Two dimensions of formalization as the drivers of managerial ambidexterity

1

The Two Dimensions of Formalization as the Drivers of

Managerial Ambidexterity

Master Thesis

Student: Karine Makaryan / Student № 11417889 MSc. Business Administration, Strategy track

University of Amsterdam, Faculty of Economics and Business Supervisor: MSc. B. Silveira Barbosa Correia Lima

University of Amsterdam, Amsterdam Business School Date: 23 – 06 – 2017

2 Statement of originality

This document is written by Student Karine Makaryan 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 Table of Contents

Abstract ... 4

Introduction ... 5

Theory and Hypotheses... 11

Managerial Ambidexterity ... 11

Two Dimensions of Formalization ... 14

The Degree of Formalization ... 15

Types of Formalization: Coercive vs. Enabling ... 17

Methodology ... 27

Sample and data collection ... 27

Measurements... 28 Statistical model ... 35 Results ... 37 Main effects ... 37 Moderating effect ... 38 Control variables ... 38 Discussion ... 41 Findings ... 41

Limitations and future research ... 44

Conclusion ... 47

4 Abstract

Previous research addressed that ambidexterity yields positive firm performance (He & Wong, 2004; Tushman, Smith, Wood, Westerman & O’Reilly, 2010; Tempelaar & Van de Vrande, 2012). While ambidexterity at the firm level of analysis is often studied, managerial ambidexterity is still underdeveloped (Raisch & Birkinshaw, 2008). However, in order to become an ambidextrous firm, ambidextrous managers are required (O’Reilly & Tushman, 2004). The current research – cross-sectional survey design – was conducted to explore what drives variation in managers’ ambidexterity. Formalization is a key element that drives managerial behaviour (Martinez & Jarillo, 1989; Van De Ven et al., 1976). This indicates that formalization is a vital contextual factor for managerial ambidexterity. Two dimensions of formalization are distinguished in this study: degree and type of formalization. The degree of formalization stresses the question whether there is a process while the type of formalization emphasizes how that bureaucracy should look like (Adler & Borys, 1996). Due to the different design and effect on managerial ambidexterity it is crucial to examine both dimensions. However, no evidence was found, in this study, that either one of the dimensions has a significant relation with managerial ambidexterity. Also, the moderating effect of the type of formalization shows no significant effect in this study. These results will be discussed in this research. Altogether, this study advances the field of research on managerial ambidexterity by theoretically and empirically contributing to a better understanding about what drives variation in managers’ ambidexterity.

Key words: Managerial ambidexterity; degree of formalization; type of formalization; enabling formalization; coercive formalization; behavioural theory of managers

5 Introduction

Due to today’s dynamic business environment, organizations are ought to act ambidextrously. Therefore, ambidexterity is developing as a crucial area of scholarly theory and empirical investigation (Turner et al., 2013). Ambidextrous firms are aligned and efficient in their management of today’s business demands while simultaneously adaptive to changes in the environment (Duncan, 1976; Gibson & Birkinshaw, 2004; Tushman & O’Reilly, 1996). Organizational ambidexterity is positively associated with sales growth, innovation and firm survival (He & Wong, 2004; Tushman, Smith, Wood, Westerman & O’Reilly, 2010; Tempelaar & Van de Vrande, 2012). Hence, previous research has shown that organizational ambidexterity is positively related to a firm’s performance(He & Wong, 2004; Raisch & Birkinshaw, 2008; Stettner & Lavie, 2013). This effect is more valuable under environmental uncertainty and when a firm has sufficient resources (Jansen et al., 2005; Jansen, Vera & Crossan, 2009). This is often the case with larger rather than smaller firms (Yu & Khessina, 2012). These studies have documented the effects of ambidexterity at a firm and business unit level.

While there is numerous researchdone at business unit and firm level, seemingly less research focused on the individual level. However, every organizational phenomenal requires analysing the individual as the core factor within the organization for a better understanding of the organization (Felin & Foss, 2005). Without individuals there would be no organization. Felin and Foss (2005, p. 452) argue that: “Individuals after all provide the nested antecedent to numerous collective phenomena and thus deserve careful theoretical and empirical consideration in our theorizing.” Gibson and Birkinshaw (2004, p. 209) introduced ambidexterity at the individual level, defined as “the behavioral capacity to simultaneously demonstrate alignment and adaptability across an entire business unit.” They found that both alignment and adaptability were higher at successful units rather than less successful units.

6 The individual level of ambidexterity is a promising research direction for future research (Raisch & Birkinshaw, 2008). In addition, many scholars have asked for a better understanding of the conditions under which individual ambidexterity leads to success (Lavie, Stettner & Tushman, 2010; Raisch et al., 2009). Further, the focus on the individual level will increase understanding of the roots and context of ambidexterity. The most basic level understanding on ambidexterity will provide organizations with knowledge on how to encourage and promote ambidextrous behaviour (Felin & Foss, 2005). Following up on the request for a better understanding on individual ambidexterity, this research will focus on managerial ambidexterity. Managerial ambidexterity is defined as “a managers’ behavioural orientation toward combining exploration and exploitation related activities within a certain period of time” (Mom et al, 2009, p.812). In addition, ambidextrous managers have the following characteristics: (1) they encounter contradictions; (2) they are multitaskers; (3) and they renew and refine their skills, expertise and knowledge (Mom et al., 2009). The relevance of investigating managerial ambidexterity is addressed by O’Reilly and Tushman (2004) who identified the fact that ambidextrous organizations need ambidextrous managers. Hence, in order for a firm to achieve ambidexterity a firm needs ambidextrous managers (O’Reilly & Tushman, 2004, p.81). Therefore, this paper will address this gap in the literature and contribute to a better understanding about the drivers of managerial ambidexterity.

Previous research has shown valuable findings on managerial ambidexterity. O’Reilly and Tushman (2011) have shown that managers commonly focus on short term profit at the expense of uncertain long-term returns. Additionally, they provide examples of organizations, such as IBM and Cisco, who managed to be successfully ambidextrous. In each of the success cases, a clear and identifiable manager, showed ambidextrous behaviour by resolving conflicts and simultaneously making resource allocation decisions (O’Reilly & Tushman, 2011). So, the managers were excellent multitaskers, which is a crucial characteristic of an

7 ambidextrous manager (Mom et al., 2009).

Integrating different and conflicting set of tasks is the current challenge of managers in the dynamic work environment. These activities require mechanisms for coordination. Formal structural coordination mechanisms such as ‘formalization’ are one of the most crucial mechanisms for coordination (Martinez & Jarillo, 1989). “Formalization of a managers’ tasks refers to the degree to which rules and codes describe a particular task; provide guides for decision making; and provide guides for conveying decisions, instructions, and information and the degree to which the manager has to conform to the task description” (Mom et al., 2009, p. 814). Previous research has confirmed the importance of coordination mechanisms as key elements that drive managerial behaviour (Martinez & Jarillo, 1989; Van De Ven et al., 1976).

