“The influence of the organizational structure on the network of management accountants within a company”

(1)

“The influence of the organizational structure on the network of

management accountants within a company”

ABSTRACT

University of Groningen

Faculty of Economics and Business

MSc. Business Administration

Specialization: Organizational and Management Control

Joost van Burgsteden

s2045168

Supervisor: W.A. Moossdorff MSc

Second supervisor: Dr. B. Crom

Words: 24.250

Groningen, 23-06-2014

Keywords:

Social network theory, Organizational structure, Bean counter, Business partner

Abstract

This research focused on how the organizational structure influences the network of the bean counter and business partner within a company. It makes use of two case studies,

which were conducted at the manufacturing organization Philips Drachten and the UMCG hospital. Using the social network theory and the classification of the organizational structure as either mechanistic or organic, this study found that two elements of the organizational structure do have an influence on these networks. The first of these elements is the communication channel management accountants use in order to communicate with other actors, as this allowed management accountants at both organizations to have a wider array of contacts. The second element consists of emphasis of work practices, which was found to put restrictions on the network of the

(2)

Master thesis University of Groningen Page 1

1.Introduction

The role of management accountants has changed throughout the years (Yazdifar and Tsamenyi, 2005; Burns and Baldvinsdottir, 2005; Lambert and Sponem, 2012). According to Paulsson (2012), the role of management accountants has moved from that of passive producers of financial information towards a role in which they are more proactive; they are involved in decision-making and the focus is not solely on financial information anymore. This can be classified as a move from a bean counter role towards a business partner role. According to Granlund and Lukka (1997), the main differences between these two types of management accountants is that the bean counter is mainly focused on the past, is not involved in the operations of the company, and is expected to produce mainly financial reports. This is in contrast with the business partner, who is focused mainly on the present and future and is involved in the operations and decision-making processes of the company (next to producing financial reports).

Prior research on the roles of management accountants is mainly on the individual level (Coad, 1999; Byrne and Pierce, 2007). However, these papers do not focus on the networks of both types of management accountants. Windeck et al. (2013) found that because of the new business partner role of management accountants, frequent interaction and cooperation between management accountants and managers arises. However, I was unable to identify any research paper which focuses on the relationships of the bean counter and business partner with actors besides the manager within an organization. This is a shortcoming of the literature, as business partners are more involved in decision-making processes and will therefore have to deal with multiple people (next to the manager) within a company (Siegel et al., 2003). Byrne and Pierce (2007) provided an interesting view on the characteristics associated with the different roles of management accountants, which include personal qualities, flexibility, and interpersonal and communication skills. Hopper (1980) examined how accountants and managers perceive the roles of management accountants when the management accounting functions were centralized versus when they were decentralized. These papers might be helpful in understanding the role of management accountants; perhaps they can even be used to investigate the relationships of bean counters and business partners with other actors in an organization. However, as these studies did not focus on these relationships, the evidence remains scarce.

(5)

Master thesis University of Groningen Page 4 information at the right time in order to enable them to make the most optimal decisions. The relevance from a theoretical perspective is that this study tries to make a start with filling the existing literature gap and that it can be used as a starting point for further research in this area.

Following from the previous arguments, the main research question is:

How does the organizational structure influence the network of management accountants, being classified as bean counters or business partners, within a company?

In order to answer this question, three sub-questions have been developed:

1. How can the organizational structure of the company be described in terms of organic versus mechanistic?

2. How can the social network of the management accountant be described within a firm? 3. How does the organizational structure influence the social network of the management

accountants?

The remainder of this paper is structured as follows. The next section provides a background of the relevant literature. After that, the methodology of this paper will be discussed, which describes how this research has been performed. The methodology section is followed by an overview of the results. The subsequent section provides a discussion of the findings, and the paper ends with a conclusion, along with the limitations of this study and suggestions for further research.

