performance management systems (spms) are related to self-efficacy. Spms are part of the interactive control system because of information they provide managers in order to shift the focus of organizations to areas that need attention

(24)

24

(Simons, 1995). Burney and Widener found that self-efficacy positively relates to employee performance. Hence arguing that SPMS relates to performance via self-efficacy.

Kaplan and Norton (1996) argue that spms procure subordinates with a better sense of understanding of how achieving individual goals will affect the firm’s ability to achieve goals. “Self-regulation of goals deemed important will increase employee efficacy since self-regulation makes salient the requirements necessary to effectively perform their tasks” (Burney & Widener, 2013, p. 120). That is that employees will keep their focus and work toward organization’s goals through performance measures which are essential and self-regulated. Gist and Mitchell (1992, p.186) define self-regulation as “a comprehensive process of cognitive, individual determination of behavior”.

According to Simons (1995) the four levers of control work together and complement each other. “The positive and negative mcs create a dynamic tension between innovation and strategic renewal on one hand and goal achievement on the other, both needed to secure organization’s long-term success” (Kruis, Speklé & Widener, 2016, p. 9). As mentioned, the positive levers are interactive and beliefs control. Taking a closer look at interactive systems, which rely on information for managers, this information is made available through

interactive systems such as Performance Management systems. Suggesting that a good use of pms can increase self-efficacy of employees.

Burney and Widener (2013) state that through the use of pms the nature of jobs has changed, going from job-specific tasks to a broader attention to the firm’s organizational goals. This positively influences self-efficacy. Boswell et al (2008, p. 363) state that “there is a paradigmatic from employees performing narrowly defined job duties to an exception that employees understand ‘the big picture’ and help contribute as needed to the attainment of firm goals”. In order to encompass organizations goals, management must communicate the core

(25)

25

values of the company. In order to instill collective pride and establish the importance of responsibility an organization relies on a beliefs system.

Based on the aforementioned findings, the following is hypothesized.

The use of the four levers of control will have an effect on the self-efficacy of employees

H3a. The use of positive msc (beliefs and interactive control systems) will have a positive

effect on self-efficacy

H3b. The use of negative msc (boundary and diagnostic control system) will have a negative

effect on self-efficacy

According to numerous studies transformational leaders have a positive influence on the self-efficacy of employees (Prozacha et al., 2017; Kirk et al., 2002; Chen & Bliese, 2003). Aforementioned, Bass & Avolio (1994) argues that transformational leadership consists of three behavioral components. These are: individualized consideration (individual coaching and mentoring), intellectual stimulation (enhancing creativity and innovation) and

charisma/inspirational motivation (inspire and excite in solving problems). In order to enhance self-efficacy of followers it is expected that transformational leaders prefer using positive msc (beliefs and interactive mechanisms) in order to demonstrate these specific behavioral characteristics (Bass, 1991; Simons, 1995).

Based on the aforementioned set of hypotheses this research explores the effect of transformational leadership directly and through the use of mcs on the self-efficacy of employees.

(26)

26

The design of this research including it hypotheses is depicted in the figure below.

Figure 2: Conceptual model

Based upon the conceptual model this research examines the influence of transformational leadership on self-efficacy. It is hypothesized that transformational leadership has a positive influence on self-efficacy. As a mediating factor it is hypothesized that managers tend to have an preference to positive msc over negative mcs in order to affect change in their employees. Further, it is derived from literature that the use of either negative or positive mcs by

(27)

27

3. Data and Method

This chapter explains and underpins the methodology of this research. First the research design is addressed, followed by the data collection and lastly the data analysis is clarified. This chapter represents the start of the empirical study of this thesis by applying research methods to gather results in order to answer the research question.

3.1 Research design

Because of the type of research question - what is the effect of transformational leadership on

the SE of employees through the use of mcs?- the best research method is a quantitative study.

Based on the research onion by Saunders et al (2008) a roadmap is given into the different layers of this research process.