In line with this, Mom et al. (2009) studied the relation between formal structural coordination mechanism and managerial ambidexterity. The formal structural coordination mechanism contains (1) managers decision-making authority and (2) formalization of a managers’ task. However, they did not find the expected negative relationship between the formalization of a managers’ task and managerial ambidexterity. Also, Jansen et al. (2006) studied the relationship between formalization and explorative/exploitative innovation. The expected positive relationship between formalization and exploitative innovation is confirmed in their research, yet their findings did not provide support for a negative relation between exploratory innovation and formalization. Despite the different level of analysis and dependent variables of these two studies mentioned above, both failed to clarify the negative relation with formalization. Thus, the proper degree of formalization which enhances managerial ambidexterity remains absent.

Accordingly, a better understanding is essential to clarify what the mechanisms are for achieving managerial ambidexterity. Previous research which tried to interpret the relation

8 between formalization and ambidexterity did not distinguish various types of formalization. Adler and Borys (1996) distinguished two types of formalization: formalization designed to enable employees to master their tasks, and formalization designed to coerce effort and compliance from employees. There are four generic features that distinguish enabling and coercive types of formalization: repair, internal transparency, global transparency, and flexibility (Adler & Borys, 1996). Despite the fact that the enabling type of formalization includes processes for effecting change, it has often been neglected in previous research.

Numerous studies have shown that the enabling type of formalization reduces role conflict and ambiguity, thereby increasing work satisfaction and reducing feelings of alienation and stress (Jackson & Schuler, 1985; Podsakoff, Williams & Todor, 1986). In addition, the enabling type of formalization is dynamic rather than static and facilitates change. These characteristics will support managers at hosting contradictions and refining and renewing their knowledge. Therefore, this will empower managerial ambidexterity. On the other hand, several studies showed that the coercive type of formalization is positively related to absences, propensity to leave, physical and psychological stress, feelings of powerlessness and self-estrangement and negatively associated with innovation and job satisfaction (Rousseau, 1978; Kakabadse, 1986; Arches, 1991). Due to the fact that these characteristics inhibit a managers’ ability to host contradiction, be a multi-tasker, and refine and renew their expertise, the coercive type of formalization will decline managerial ambidexterity. This suggests that formalization has positive and negative effects on managerial ambidexterity depending on the type of formalization.

Thus, it is important to consider both types of formalization since they have different effects on managerial ambidexterity. While the coercive type, the traditional view of formalization, is expected to have a negative effect on managerial ambidexterity, the enabling type of formalization will most likely have a positive effect on managerial ambidexterity.

9 Hence, this research yields to examine the relation of two dimensions of formalization (degree and type) on managerial ambidexterity. With these new insights more sufficient understanding will be gained about the drivers of managerial ambidexterity.

Taken together, the abovementioned statements raise the following research question:

RQ: To what extent do the two dimensions of formalization, degree and type, have an effect on managerial ambidexterity?

This research aims to have several contributions. Firstly, clarifying the picture of how managerial ambidexterity can be achieved. Thus, help develop ambidexterity at the managers’ level of analysis, since this is still underdeveloped (Gupta et al. 2006, Raisch and Birkinshaw 2008). Secondly, this study will develop hypotheses on multiple direct effects; degree of formalization and type of formalization, and the moderating effect of the type of formalization. Furthermore, the crucial distinction will be made between the different types of formalization; enabling and coercive (Adler & Borys, 1996). These important insights will contribute to a better understanding of the drivers of managerial ambidexterity. This knowledge might support managers at becoming ambidextrous, which will consequently lead to higher firm performance (He & Wong, 2004; Felin & Foss, 2005; Raisch & Birkinshaw, 2008; Stettner & Lavie, 2013). Thirdly, this paper will examine multiple future research suggestions of recent studies (Mom et al., 2009; Mom et al., 2015; Raisch & Birkinshaw, 2008). Finally, by clarifying important antecedents of managerial ambidexterity, managers and organizations can recognize these aspects and alter their behaviour more towards ambidexterity.

In the next section, the concept and drivers of managerial ambidexterity (two dimensions of formalization) will be expanded. Followed by associated hypotheses. The methodology section provides analysis about the sample, data collection and the development and validation of the measurement instruments. To test the hypotheses, data will be gathered

10 through a questionnaire. The survey will be conducted among managers of several firms. Then, the empirical findings will be illustrated in the results. Next, the findings of the results will be presented and the implications and suggestions for future research will be given. Finally, an overall conclusion will be presented.

11 Theory and Hypotheses

The following section contains main insights regarding the existing literature on managerial ambidexterity in relation to formalization and presents the hypotheses of this study. First, managerial ambidexterity is discussed in more detail. Hereby will be discussed: how and why the contextual factor, formalization, is influencing a managers’ ability to become ambidextrous. Followed by introducing the two dimensions; degree and type of formalization. Next, the impact of the degree of formalization on managerial ambidexterity will be addressed and the first hypothesis will be presented. Despite the extent of formalization, the type of formalization, the second dimension, is also expected to influence a managers’ ambidexterity. Therefore, the direct and moderating effect of the type of formalization are discussed, and the rest of the hypotheses are presented next.

Managerial Ambidexterity

A managers’ primary role includes dissemination and gathering of information, making managers the centre for information flows within the company (Mintzberg, 1973). Notwithstanding that this is common for all managers, the information content of these roles differs by hierarchical level, functional area, and business unit or department (Walsh, 1988). Despite that, currently managers are ought to obtain and process contradicting streams of information (Mom et al., 2009). Also, nowadays, managers face inconsistent behavioural expectations regarding the development of current competences and strategies, and the need to enforce new competencies and strategies (Floyd & Lane, 2000). O’Reilly and Tushman (2004) declare that one of the most crucial aspects for an organization to become ambidextrous is having ambidextrous managers – executives who are able to understand the needs of distinctive kinds of businesses. Moreover, Mom et al. (2009) refer to managerial ambidexterity as a managers’ behavioural orientation toward combining exploration and

12 exploitation related activities within a certain period of time. Exploitation is associated with efficiency, focus, convergent thinking, and reducing variance; while exploration is associated with experimentation, flexibility, divergent thinking, and increasing variance.