2.Literature review

In this section, the literature with respect to organic-mechanistic organizational structure, the social network theory and the role of management accountants will be discussed

2.1 Organizational structure

(6)

Master thesis University of Groningen Page 5 also supported by Aiken et al. (1980) and Covin and Slevin (1989). A further clarification of the distinction between the mechanistic and organic organizational structure is provided by Courtright et al. (1989). They state that “mechanistic forms of organization, characterized by hierarchical control, are more suited to stable environments, which purportedly afford a relatively high level of control over tasks. Organic organizational forms, characterized by dispersed control, are more suited to unstable conditions, under which task accomplishment and innovation should shift to the most knowledgeable parties” (Courtright et al., 1989, p.773).

2.1.1. Model of Khandwalla

For this study, the model of Khandwalla (1977) is used in order to evaluate the ‘organicness’ of the organizational structures of the companies involved in this research. This model is shown in Figure 1 and it has been used extensively in previous literature (e.g. Covin and Slevin, 1989; Naman and Slevin, 1993; Slevin and Covin, 1997; Ambrose and Schminke, 2003). Therefore, as this model has proven to be useful in classifying an organizational structure as organic or mechanistic, it will be used for this research.

Organic structure Mechanistic structure

Communication channel Open Restricted

Managerial style Allowed to vary freely Uniform

Decision-making authority Participation and group consensus used frequently

Superiors make decisions with minimum consultation and involvement of subordinates

Adaptation Free adaptation if

circumstances change

Holding fast to true and tried management principles

Emphasis Getting things done Formally laid down procedures

Control Loose and informal Tight by using sophisticated

control systems

On-job behavior Flexible Conform to job description

Figure 1: Organic-mechanistic continuum (Khandwalla, 1977, p.411; Sandberg, 2008, p.40)

However, an important remark needs to be made. Donaldson (2001) notes that the mechanistic and organic structures are two polar extremes. He states that, “the mechanistic and organic types each mark the poles of a continuum that registers differences in degrees of mechanisticness and organicness” (Donaldson, 2001, p.38). According to him, companies can lie at any point on this continuum. Burns and Stalker (1994) also acknowledge this point; they note that the mechanistic and organic structure represents a polarity and not a dichotomy. This means that a company facing relatively stable conditions, while at the same time dealing with a change in the environment might “operate with a management system which includes both types” (Burns and Stalker, 1994, p.122).

2.2 Social Network Theory

(7)

Master thesis University of Groningen Page 6 people included in these webs are connected with others. These connections are classified as ties by Granovetter (1973), which is a dyadic relationship between actors (Borgatti and Foster, 2003).

2.2.1 Tie Strength

Granovetter (1973) notes that ties can have different strengths. He defines the strength of a tie as “a (probably linear) combination of the amount of time, the emotional intensity, the intimacy (mutual confiding), and the reciprocal services which characterize the tie” (p.1361). Using Granovetter’s (1973) classification, a tie can be classified as strong, weak, or absent. Strong ties consist of people who are close to us (Granovetter, 1983). According to Krackhardt (1992, p.218), one of the main advantages of strong ties is that they “constitute a base of trust that can reduce resistance and provide comfort in the face of uncertainty”. Furthermore, McPherson et al. (2001, p.416) state that because people who have strong ties with others are more similar to each other, “contact (…) occurs at a higher rate than among dissimilar people”. They note that people’s networks are often homogeneous with respect to sociodemographic, behavioral, and intrapersonal characteristics. These homogenous networks, in which people have strong ties with each other, are called cliques by Doreian (1979). Another advantage of strong ties is mentioned by Granovetter (1983, p.209), who notes that “strong ties have greater motivation to be of assistance and are typically more easily available”. On the other hand, the main advantage of weak ties is information diffusion. According to Brass et al. (1998, p.23), “as the size of a network increases, the possibility of fragmentation (individuals forming subgroups) increases. Because similarity breeds attraction and interaction, subgroups of similar people form”. However, if information is only transferred using strong ties, it is likely that this information will only diffuse within a clique of a network. Therefore, weak ties allow information diffusion throughout an (organizational) network consisting of a number of cliques. This is consistent with the findings of Podolny and Baron (1997, p.674), who state that “the more weak ties an individual has in his/her network, the more valuable the network is as a source of information”. The difference between strong and weak ties will also have an influence on the manner of communication. According to Granovetter (1983), people with strong ties communicate using restricted codes, in which much meaning is implicit and taken for granted because they are familiar with each other. On the other hand, weak tie communication consists of elaborated codes, which are more complex and universal, as the communicating people differ from each other. This point is supported by Hansen (1999), who found that complex knowledge requires a strong tie in order to be transferred. The last tie that Granovetter (1973) classifies is that, next to strong ties and weak ties, a tie can also be absent. An absent tie can exist in two situations (Granovetter, 1973):