This thesis tries to explain how transformational leadership influences the self-efficacy of employees through mcs. This creates a particular set of circumstances and individuals coming together that creates a unique type of social interaction (Saunders et al, 2008). These social actors are unique to their environment. In this case Rabobank Amsterdam contributes to a unique set of managers and subordinates working together to achieve organizational goals.

The research approach is deductive as it tries to explain the causal relationship between the variables in this study. This explanation is based on various stages, which starts with defining the research question. Based on theory the hypotheses are derived. Based on the hypotheses, data is collected and analyzed to confirm or reject the hypotheses. Because of the expected causal relationship between the variables the research is therefore based on

explanatory research.

Collecting data via an electronically spread survey is needed to answer the research question. According to Saunders et al (2008) a survey is “a research strategy which involves

(28)

28

the structured collection of data from a sizeable population”. The survey will take the form of a questionnaire. The questionnaire has been made in Qualtrics and sent electronically to participants, in this case the employees of Rabobank Amsterdam via e-mail. Participation of respondents is voluntary. The sampling technique used in this study is a non-probability sampling.

The time horizon of this study is cross-sectional because the collection of data from participants is from one period in time. This is often called a snapshot. After the respondents have filled the questionnaire an additional follow-up will not occur. A reminder email was sent to all respondents.

3.2 Data Collection

To test the hypothesis the survey was conducted at Rabobank Amsterdam. This specific bank has about 450 employees. This random sample contains respondents from different

departments, functions and levels of education. Because of the perceived self-efficacy based on the transformational leadership style of managers, all employees have been included in the research. The questionnaire was setup in a way which will applicable to managers as non-managers. This will give the study a broader population instead of limiting it to only

managers or non-managers. The study is limited to the Rabobank in Amsterdam. The reason for this is the fact that there is no objection from higher management to conduct the study there. With the sample size being 450 employees the minimal respondents of 100 can be achieved to draw valid conclusions based on the questionnaire (Field, 2013). Also, all e-mail addresses of the employees working there are available to send an email containing the link to the online survey. This will make reaching the respondents faster and easier. Also, conducting the study in one company enables the control of the settings of relevant factors such as

strategy, external environment, technology, structure etc. Meaning that these factors are the same to all respondents.

(29)

29

The online survey was sent on the 23th of December 2017 to all 450 respondents by e-mail. The survey was available until the 12th of January 2018. On the 2nd January a reminder e-mail was sent to the population to participate in the survey. During this period 144

respondents started the questionnaire but only 109 respondents (N=109) completed the survey. See appendix A for the questionnaire sent to the respondents.

Independent variable – Transformational leadership

The independent variable in this study is leadership style. The IV has one variable which is transformational leadership. Transformational leadership is measured in different studies which used questionnaires to identify this specific leadership style (Nguyen et al, 2016). According to the study of Nguyen et al (2016) a seven-point Likert model (1 being not at all and 7 being frequently, if not always) of 8 questions was used on 152 respondents to

determine the extent of transformational leadership style. An example question is “indicate to which extend the following leadership behaviors represent your superior, makes subordinates feel proud of the group”. Based on the Cronbach’s α (0.871) the internal reliability was tested. Based on the statistics for average variance extracted the validity was tested (0.726).

Dependent variable – Self-efficacy

To test the SE in this research the website of the International Personality Item Pool (IPIP) is used. The IPIP is website where a number of personality scales are available. The scales are based on reliability and validity. There is an existing IPIP scale “self-efficacy”. The

Cronbachs α is 0.78. There are six questions and four reversed questions in this scale. A reverse coding indicates a relatively low score on those questions which indicate a high level of self-efficacy.