Furthermore, ambidextrous managers are ought to effectively manage strategic contradiction (Mom et al., 2009). Preceding, Barnard (1968, p.21) stated: “It is precisely the function of the executive to facilitate the synthesis in concrete action of the contradictory forces, to reconcile the concrete forces, instincts, interests, conditions, positions, and ideals”. Thus, an ambidextrous manager should possess mental balancing and attend processes of the past while simultaneously preparing innovations that will define the future. Also, previous research demonstrates that inconsistent activities can both succeed at the same time (Smith & Tushman, 2005). Effectively managing these contradictions is dependent on a managers’ ability to recognize and embrace the contractions (Smith & Tushman, 2005).

Subsequently, ambidextrous managers should contain the ability to fulfil multiple roles simultaneously (Floyd & Lane, 2000). So, they are multi-taskers who will wear more than one hat. This also indicates that an ambidextrous manager is a generalist rather than a specialist. Further, ambidextrous managers will seek out favourable changes beyond the limits of their own jobs. They are the ones taking initiatives and reaching out to others trying to combine their efforts. Being a mediator is another feature of an ambidextrous manager, always searching to build internal linkages (Birkinshaw & Gibson, 2004).

As mentioned earlier, managers are the nexus for knowledge flow within an organization. However ambidextrous managers are ought to retrieve different kinds of knowledge. They act creatively as well as collectively; generating high quality and in-depth knowledge while integrating this into collective action (Sheremata, 2000). This characteristics needs to be altered regarding the current circumstances of the managers’ job. This implies the urge for both refining and renewing their knowledge, skills, and expertise

13 (Mom et al., 2009).

All the above mentioned arguments sum three main characteristics of ambidextrous managers; (1) ambidextrous managers host contradictions, (2) ambidextrous managers are multitaskers, and (3) ambidextrous managers both refine and renew their knowledge, skills, and expertise (Mom et al., 2009).

It is important for managers to achieve these characteristics, because several researches addressed that, in order to stimulate ambidexterity in an organization, the most important source is enforcing ambidextrous managers (Bartlett & Ghoshal, 1986; Tushman & O’Reilly, 1996; Gibson & Birkinshaw, 2004). As mentioned earlier, when ambidexterity is present in an organization this will improve a firm’s performance (He & Wong, 2004). Gibson and Birkinshaw (2004), argue that the level of performance will increase with high levels of ambidexterity. In addition, achieving ambidexterity within a firm is one of the most critical aspects for firm’s dynamic capabilities (Eisenhardt & Martin, 2000; Teece, Pisano & Shuen, 1997). By investigating the paradox for managers being involved in exploration and exploitation activities simultaneously, we “move beyond oversimplified and polarized notions to recognize the complexity, diversity, and ambiguity of organizational life” (Cameron & Quinn, 1988).

Ambidextrous managers are expected to; increase efficiency while also fostering creativity and create change, think globally while acting locally, monitor their employees and report to their senior executives (Lewis, 2000). For managers, to act ambidextrous, thus manage these paradoxes, it is important that there is supportive context (Gibson & Birkinshaw, 2004). When managers seek to leverage participation, increase efficiency, trust, and commitment they elaborate the complicating nature of formalization (Lewis, 2000). Formalization of a managers’ task endures tensions between control and flexibility (Hatch, 1993; Sitkin & Bies, 1993). Coordination mechanisms, such as formalization, are one of the

14 key drivers of managerial behaviour (Martinez & Jarillo, 1989; Van De Ven et al., 1976). So, in order to advance the field of research, it is important to examine what the right degree and type of formalization is to be the supportive context that managers need in order to become ambidextrous. The impact of formalization of a managers’ task will be discussed in more detail later in this study.

Two Dimensions of Formalization

As mentioned above, formalization is a contextual factor which impacts a managers’ ability to become ambidextrous. The right surrounding and set of circumstances will support managerial ambidexterity (Gibson & Birkinshaw, 2004). The degree of formalization, the first dimension, is referred to as: “The proportion of jobs that are codified and range of variation allowed within jobs. The higher the proportion of codified jobs and the less the range of variation allowed, the more formalized the organization” (Hage, 1965).

While previous research mainly focused on this dimension (degree) of formalization, it is important to distinguish among a second dimension of formalization; the type of formalization. The first dimension stresses the question whether or not there is a process while the second dimension emphasizes how that bureaucracy should look like (Adler & Borys, 1996). Distinguishing amidst types of formalization creates the possibility to theorize about ‘good’ and ‘bad’ rules and procedures as experiences by managers. Adler and Borys (1996) introduced this two-dimensional view. This view implies that the effect of formalization depends both on the degree of formalization and the type of formalization. Mom et al. (2009) studied the relationship between formalization of managers’ task and managerial ambidexterity, however, neglected the second dimension, the type of formalization. This could be a possible reason why they did not find the expected relation between formalization and managerial ambidexterity.

15 So, the impact of the degree of formalization on managerial ambidexterity will be discussed next. Followed by the second dimension, the type of formalization. This part will address the questions why we need to distinguish among types of formalization and what the influence of this distinction is on managerial ambidexterity.

The Degree of Formalization

The first dimension, influencing managerial ambidexterity, is the degree of formalization. Pugh et al. (1963, p. 303) state the following regarding the formalization of a managers’ task: “Formalization includes (1) statements of procedures, rules, roles (including contracts, agreements, and so on), and (2) operation of procedures, which deal with (a) decision seeking (applications for capital, employment, and so on), (b) conveying of decisions and instructions (plans, minutes, requisitions, and so on), and (c) conveying of information, including feedback”. Subsequently, the extent to which variation is allowed and the magnitude of codification within jobs define the degree of formalization (Hage, 1965).

A high degree of formalization results in the development of expertise in a limited area (Hage, 1965). Due to high degree of written and filed procedures, less room maintains for managers’ knowledge acquisition outside their current expertise (Pugh et al., 1963; Mom et al., 2007). However, ambidextrous managers are generalists rather than specialist and therefore are required to consume knowledge outside their expertise.

Also, managers need to find a balance in persuading subordinates to fulfil the demands of the company while building a positive rapport (Pace, 2003). Blau and Scott (1962) refer to this strain between complexity and formalization, as the dilemma of the professional and the bureaucracy. In order to deal with this dilemma and complexity, managers need access to flexibility (Pace, 2003; Hage, 1965). Therefore, the higher the degree of complexity of a certain task, the lower the degree of formalization should be, since

16 low degree of formalization allows more flexibility compared to high degree of formalization (Hage, 1965). Hence, since hosting contractions, thus complex tasks, is a characteristic of an ambidextrous manager, low formalization is expected to encourage managerial ambidexterity.

Hage (1965) also suggests that high degree of formalization is associated with high efficiency. Considering that high formalization results in expertise in a limited area this will enforce efficiency and reduce errors being made (Hage, 1965). Efficiency is, however, associated with exploitation rather than managerial ambidexterity. Despite the fact that efficiency is necessary to achieve ambidexterity, fully focussing on efficiency and neglecting the importance of flexibility and innovation will hinder managers at becoming ambidextrous (Adler et al., 1999; Gibson & Birkinshaw, 2004). Thus, ambidextrous managers increase efficiency while simultaneously creating innovative opportunities for the future.