1. A tie is called absent if there exists no relationship between two actors.

2. A tie is called absent if there is a relationship between two actors; however, this relationship has no substantial significance. An example includes two people living in the same street who nod to each other when meeting on the street. In this case, the interaction is negligible.

2.2.2. Operationalization of tie strength

(8)

Master thesis University of Groningen Page 7 informants in different ways. We cannot assume that asking informants to assign tie strength based on this single word will assign values equally along all the dimensions informants think about when applying it to their alters”. In order to overcome this problem, Capaldo (2007) used the factors of duration, frequency and intensity to operationalize the strength of a tie. Marsden and Campbell (1984) also used the duration and frequency of a relationship to measure tie strength and they state that “closeness is used as a measure of the intensity of a relationship”. Therefore, supplementing the relatively subjective concept of closeness with objective measures of the duration and frequency of the relationship might prove helpful in operationalizing tie strength. This method is also used by Pettigrew (1997), who used closeness as a primary indicator of tie strength, but supplemented it with secondary indicators (e.g. how long people know each other). Therefore, for this study, it is proposed that network ties are strong if there is a high degree of closeness between actors, if contact between actors is frequent and if the relations with other actors in a company are of a relatively long duration.

2.3 The role of management accountants

As was already illustrated in the introduction, a move from a bean counter to a business partner role for management accountants can be observed nowadays. However, to my best knowledge, no research has been done on the social networks of both types management accountants. This is remarkable, as the networks of the bean counter and business partner are likely to be different because the bean counter is more focused on his own department while the business partner, in his role as advisor, has to get in contact with people outside his accounting department.

2.3.1 Bean counter vs. business partner

According to Granlund and Lukka (1997, p.240), a bean counter can be described as someone who aims to “write accurate and correct financial history, whose personal communication is typically directed into the accounting function only and whose communication to the outside of the accounting function concentrates on written reports”. In contrast, the business partner is “required to have – in addition to the knowledge of the principles of financial analysis – a good understanding of the business the firm is in, fluent communication skills and an ability to run change projects to” (Granlund and Lukka, 1997, p.240). Thus, the business partner role still involves a bean counter role; however, its role is extended as the management accountant is now more involved in the company. Therefore, it is likely that differences exist in the social networks of bean counters and business partners, as the latter are more involved with other actors in the company then the former. However, Byrne and Pierce (2007, p.469) state that “the adoption of a ‘business partner’ model for management accountants is found to be ambiguous, conditional and uncertain”. They note that management accountants often view themselves as having a role of decision-maker, which is in contrast with the opinion of other players in the organization (e.g. operation managers), who view “them (management accountants) more in a role that involved making suggestions, recommendations and influencing outcomes” (Byrne and Pierce, 2007, p.491). In this case, management accountants do not make decisions themselves, but provide information to decision-makers. Nevertheless, Byrne and Pierce (2007) acknowledge that management accountants are becoming more involved in companies, moving away from having a solely bean counting role. Another remark, made by Feeney and Pierce (2007), is that the concept of management accountants having a role of either a bean counter or business partner must be seen as two extremes on a continuum. They describe the concept of a hybrid accountant, which is someone “who assists managers in maintaining effective control, while also performing as a business partner as part of the management team” (Feeney and Pierce, 2007, p.16). Furthermore, Albu et al. (2008) characterize a hybrid accountant as having both management accounting and financial accounting tasks.