Mediating variable – msc (management control systems)

The mediating variable is based on four variables being the belief system, interactive system, boundary system and diagnostic system. The measurement of these variables is based on the study of Widener (2007). To measure these variables Widener used a seven-point Likert scale

(30)

30

on 122 respondents (these respondents were all CFO’s). The scales are ranging from 1 (strongly disagree) to 7 (strongly agree). The internal reliability of these variables was above the desired 0.75 (Cronbach’s α > 0.75). This makes the research used by Widener suitable for this study. The validity of the study was examined extensively, based on factor analysis and a lack of significant cross-loadings.

Control Variables

In this study five control variables are added. These are gender, age, tenure, function, department and education. To measure if transformational leaders tend to use positive over negative mcs it is necessary to examine the leadership behaviors in various departments within Rabobank Amsterdam. By examining the way these departments are led by managers it will help establish a relationship between the examined variables in this study. The internal and external factors in the organization are similar for all employees and managers. As aforementioned managers have freedom at Rabobank Amsterdam in how to lead their subordinates, this makes it interesting to find a relation between their leadership style and their use of mcs and how this affects the self-efficacy of employees. Rabobank Amsterdam has over 20 job profiles which makes it difficult to specify every job profile in the

questionnaire.

Likert scale

The Likert scale used in this study is a 6 point Likert scale. The reason for this is restricting respondents to choose a middle option, meaning to remove the choice of respondents as not to commit to a choice (Lietz, 2009). Another reason for using a 6 point Likert scale instead of an 5 point Likert scale is that the expected value of the validity coefficient increases with 0.04 for each extra option (Lietz, 2009).

(31)

31

4. Results

In order to test the hypotheses from chapter two, statistical analysis was performed on the collected data. These analyses include a reliability, correlation, regression and mediation analysis. This chapter showcases the results of these tests.

Data Analysis

The data was exported from Qualtrics to SPSS. Based on the completed surveys (N=109) it is assumed that the data is normally distributed (Field, 2013). To confirm the assumption of the data being normally distributed a kurtosis and skewness test was conducted. The kurtosis test is a check on the sharpness of the peak of the distribution and the skewness test checks whether the distribution is asymmetrical (Field, 2013). Before testing the data for normality and reliability the counter-indicative questions were transformed. Also, two groups were computed of the variables beliefs system and interactive system, this is named positive mcs and the other two variables (boundary and diagnostics system) were grouped as negative mcs. For all variables, the scale means were computed and the data recoded into Self-Eff, Beliefs, ICS, Boundary, Diagnost, Trans, Pos_MCS and Neg_MCS.

Descriptive Statistics

A frequency test was conducted in order to determine the normality of the data. Below in the table is the descriptive statistics of each variable.

Table 2: Descriptive statistics

Variables N Min Max Skew Kurt

1. Self-Eff 106 3.10 6.00 -0.57 0.47 2. Beliefs 108 1.25 6.00 -0.61 0.14 3. ICS 109 1.50 5.50 0.02 0.22 4. Boundary 108 2.00 6.00 -1.03 1.63 5. Diagnost 106 2.36 6.00 -0.49 0.37 6. Trans 106 2.88 6.00 -0.65 0.40 7. Pos_MCS 109 1.70 5.20 -0.13 0.01 8. Neg_MCS 108 2.93 6.00 -0.48 0.58

**. Correlation is significant at the 0.01 level (2-tailed) *. Correlation is significant at the 0.05 level (2-tailed)

(32)

32

As the skewness and kurtosis remain between to -2 and 2, no mutations of the data were deemed necessary. Looking at the variable Boundary through the frequency distribution there are more positively answered questions which results in a high sharpness of the peak or leptokurtic distribution. As the variable does not exceed the -2 and 2 boundaries, this is acceptable.

Reliability Analysis

After testing the normality of the data was subjected to a reliability analyses. The reliability is determined by the Cronbach’s alpha for each variable. A variable is considered reliable if the Cronbach α > 0.7 (Field, 2013).