While flexibility and innovative activities are essential for ambidextrous managers, previous research indicates that high degree of formalization permits innovation and openness in the system, contributes to routinized activities and singleness of purpose, and for that reason is negatively related to initiations of becoming an ambidextrous manager (Knight, 1932; Stepherd, 1945; Pierce & Delbecq, 1977). Consequently, this will also negatively influence a managers’ ability to fulfil multiple roles simultaneously and their capability of pursuing favourable changes beyond the limits of their own jobs.

Thus, in order to have the ability to host contradictions, formalization of a managers’ task should be low (Mom et al., 2009). Low formalization is characterized by "the adjustment and continual redefinition of individual tasks" (Argyris, 1972, p. 184). Higher degree of formalization, in contrast, lowers adaptiveness and assertiveness of an individual (Hage, 1965; Miller, 1987). As mentioned earlier, ambidextrous managers both refine and renew their knowledge, skills, and expertise and therefore will benefit from low degree of

17 formalization of a managers’ task. All the arguments stated above suggest the following hypothesis.

Hypothesis 1. A high degree of formalization of managers’ tasks has a negative effect on managerial ambidexterity.

Types of Formalization: Coercive vs. Enabling

Despite distinguishing low and high degree of formalization, it is also important to distinguish different types of formalization to completely understand its effect on managerial ambidexterity. The behaviour of managers is often formalized to reduce variability and ultimately to predict and control their behaviour. This is realized by using mechanisms called standardization of work processes (Mintzberg, 1979). While previous research mainly focused on this impact of different degrees of formalization, less attention is paid to different types of formalization. Previous studies discuss two contrasting views of formalization; enabling and coercive formalization. Adler & Borys (1996) argue that these two types of formalization have very different outcomes. Burns and Stalker (1961) also distinguish two types of bureaucracy; mechanistic and organic. The mechanistic model is described as “the precise definition of rights and obligations and technical methods attached to each functional role” while the organic model is characterized by “the adjustment and continual redefinition of individual tasks” (Burns and Stalker, 1961). Since enabling formalization supports rather than controls a manager, this suits the organic rather than the mechanistic model. Controversially, coercive formalization has a controlling rather than supporting function, which fits the mechanistic model. The content is different concerning these two models. While the content of information is instruction and order at the mechanistic model, the content of information at the organic model is information and advice. So, due to the different

18 outcomes and content of information it is important to understand both types of formalization in order to examine its effect on managerial ambidexterity.

The central feature of formalization - the degree of written procedures, rules, and instructions – is Weber’s ideal type (Mintzberg, 1979). However, Weber’s analysis resides with ambiguity (Parsons, 1947). He determined two very diverse sources of authority in bureaucracies: "incumbency in a legally defined office" and "the exercise of control on the basis of knowledge" (Weber, 1947). Building on that, Adler and Borys (1996) identified two contrasting views of formalization: the negative (coercive) and positive (enabling) view. Managers’ attitude towards formalization depends on the characteristics of the type of formalization which they are confronted with (Adler & Borys, 1996). According to the negative view, coercive type of formalization, formalization demotivates managers, restrains creativity and supports dissatisfaction. According to the positive view, enabling type of formalization, it ensures guidance and clarity, thereby helping managers to feel more effective and releasing stress (Adler & Borys, 1996). Hence, for this research two generic types of formalization will be distinguished: “formalization designed to enable employees to master their tasks, and formalization designed to coerce effort and compliance from managers” (Adler & Borys, 1996, p. 62). Formalization of a managers’ task is either coercive or enabling. The next sections explain each type in more detail.

Coercive formalization. Adler (1999) identified several features that characterize the structure of coercive formalization. Firstly, the focus on performance standards is attendant to highlight poor performance. Then, this type is trying to minimize game-playing and monitor costs. Next, it functions to keep managers out of the control loop. Finally, they are instructions to be followed, not changed. This type of formalization is the more familiar one. If bureaucracy always served for this purpose – assure that irresponsible managers do the right thing – then efficiency would trade off motivation and creativity (Adler, 1999).

19 Enabling formalization. Unlike the coercive type of formalization, the enabling type of formalization serves the purpose of support rather than control. Enabling formalization procedures are designed with the participation of the managers in the interest of identifying best practices and opportunities for growth and improvement (Alder, 1999; Hoy, 2000). The enabling type of formalization holds several features that define its structure. Firstly, the focus is set on best practices. Information on best practices is crucial for achievement of such performances. The second feature also concerns best practices; formalization systems are best practice templates to be improved. Thirdly, flexibility in these systems is beneficial. So, customization and improvisation on different levels of skills and experience is feasible. Finally, enabling formalization systems should offer managers the ability to control their own work rather than being controlled by the system (Adler, 1999). Accordingly, enabling formalization assists managers in such a manner that it helps rather than hinders them while doing their job (Hoy, 2000). Previous research also confirmed that enabling formalization improves work satisfaction (Michaels, Cron, Dubinsky & Joachimsthaler, 1988), increases innovation (Craig, 1995), helps managers to become effective (Adler & Borys, 1996), reduces role conflict and stress (Senatra, 1980; Adler & Borys, 1996), and provides guidance and clarity for the managers’ responsibilities (Adler & Borys, 1996). Hence, enabling formalization procedures are “helpful and lead to problem solving among members rather than rigid, coercive activities that demand conformity” (Hoy, 2000, p.530). Table 1 summarizes the characteristics of both coercive and enabling formalization.

Table 1

Coercive Versus Enabling Features of Structure (Adler, 1999, p. 44)

20 Systems focus on performance standards so as to

highlight poor performance.

Focus on best practice methods: information on performance poor performance is not much use without information on best practices for achieving them.

Standardize the systems to minimize gameplaying and monitoring costs.

Systems should allow customization to different levels of skill/experience and should guide flexible improvisation.

Systems should be designed so as to keep managers out of the control loop.

Systems should help people control their own work.

Systems are instructions to be followed, not challenged.

Systems are best practice templates to be improved. According to Adler & Borys (1996), the two types of formalization could be separated based on three dimensions: (1) the features of enabling and coercive formalization, (2) the design process, and (3) the implementation process. Regarding the first dimension, four design principles are used to distinguish features between the two types of formalization: repair, internal transparency, global transparency, and flexibility. The second dimension concerns whether individuals should be involved in the design process of formalization or not. Lastly, the third dimension involves the implementation of this formalization process (Adler & Borys, 1996). However, this research will only discuss the first dimension (the features of enabling and coercive formalization).