Table 3: Reliability Analysis

Variables Cronbach α 1. Self-Eff 0.852 2. Beliefs 0.878 3. ICS 0.659 4. Boundary 0.862 5. Diagnost 0.928 6. Trans 0.918 7. Pos_MCS 0.721 8. Neg_MCS 0.894

Based on table 3 all variables are considered reliable but one, that is ICS. Here the α= 0.66, which is below the margin of 0.7. During the analysis removal of questions from the scale was considered. No removal of questions would significantly change the reliability score of the scale. Although on its own the ICS has a low reliability the combined Cronbach’s alpha of the Pos_MCS is above the desired 0.7. Therefore, no items are deleted form the construct ICS.

Correlation

After the reliability analysis was performed the data was subjected to a correlation analysis. A correlation analysis indicates whether there is a presence of significant correlation amongst the constructs. See the results of the correlation test below.

(33)

33

Table 4: Means, Standard Deviations, Correlations

Variables M SD 1 2 3 4 5 6 7 8 1. Self-Eff 4.85 0.703 -0.85 2. Beliefs 3.97 0.974 0.12 -0.88 3. ICS 3.37 0.719 0.022 0.152 -0.66 4. Boundary 4.73 0.732 .394**_.417** 0.024 -0.86 5. Diagnost 4.54 0.762 .243*_.213* 0.188 0.119 -0.93 6. Trans 4.72 0.703 .307**_.300**_{0.068 .298}**_.674** -0.92 7. Pos_MCS 3.61 0.622 0.091 .729**_.788**_.277**_.263**_.235* -0.72 8. Neg_MCS 4.6 0.61 .35** _.32** _{0.17 .430}**_.948**_.709**_.319**_-0.84

**. Correlation is significant at the 0.01 level (2-tailed) *. Correlation is significant at the 0.05 level (2-tailed)

Based on the correlation analysis it seems that there is a significant correlation between Self-Eff and the other variables such as Beliefs, ICS, Boundary, Diagnost, Trans, Pos_MCS and Neg_MCS. Also, looking at table 4 there seems to be multicollinearity between Neg_MCS and Diagnost (correlation well above .8 (Field, 2013)). This also applies for Pos_MCS and Beliefs and ICS. As Neg_MCS and Pos_MCS are made up of the four levers (Diagnost and Boundary for Neg_MCS and Beliefs and ICS for Pos_MCS) these show a high

intercorrelation. Although not high enough for multicollinearity, it is notable that Trans correlates strongly with Neg_MCS. Further investigation on the VIF statistics showed there is no multicollinearity because the VIF values are well below 10 (Field, 2013).

Preliminary Regression

To start the analysis a multiple regression was performed to investigate the direct effect of transformational leadership on efficacy and mcs and the direct effect of mcs on self-efficacy.

(34)

34

Table 5: Regression Analysis

Model R R ² B SE t 1 .310a _0.096 Trans .276** 0.083 3.323 2 .310b _0.096 Trans .276** 0.086 3.223 Pos_MCS -0.001 0.094 -0.006 3 .361c _0.13 Trans 0.115 0.117 0.983 Pos_MCS -0.047 0.095 -0.493 Neg_MCS .275** 0.138 2.002

In the first model, one predictor was entered: transformational leadership. This model was statistically significant F (1, 104) = 11.04; p < .05. R2= 0,096

After entry of Pos_MCS the total variance explained by the model as a whole remained 9.6% F (2, 103) = 5.47; p < .001. This stayed the same, there was no additional % of variance in self-efficacy after controlling for transformational leadership. F (2, 103) = 0.00; p < .001).

In model 3 Neg_MCS was introduced. The entry of this variable lead to the rise of the R2 to 13%. F (3, 102) = 5.09; p < .001. Adding Neg_MCS changed the predictive value of the model but also the effect of each of the predictor values. Transformational leadership is no longer significant, and Neg_MCS has a significant positive effect.

The hypotheses were tested by doing a mediation analysis of the direct and indirect

relationship of transformational leadership on self-efficacy through positive and negative mcs. This analysis was done using the PROCESS macro in SPSS 24.