Features of enabling and coercive formalization. As mentioned above four design features – repair, internal transparency, global transparency, and flexibility – distinguish coercive and enabling formalization. The following section identifies these features.

Repair. This feature concerns the ease with which managers repair a process themselves instead of allowing the disruption to end the process. When a manager is capable of repairing easily and agile, errors will be corrected with little effort. The question that should be asked here is “to which extent are organizational members allowed to solve breakdowns and continue without further interruptions to their activities?” (Jørgensen &

21 Messner, 2009, p. 102). Thus, when the ease allowance of repairing processes is high, the degree of formalization of a managers’ task is low.

In regard to the coercive type of formalization any deviation from the standard procedure is seen as possibly harmful. The processes are designed to control managers rather than support them. Hence, it will be less likely for managers to shirk. So, the repair tasks are not designed to help the process operate well, nor to navigate a managers’ task, nor to help a manager identify future opportunities. The procedures are discussed with the managers, however, the managers are not in the position to discuss back (Adler & Borys, 1996). Also, repair tasks are separated from routine tasks. Since repair tasks concern unexpected activities they are part of non-routine tasks. Whenever an unexpected breakdown occurs, little attention is paid to managers who have suggestions for improvement. Ultimately, this all results in covert and inefficient work behaviours (Adler & Borys, 1996). This behaviour will hinder ambidextrous managers at refining and renewing their knowledge and activities. By reason of not being able to navigate and change their own tasks and no access to flexibility, the coercion logic will obstruct ambidextrous managers (Adler & Borys, 1996; Mom et al., 2009).

While in the coercion logic managers do not have any input regarding repair efforts, in the enabling logic managers are allowed to repair the processes themselves. In contract to the enabling logic, these repair efforts are recognized as an attempt for improvement (Adler & Borys, 1996). So, instead of viewing these breakdowns as a negative aspect they become opportunities for growth. Diversions from prescribed managerial tasks are considered as need for more training or inadequate prescribed methods (Adler & Borys, 1996). Unlike repair in the coercion logic, repair in the enabling logic is part of the routine task of managers. Jørgensen and Messner (2009) argue that repair is better executed by managers rather than other employees. Pursuing improvement and growth is in line with an ambidextrous manager

22 who is constantly refining and renewing his/her knowledge and expertise. Therefore, the feature repair, viewed from an enabling point of view, will enforce managers to act ambidextrous (Adler & Borys, 1996; Mom et al., 2009).

Internal transparency. The next feature is internal transparency. Regarding internal transparency, Adler & Borys (1996, p.72) state: “When equipment is designed to reduce reliance on users' skills, there is little reason to provide users with any visibility into its internal workings.” In other words, when an unexpected breakdown appears, there will be no information provided to the managers about this. Information is only given to the person who is responsible for the repair. This enforces specialists rather than generalists, as managers. So, internal transparency includes visibility and understanding of equipment. Since coercive formalization consists of rules and procedures to be followed and not changed, internal transparency is not a necessary factor in order for managers to execute their task. This, however, inhibits a managers’ ability of being the nexus of knowledge within an organization (Mintzberg, 1973). For a manager to become ambidextrous is it crucial to focus on a certain task while simultaneously experimenting for future opportunities (Mom et al., 2007; Mom et al., 2009). According to the coercion logic no information is provided to the manager which consequently inhibits a managers’ ability to become ambidextrous and pursuit innovation that defines growth and improvement (Adler & Borys, 1965; Smith & Tushman, 2005).

In contrast, regarding the enabling logic; visibility and clarity in the processes that regulates a managers’ task is key. A deeper understanding of these processes is also important since the reliance on managers’ skills is high. Further, internal transparency will provide feedback on a managers’ performance by comparing their performance with previous standards (Adler & Borys, 1996). This will contribute to a further development of managerial tasks. Enabling formalization consist of rules and procedures which are open for change and

23 improvement. For this reason, internal transparency is vital for a manager to execute his/her tasks. This view of internal transparency supports a managers’ ability in being an ambidextrous manager and nexus of diverse knowledge flows within an organization (Mintzberg, 1973). Transparency enables deeper understanding and less information asymmetry (Adler & Borys, 1965) which will support a manager at handling contradicting tasks and thereby explant their ambidextrous behaviour (He & Wong, 2004; Mom et al., 2009).

Global transparency. While internal transparency refers to the internal functioning of the procedure as used by managers; global transparency refers to the consciousness of managers for a broader whole within which they are working (Adler & Borys, 1996). In the logic of coercive formalization this indicates that managers only have access to specific information concerning the task that they are responsible for at that moment. Managers are not allowed to move from their specific task. Thus understanding of the broader system is not required. However, an ambidextrous manager is a generalist rather than a specialist (Birkinshaw & Gibson, 2004). This expresses the increase of the negative relation of coercive formalization towards managerial ambidexterity.

By contrast, concerning the enabling formalization logic, a broad range of information is perceived as favourable. The information provided from a wide range will help managers to interact with the whole organization and environment. Thus, global transparency is required to broaden managers’ understanding of all processes. This will guide them to understand the position of their managerial task within the organization as a whole (Adler & Borys, 1996). In order to balance explorative and exploitative tasks simultaneously, this understanding is essential. Being able to understand and act on this paradox, which ambidextrous managers are ought to do, acquiring and renewing of knowledge is a critical aspect (Lewis, 2000; Mom et al., 2007).

24 Flexibility. The last feature that distinguishes coercive and enabling formalization is flexibility. Flexible systems encourage managers to adjust the activity and add functionality to suit their work demands. This feature illustrates a managers’ power to adjust the system (Jørgensen & Messner, 2009). The coercive procedure implies a detailed description for the managers’ task with no authority for deviation (Adler & Borys, 1996). Thus, the manager implements the tasks precisely as handed to him or her. However, the lack of flexibility inhibits managers to be creative and innovative, which are, among others, necessary features of an ambidextrous manager (Mom et al., 2009).

Contrasting, the enabling formalization logic considers deviation as opportunities for development rather than risk. Standardized methods are there to provide suggestions and advice rather than control. Managers have the opportunity to choose whether they retain control over themselves. The enabling type of formalization encourages managers to change and add functionality to suit their tasks (Adler & Borys, 1996). This flexibility of enabling formalization encourages managers to act ambidextrous.