Mediation

To establish if there is a relationship between the variables a mediation analysis was done. Although Neg_MCS and Pos_MCS are correlated, these variables have been treated as non-serial mediators. In PROCESS model 4 was used to determine the effect of each mediator separately. In each model one of the MCS mediators was tested, and the other mediator was

(35)

35

added as a covariate. This was done because of the positive correlation between Pos_MCS en Neg_MCS. The addition made it possible to use model 4 to separately test the two mediation effect.

Table 6: Mediation Analysis Pos_MCS as mediator

Consequent

M (Pos_MCS) Y (Self_Eff)

Antecedent Coeff SE P Coeff SE P

X (Trans) -0.012 .115 .919 0.115 .117 .344 M (Pos_MCS) -.047 .086 .588 Neg_MCS 0.350 .145 < .05 0.275 .128 < .05 Constant 2.047 .403 < .01 3.217 .685 <.001 R2 = 0.107 R2 = 0.130 F(2,103) =7.31, P<.01 F(2,103) = 2.5399, P< .1

(36)

36

Table 7 Mediation Analysis Neg_MCS as mediator

Consequent

M (Neg_MCS) Y (Self_Eff)

Antecedent Coeff SE P Coeff SE P

X (Trans) 0.588 .060 <.001 0.115 .118 .334 M (Neg_MCS) 0.275 .128 < .05 Pos_MCS 0.169 .052 < .01 -0.047 .086 .588 Constant 1.221 .324 < .05 3.217 .491 < .01 R2 = 0.533 R2 = 0.130 F(3,102) =58.75, P<.001 F(3,102) = 2.539, P<.1

In the tables above the results of the mediation analysis for Pos_MCS and Neg_MCS are presented.

The results show a non-significant effect of Trans on Self Eff. (B= .115, p= .344). This means that Hypothesis 1 is rejected.

Table 6 shows no significant positive effect of Trans on Pos_MCS (B= -,012, p= .919) Thus hypothesis 2a is rejected. For Neg_MCS the effect is also significant, but positive instead of negative (B= .588, p< .01). Therefore hypothesis 2b also has to be rejected.

Hypothesis 3a is rejected due to the fact that there is no significant relationship between Pos_MCS and Self Eff (B= -.047, p=0.588). The relationship between Neg_MCS and self-efficacy is significant, but the effect is contrary to the expected value. Hypothesis 3b (B= .275 p< .05) is rejected.

(37)

37

In the preliminary tests a significant effect of transformational leadership on self-efficacy was expected. Covariance of the two mediators negatively influenced this

relationship. When not controlling for the Neg_MCS in the mediation model a significant effect on Self-Eff is observed (B=.275, p< .05). Based on above findings an overview of the outcome is presented in figure 3 below.

Figure 3: Statistical model

Additional analysis

Based on the above analysis there is no significant relation, both direct and indirect, on the variables examined in this study hypothesis none of the hypothesis are supported. Although for hypotheses 2b and 3b statistically significant relationships are found, these are contrary to the expected results. This means that transformational leadership does not directly contribute to the self-efficacy of employees at Rabobank. Although hypothesis H2b and H3b are not supported, it seems that self-efficacy increases when managers use negative mcs. This is contrary to the findings in literature.

In the preliminary analysis, an OLS regression was done. This analysis showed a low Durbin Watson score indicating positive autocorrelation in the data. The scatterplot of

(38)

38

residuals and predicted values showed no discernable patterns that would indicate heteroskedacity or other disturbing patterns and the analysis was conducted with the full sample.

Because of the findings form the analysis, the data was examined again and divided based on subgroups in respondents. There are three groups of employees in the dataset: Corporate, Private and Other (eg staff, HR). The dataset was split and the data was analyzed again. The OLS was repeated to find differences in the results. The results of this can be found in table 8.