Another way to distinguish coercive and enabling formalization is their ability of deskilling and usability approaches. While the coercive type of formalization is deskilling a managers’ approach, the enabling type of formalization strives for usability (Adler & Borys, 1996). Usability is the capability of leveraging a managers’ skill instead of replacing it (Salzman, 1992). So, whenever systems or managers themselves break down, the mental models gained from the usability approach enables managers to regain control. Contrasting, in the deskilling approach, the managers’ mental models are redundant since managers are expected to precisely follow instructions (Adler & Borys, 1996). This demonstrates the relation between the types of formalization and managerial ambidexterity. The enabling type of formalization, which handles a usability approach, will encourage managers to become ambidextrous. The managers gain control and possess flexibility which permits them to act

25 creatively as well as collectively (Sheremata, 2000). The coercive type of formalization, which maintains a deskilling approach, inhibits managers at becoming ambidextrous. Since they have no control and need to follow strict procedures, they are not able operate explorative and exploitative characteristics, such as efficiency and experimentation, simultaneously (March, 1991; Gibson & Birkinshaw, 2004).

In sum, high degree of formalization of a managers’ task has a negative effect on managerial ambidexterity because, among others, it permits innovation and openness in the system (Pierce & Delbecq, 1977). The coercive type of formalization will enforce this negative effect on managerial ambidexterity since the purpose of this type of formalization, among others, is to control and provide instructions rather than support and development. Moreover, coercive formalization is used to punish and hinder managers. It also discourages interactions among managers which obstructs growth and innovation (Adler & Borys, 1996; Adler, 1999; Hoy, 2000). Hence, coercive formalization causes more problems than it solves (Hoy, 2000). This means that the negative effect of the degree of formalization on managerial ambidexterity is reinforced by the coercive type formalization. These arguments result in the following hypothesis:

Hypothesis 2. The coercive type formalization moderates the negative effect of a high degree of formalization on managerial ambidexterity.

All in all, when a managers’ task contains an enabling type of formalization, rather than a coercive type of formalization, this enables authentic communications and provides opportunities for growth and development. Besides that, these types of procedures are guides to solutions rather than strict and permanent rules. So, they are not steady but modified to meet new goals and solutions to problems. When the rules and procedures are hindering, they are set aside to encourage progress. In addition, these types of procedures facilitate learning

26 and help rather than hinder managers at doing their job (Adler & Borys, 1996; Adler 1999; Hoy, 2000). Therefore, these characteristics of enabling formalization will support a manager at hosting contradictions, being a multi-tasker, and at refining and renewing their knowledge, skills, and expertise (Mom et al., 2009). The abovementioned claims suggest the following hypothesis:

Hypothesis 3. The enabling type of formalization has a positive effect on managerial ambidexterity.

The line of reasoning discussed earlier suggest that, notwithstanding a low or high degree of formalization, the enabling type of formalization will positively affect managerial ambidexterity. Therefore, the negative effect of the degree of formalization on managerial ambidexterity will be decreased due to the enabling nature of formalization. This raises the following hypothesis:

Hypothesis 4. The enabling type of formalization moderates the negative effect of a high degree of formalization on managerial ambidexterity.

27 Methodology

In the method section the design and research approach of this study will be discussed. Firstly, the sample and data collection will be explained. This includes which strategies and techniques are most appropriate to gather information and respondents. Secondly, the measures and validation are clarified. Thus, this section defines which scales are used to measure the different variables. In addition, the Exploratory Factor Analysis (EFA) for the dependent and independent variables will be discussed in this section. Furthermore, all measures were drawn from extant research. Finally, the final section reveals how raw data was compiled and analyzed.

Sample and data collection

In order to answer the research question and hypotheses, this study followed quantitative techniques. A cross-sectional survey design was used to gather data for the explanatory variables and answer the hypotheses. Previous researches, studying the antecedents of individual ambidexterity, also conducted a survey for gathering data (e.g., Mom et al., 2009; Mom et al., 2015). The survey was conducted online, using qualtrics. Using survey design makes it possible to test hypotheses concerning cause and effect. Since this is quantifiable this allows the findings of correlations. Additionally, surveys are generally viewed as one of the more reliable methods of data collection (https://revisesociology.com/2016/01/11/social-surveys-advantages-and-disadvantages/). Moreover, since the population is large and the sampling frame is unknown, this study was examined using a non-probability convenience sampling method. Managers from different manufacturing and service industries were asked to fill in the survey. Considering that managers are a critical aspect for an organization to become ambidextrous and that they are expected to fulfil conflicting and inconsistent tasks, this study investigated managers on

28 becoming ambidextrous (Floyd & Lane, 2000; O’Reilly & Tushman, 2004). The managers represent a wide variety in terms of organizational and functional tenure, hierarchical level, gender, educational level, and age, which minimizes threats to external validity. The names and identities of the managers were not revealed, to ensure confidentiality. The average age of the managers is 36 years, the average job tenure within the firm is 6,5 years, and the average span of control is 331 employees. Since list-wise deletion is one of the most common techniques of handling missing data, it was used to clean the data in this study (Peugh & Enders, 2004). A total of 157 employees completed and returned the survey of which 106 managers. The further analyses only consist of managers.

Measurements

Dependent variable. The dependent variable for this research is managerial ambidexterity. Most scales for ambidexterity are scales of firm or business unit level which are constructed by combining exploitation and exploration measures (Gibson & Birkinshaw, 2004; He & Wong, 2004). For the managerial level of analysis, Mom et al., (2009) followed this practice and developed exploitation and exploration measures at the manager level of analysis. This study used their approach and examined the extent to which managers engaged in exploration and exploitation activities simultaneously during the last year (Mom et al., 2009). Thus, managerial ambidexterity was measured by the 14 item seven-point Likert scale (Mom, Van Den Bosch & Volberda, 2007; Mom et al., 2009). Seven items for exploration (α=.79) and seven items for exploitation (α=.75). A multiplication of exploration times exploitation generated the variable for managerial ambidexterity. Managers were requested to indicate “To what extent did you, last year, engage in work related activities that can be characterized as follows” (1 = to a very small extent to 7 = to a very large extent). Exploration items include: “Searching for new possibilities with respect to products/services, processes,

29 or markets,” “Activities of which the associated yields or costs are currently unclear,” and “Activities requiring quite some adaptability of you”. Exploitation items include: “Activities which you can properly conduct by using your present knowledge,” “Activities which serve existing (internal) customers with existing services/products,” and “Activities of which a lot of experience has been accumulated by yourself.” Hence, managers’ ambidexterity was assessed by computing the multiplicative interaction between managers’ exploration activities and managers’ exploitation activities.

An exploratory factor analysis (EFA) was conducted to check for convergent and discriminant validity. The Kaiser–Meyer–Olkin measure verified the sampling adequacy for the analysis, KMO = .742. Bartlett’s test of sphericity χ² (91) = 788.843, p < .001, indicated that correlations between items were sufficiently large for EFA. Exploratory factor analysis (table 2) with Varimax rotation with all 14 items, showed two factors were retained after rotation; one exploration scale with the seven exploration items (Eigenvalue = 2.6) and one exploitation scale with the seven exploitation items (Eigenvalue = 3.8). The first item of exploitation shows high cross-loading on the factor of exploration as well, this could be due to the content of the item (activities of which a lot of experience has been accumulated by yourself). Since delating this item does not contribute to vital improvement on the factor analysis or regression analysis, I decided not to exclude this item. Regarding the rest of the items, no item cross-loading was greater than 0.21.

Table 2

Factor Analysis for Managerial Ambidexterity (Mom et al., 2009)

Factorᵇ

Itemsª

To what extent did you engage in work related activities that can be

characterized as follows: 1 2

An employee's exploration activities (α=.79)

30

Evaluating diverse options with respect to products/services, processes, or markets 0.72 0.06

Focusing on strong renewal of products/services or processes 0.62 0.12

Activities of which the associated yields or costs are currently unclear 0.73 -0.07

Activities requiring quite some adaptability of you 0.61 0.17

Activities requiring you to learn new skills or knowledge 0.51 0.11

Activities that are not (yet) clearly existing company policy 0.69 -0.20

An employee's exploration activities (α=.75)

Activities of which a lot of experience has been accumulated by yourself 0.41 0.49

Activities which you carry out as if it were routine 0.10 0.67

Activities which serve existing (internal) customers with existing services/products 0.21 0.42

Activities of which it is clear to you how to conduct them 0.00 0.79

Activities primarily focused on achieving short-term goals 0.14 0.61

Activities which you can properly conduct by using your present knowledge 0.00 0.78

Activities which clearly fit into existing company policy -0.07 0.67

ª Items are quoted from survey. All items were measured on a seven-point scale (1=to a very small extent to 7=to a very large extent).

ᵇ Extraction Method: principal component analysis. Rotation Method: Varimax with Kaiser normalization.

Independent and moderating variables. This study contains two independent variables, also referred as the two dimensions of formalization: the degree of formalization of a managers’ task and the type of formalization. The degree of formalization was measured using a four item five-point Likert scale from Deshpande and Zaltman (1982). The measurement reveals to which extent a managers’ task is defined by regulations, rules, or procedures (α=.78). The items were: “Whatever situation arises, I have procedures to follow in dealing with it,” “I have to follow strict operational procedures at all times,” “Rules occupy a central place in my work related activities,” and “There is a written job description for going about my tasks.” All items were scored on a 5-point scale, ranging from "to a very small extent" to "to a very large extent.” Furthermore, Hoy (2000) developed a measurement for enabling formalization. Initially, they tested for two independent factors: enabling formalization and enabling centralization. However, their principal-axis factor analysis

31 suggested that the two independent factors were strongly related. Hence, this indicates that whenever the rules are enabling so is the hierarchy and vice versa. Also, the original scale was intended for teachers, however, for this research it is applied for managers. The type of formalization was measured by the 21 item five-point Likert scale (1 = never to 5 = always) of Hoy (2000). 12 items for enabling formalization (α=.87) and nine items for coercive formalization (α=.80). Enabling type of formalization items include: “Administrative rules in this company provide opportunities for professional growth and development,” “Administrative rules in this company help me solve problems,” and “The administrative hierarchy of this company facilitates learning.” The coercive type of formalization measured whether the rules and procedures are hindering rather than constructive. Items include: “Administrative rules in this company are hindering,” “Administrative rules in this company discourage positive interactions among colleagues,” and “Administrative rules in this company are used to punish employees.” Due to counter-indicative items, all nine items of coercive formalization were recoded into different variables. Thus, a high score on this variable indicates that the type of formalization is enabling rather than coercive and vice versa. So, while the (enabling) type of formalization has a (positive) direct effect on managerial ambidexterity, this variable also moderates the relation between the degree of formalization and managerial ambidexterity.

To test for convergent and discriminant validity, among coercive and enabling formalization and the degree of formalization, an exploratory factor analysis (EFA) was conducted. The Kaiser–Meyer–Olkin measure verified the sampling adequacy for the analysis, KMO = .794. Bartlett’s test of sphericity χ² (300) = 1318.95, p < .001, indicated that correlations between items were sufficiently large for EFA. Table 3 shows the factor loadings after rotation. Varimax rotation was used with all 21 items. Three factors retained after rotation; one for the degree of formalization (Eigenvalue = 5.8), one for enabling

32 formalization (Eigenvalue = 3.5), and one for coercive formalization (Eigenvalue = 2.6). One item shows high cross-loading on the other factors (Administrative rules in this company are set aside when they hinder progress). Deleting these items, however, does not add enough improvement in the factor analysis or regression analysis. Therefore, no items were deleted. The rest of the items do not show cross-loading greater than 0.27.

Table 3

Factor Analysis for the Degree of formalization and the Type of Formalization (Deshpande & Zaltman, 1982; Hoy, 2000)

Factorᵇ Itemsª

For each item, please answer as it applies to you:

1 2 3

The degree of formalization (α=.78)

There is a written job description for going about my tasks -0.07 0.05 0.65

Rules occupy a central place in my work related activities

-0.08 0.00 0.85

I have to follow strict operational procedures at all times

-0.06 -0.12 0.82 Whatever situation arises, I have procedures to follow in dealing with it

0.02 -0.03 0.80 Please rate to which extent each item describes behavior in the company you are currently working:

Enabling formalization (α=.87)

Administrative rules in this company enable authentic communication between me and my colleagues

0.47 -0.06 -0.03

Administrative rules in this company provide opportunities for professional growth and development

0.69 -0.06 -0.2

Administrative rules help rather than hinder 0.68 -0.14 -0.18

Administrative rules in this company are guides to solutions rather than strict procedures 0.73 -0.07 0.12 Administrative rules in this company are guides to solutions rather than rigid procedures 0.68 -0.08 0.00

Administrative rules in this company are modified to meet new goals 0.63 0.10 0.00

Administrative rules in this company suggest solutions to problems 0.70 0.09 -0.25

Administrative rules in this company help me solve problems 0.77 0.01 -0.20

Administrative rules in this company are set aside when they hinder progress 0.39 -0.30 0.20

The administrative hierarchy of this company enables me to do my job 0.54 -0.08 0.11

The administrative hierarchy of this company facilitates the mission of the company 0.62 -0.17 0.03

33

Coercive formalization (α=.80)

Administrative rules in this company are used to punish employees 0.03 0.60 0.02

Administrative rules in this company are hindering 0.02 0.55 -0.05

Administrative rules in this company discourage positive interactions among colleagues -0.10 0.63 -0.13

The administrative hierarchy of this company obstructs innovation -0.19 0.73 -0.06

The administrative hierarchy of this company causes more problems than it solves -0.41 0.61 -0.02 Administrative rules in this company are substitutes for professional judgment 0.27 0.40 -0.14

The administrative hierarchy obstructs employee achievement 0.02 0.79 0.00

In this company authority is used to undermine employees -0.15 0.64 0.12

My boss’ way is the only way to do the job -0.17 0.51 0.07

ª Items are quoted from survey. All items were measured on a seven-point scale (1=to a very small extent to 7=to a very large extent).