Table 8: Regression Analysis

("Corporate") Model R R ² R ² Change Durbin-Watson B SE β t 1 .158a _.025 _.025 Trans .158 .125 .158 1.257 2 .172b .030 .005 Trans .155 .126 .155 1.229 Pos_MCS .065 .119 .069 0.547 3 .272c .074 .044 1.318 Trans -.001 .155 -.001 -.009 Pos_MCS .018 .121 .019 .152 Neg_MCS .278 .164 .268 1.695

a. Predictors: (Constant), Trans

b. Predictors: (Constant), Trans, Pos_MCS

c. Predictors: (Constant), Trans, Pos_MCS, Neg_MCS

Table 9: Regression Analysis ("Private & Other")

Model R R ² R ² Change

Durbin-Watson B SE β t 1 .474a _.225 _.255 Trans .389 .114 .474 3.408 2 .490b _.240 _.015 Trans .429*** .123 .524 3.487 Pos_MCS -.151 .170 -.133 -0.888 3 .510c _.260 _.020 _2.030 Trans .272*** .198 .331 1.371 Pos_MCS -.182 .173 -.161 -1.056 Neg_MCS .261 .258 .248 1.014

a. Predictors: (Constant), Trans

b. Predictors: (Constant), Trans, Pos_MCS

(39)

39

The analysis shows the expected results for the group “Private and Other” but that results for the group “Corporate” indicates a strong deviation. None of the relationships between the dependent and independent variables are significant. Also, the low Durbin Watson score compared to the normal score for the group “Private and Other” indicates that the responses from the Corporate division are the source of the disturbance in the main sample. The regression analysis for the “Private and Other” group show a positive relationship between transformational leadership and self-efficacy that gets diminished when Pos_MCS and Neg_MCS get added to the model. These results are discussed further in Chapter 5.

The sample size of these groups (Corporate N=64, Private & Other N=42) does not merit further mediation analysis as these results would have a high probability to show false positive or negative results.

(40)

40

5. Discussion

The main goal of this study was to examine the relationship between transformational

leadership and self-efficacy of employees and the mediating role management control systems play (divided in positive and negative mcs). A quantitative study was conducted at Rabobank Amsterdam, were in total 109 respondents participated (N=109). Three hypotheses (including sub hypotheses) were formulated in order to answer the following research question:

What is the effect of transformational leadership on the self-efficacy (SE) of employees

through the use of management control systems (mcs)?

In this chapter the findings of this study are discussed and how these findings contribute to existing literature, followed by the implications and limitations.

5.1 Discussion on results of statistical analysis

The findings in this study mostly reject the formulated hypotheses. Below, the results of the data and the findings of each hypotheses based on theory are discussed.

The first hypothesis, H1, dove into the relationship between transformational leadership and self-efficacy of employees. Empirical results proposed a weak but positive effect (Prozacha et al, 2017). The result of this study partially confirm this hypothesis. Looking at the initial mediation analysis it seemed that there is no statistically relationship between transformation leadership and self-efficacy. But after splitting the respondents into their respective segments hypothesis H1 for the segment “Private & Other” is significant whereas the hypothesis for the segment “Corporate” is rejected. This finding of a positive relation is consistent with the study of Prozacha et al (2017). Although the overall hypothesis is rejected one can clearly see a discrepancy in how these segments view transformational leadership and self-efficacy. An explanation can be that respondents working in the segment

(41)

41

“Corporate” enjoy more autonomy in carrying out their job hence arguing that managers marginally contribute to their effectiveness, whereas the respondents working at the segment “Private & Other” are more driven by managers based on data from Management

Performance systems.