ᵇ Extraction Method: principal component analysis. Rotation Method: Varimax with Kaiser normalization.

Control variables. Work experience can be defined as “the number of months spent in a particular job” (Quinones et al., 1995, p. 888). Experience is the most significant predictor of behavior, compared to other demographical characteristics (Taras et al., 2010). In addition, work experience is one of the dominant drivers of managerial behavior (Sturman, 2003; Taras, Kirkman & Steel, 2010). Along with that, work experience affects managers’ motivation and repertoire of skills and competences (Beier & Ackerman, 2001; Collins, Smith & Stevens, 2001). Therefore, it is crucial to control for experience. Age and tenure measure the experience of a manager. Organizational tenure, the length of service within an organization, is the most frequent studied type of work experience (Quinones et al., 1995). The length of service within the firm will help managers become a more knowledgeable individual regarding knowledge about their firm as a whole (Datta, Guthrie & Wright, 2005). Organizational related experience is different from functional related experience. Functional tenure is specialized and narrow which is not beneficial for managerial ambidexterity. While organization tenure has a positive effect on managerial ambidexterity, functional tenure has as a negative relationship with managerial ambidexterity (Mom et al., 2015). So, due to the different effects it is also important to control for functional tenure. Moreover, as the size of

34 the firm might increase the positive effect of ambidexterity on firm performance (Yu & Khessina, 2012), I will control for firm size. Also, managers at larger firms have more resources. Therefore they can allocate those resources on both exploration and exploitation instead of one of the activities (March, 1991; Lewin et al., 1999). The variable was measured by asking the number of employees working in their establishment (Audia & Greve, 2006). For this variable, dummy variables were made: small, medium, and large firms. Large firms were the reference group. Further, to control for educational effects, dummy variables were included: one for managers with degrees below bachelor level, another for managers with a bachelor degree, and the reference group were managers with a master’s degree of higher (Mom et al, 2009). Higher educational levels are positively related to managerial ambidexterity since high levels of education are associated with the ability to process information and learning (Papadakis et al., 1998; Adler et al., 1999). Due to the fact that higher level managers are expected to act more ambidextrous than lower level managers (Floyd & Lane, 2000), I control for hierarchical level. Dummy variables were constructed for high, medium and lower level managers. High level managers were the reference group. Also, managers might differ across functional areas regarding ambidextrous activities (Duncan 1976). Therefore, I created dummy variables to control for functional area. Research and development (R&D), marketing and sales (M&S), operations are the three dummy variables (Mom et al., 2009). Here, M&S functioned as the reference group. A multiple choice question was used to ask this question: “Which of these functional areas is most suitable to yours?” Finally, I controlled for environmental dynamism. Jansen et al. (2009) argue that environmental dynamism is a factor which may influence a managers’ ability to engage in exploration or exploitation activities or both. To measure this item a five item (α=.69), seven point Likert scale (1 = strongly disagree to 7 = strongly agree) by Dill (1958) was used. Sample items are: “Environmental changes in our local market are intense” and “In

35 our local market, changes are taking place continuously.”

All respondents were ensured that all their answers would be kept in the strictest confidentiality. In addition, at the end of the questionnaire the participants were thanked and assured that their input has been extremely valuable. If desired, they had the opportunity to ask questions or leave comments in the provided space in the questionnaire.

Statistical model

As mentioned earlier, data for this research was collected by conducting a survey among managers. Before testing the hypotheses some preliminary steps were necessary. Firstly, the data was screened. All abovementioned variables were tested for missing values by means of a frequency test. Managerial ambidexterity had 0.9% of missing values. Further, the amount for degree of formalization was 4.7%. Finally, the type of formalization had 14.2% of missing values. To check for outliers, standardized values were used. Also, a normality check was conducted to control the skewness and kurtosis. The items of coercive formalization were all recoded due to counter-indicative items. Also, a reliability analysis was conducted for all the variables. The Cronbach’s alpha, which is a measure of internal consistency, tests whether the set of items are closely related as a group. All the items have a Cronbach’s alpha >.7, which indicates a high level of internal consistency (Cronbach, 1951). Moreover, in order to test the hypotheses, new variables were created with the use of existing items. These scale means provide the mean for all the items mentioned earlier. Finally, a correlation matrix, which is a table of correlation coefficients of all variables, was provided by SPSS. After the preliminary steps, a hierarchical multiple regression analysis was performed to investigate the proportion of variance of the degree of formalization, type of formalization, and the control variables, explained by the model. Regarding the moderating effect of the type of formalization on the relationship between the degree of formalization and

36 managerial ambidexterity, I mean-centered the independent and moderating variables before analyzing the effect. By doing so, any potential multicollinearity problems are mitigated (Lim et al., 2015). In order to test the hypotheses on the moderating effects, I used model 1 of PROCESS on SPSS (Hayes, 2013).

37 Results

This section contains the results of this study. To start, the descriptive statistics for all the variables, used in this study, are presented in order to view the data. Next, the correlation analysis shows the correlations among variables. It is reported whether the correlation is significant or not. Table 4 shows the descriptive statistics and correlations for all the variables used in this research. Finally, the hierarchal regression analysis is performed to test the hypotheses of this research. The hierarchical regression analysis is demonstrated in table 5. Model 1 presents the control variables, model 2 includes the main effects, and model 3 also includes the moderating effect.

Main effects

Model 2 in table 5 shows the main effects concerning hypothesis 1 and 3. A positive coefficient for degree of formalization means that high degree of formalization will affect managerial ambidexterity positively, whereas a negative coefficient means that it will negatively affect managerial ambidexterity. Hypothesis 1 assumes that the degree of formalization will have a negative effect on managerial ambidexterity. When considering the coefficients of model 2, the main effect of degree of formalization on managerial ambidexterity is positive and not significant (β = 0.16, p > 0.05). Hence, hypothesis 1 is not supported. Further, a positive coefficient on the type of formalization means that enabling formalization will affect managerial ambidexterity positively, whilst a negative coefficient means that enabling formalization will negatively affect formalization. Model 1 also presents the effect of the type of formalization on managerial ambidexterity (β = 0.08, p > 0.05). So, the effect of the type of formalization on managerial ambidexterity is positive but not significant. As mentioned, a positive relation means that enabling formalization will enforce