The second hypothesis, H2a and H2b, tried to explain the relationship between transformational leadership and mcs. Based on empirical studies it is assumed that transformational leaders tend to show a preference for use of positive mcs (Bass, 1991; Simons, 1995). Hypothesis H2a is rejected because of severe lack of statistical relevance, meaning no relationship between transformational leaders and their preference for use of positive mcs. Hypothesis H2b is also rejected because of lack of findings that

transformational leaders prefer to use less negative mcs. Results however showed quite the opposite of empirical findings. It seems that there is a relationship between transformational leadership and their use for negative mcs. An explanation can be that the setting of Rabobank, being a machine bureaucracy (Mintzberg, 1979) where the focus lies on compliance through the formal line of hierarchy (Chenhall, 2003). Thus trying to shift management attention to other much needed issues which makes the use for diagnostic and boundary control systems explainable. This seems to support the findings of Simons (1995) were the Levers of Control are used effectively by managers in order to focus on innovation and creativity without sacrificing control.

The last hypotheses, H3a and H3b, tries to explain the relationship between mcs and self-efficacy. H3a hypothesizes that the use of positive mcs will increase self-efficacy.

Analysis of the results show no statistical relevance to empirical findings of Kaplan & Norton (1996) and Burney & Widener (2013). Where they argue that performance management systems are positively related to self-efficacy. This study has found some evidence that negative mcs increases self-efficacy.

(42)

42

Simons (1995) states that diagnostic control systems are comparable to performance management systems. Spms are used by diagnostic and interactive control systems but the purpose for use of smps for these levers are different from each other and therefore serve other purposes (see chapters 2.3 and 2.5.2). Thus finding relevance to the studies of Kaplan & Norton (1996) and Burney and Widener (2013). Another preference for negative mcs can be found in the fact that most of the respondents have a commercial function, in which

performance management and diagnostic control play an important role (Simons, 1995). This would explain the positive effect of the Neg_MCS on self-efficacy. The expected relation between transformational leadership and the use of diagnostic control systems was expected to be negative, hence hypotheses H3a and H3b are both rejected.

5.2 Implications

As described in the above section of this study the majority of the hypotheses are rejected. Although this was the case, there are several contributions to literature that can be implied. As mentioned earlier there seems to be a positive relation between transformational leadership and the self-efficacy of employees. This however was not the case for all respondents, but the segment “Private & Other” showed that transformational leadership does indeed enhance self-efficacy. Which seems to contribute to the study of Prozacha et al (2017). This contradiction, where one group experience self-efficacy through transformational leadership, offers

interesting perspectives for further research, which will be discussed further on. There also seems to be a relation between transformational leadership and use of mcs. This contributes to the study by Simons (1995) where he discusses that mcs are used by management in order to keep focus on innovation and creativity without sacrificing control, thus achieving

organizational goals.

The findings of this study are also useful for management of Rabobank Amsterdam in order to help understand the relation between leadership type and self-efficacy of their

(43)

43

employees and use of mcs. There seems to be a discrepancy between the two segments (Corporate and Private & Other). In order to enhance organizational goals of Rabobank Amsterdam further notice can be given into using boundaries, and specifically diagnostic control systems, as employees positively react to this, and to dive further in the reasons for the discrepancy between the two groups.

5.3 Limitations

As every other study, this study has its limitations. Starting with the sample size of the study. Even though there were 109 respondents, looking into the segments where employees work, there were too few respondents to dive into the phenomena of what caused one group to experience more transformational leadership than the other.

There seems to be a high drop-out rate of respondents (32%). Respondents tend to stop the survey after having filled in the control variables/questions, which were one of the first questions. A lager sample size implies more data on the examined variables, which would benefit the study.

The sampling method of this study is also a limitation. This was a convenience sample as the researcher of this study works at Rabobank Amsterdam hence suggesting a possible bias effect of on the data of this research. This was mitigated as the sample size was large enough for analysis, although not sufficient for the analysis of discrepancy between groups.

Another limitation is the place where this study has been performed. The study was done at Rabobank Amsterdam. Thus, making it difficult to determine the generalizability of the data and therefore the results to other organizations in the Netherlands or even

organizations in other countries. Further research should focus more on different

(44)

44

banking sector can perceive self-efficacy through transformational leaders differently from employees in the health sector.

The effect of transformational leadership on self-efficacy of employees : the influence of management control systems