2.3.2. Operationalizing the role of management accountants

(9)

Master thesis University of Groningen Page 8 Furthermore, this model is supplemented with the model of Granlund and Lukka (1998), which mentions six characteristics which are often associated with the bean counter and the business partner. This model is shown in Figure 3.

Type of information

Attention to group processes

Financial and non- Financial information

Analyst Coach

(Business partner) Financial information Accountant

(Bean counter)

Educator Analysis Affect the thinking of

others

Figure 2: Four roles of management accountants (adapted from Nilsson et al., 2011, p.249)

Character Bean counter Business partner

Temporal orientation Emphasis on the past Emphasis on the present and future

Knowledge of the business in which the firm operates

Not expected Expected

The primary aim of communication

Fulfilling of formal information requirements

Active attention attraction in order to get the message through Felt scope of responsibility Narrow; covers the production

of correct accounting reports in time

Wide; covers both the production of relevant accounting figures and their application in business decisions

Cross-functional appreciation Limited High

General operating style Information collector and processor

A member of the management team and a change agent

Figure 3: Characteristics of bean counter and business partner (Granlund and Lukka, 1998, p.202)

Using these two figures, management accountants in the two case organizations can be classified as operating either more in a business partner or a bean counter role, thereby acknowledging that a stereotypical role does not exist. This method is consistent with Paulsson’s (2012) study. A remark that must be made is that viewing the change in the roles of management accountants as simply moving between the four roles as mentioned by Nilsson et al. (2011) along the two dimensions is ambiguous. According to Burns and Baldvinsdottir (2005, p.751), this “would ignore the cumulative nature of change events where context, history, institutional embeddedness and transformational agency all matter”. Therefore, changes in the roles of management accountants “need to be conceived and enacted in its broad and evolving context, and to be adaptable to alternative and largely unpredictable paths in the future” (Burns and Baldvinsdottir, 2005, p.751). However, this research only classifies management accountants as operating more in a bean counter or business partner role, leaving the question of how these roles change over time open for further research.

2.4 Research framework This paper is based on the research framework shown in Figure 4 on the next page, which is formed according to the following research question:

How does the organizational structure influence the network of management accountants, being classified as bean counters or business partners, within a company?

(10)

Master thesis University of Groningen Page 9 Lukka (1998). Finally, the social networks of both types of management accountants will be analyzed by measuring the strength of the ties included in the network using the concept of closeness (Marsden and Campbell, 1984) and the frequency and duration of the relationship. As can be seen, the framework shows that the proposition of this paper is that the organizational structure (consisting of organic and mechanistic elements) influences the social networks of the bean counter and the business partner. It is expected that a more organic organizational structure enables a management accountant to have a larger network, as this structure probably places less restrictions on the size and content of the network. Furthermore, based on the frameworks of Nilsson et al. (2011) and Granlund and Lukka (1998), it is expected that the business partner has more contacts outside his own department compared to the bean counter as his role includes giving advice. In the remainder of this paper, the specific effects will be investigated; however, first the methodology is discussed in the next section.

Organizational Structure Social Network Bean Counter Social Network Business Partner

3.Methodology

In this section, the research methodology of this paper will be described. First, a description is provided of the research method and the sample, which is followed by an elaboration on the data collection and data analysis processes used for this study.

3.1 Research Method

This research is of a theory development type. According to Van Aken et al. (2012), the theory development process consists of four steps. These steps include:

1. Observation of a business phenomenon, which has not been addressed in the literature or “when the literature related to the business phenomenon is still very exploratory in nature” (Van Aken et al., 2012, p.5).

2. The business phenomenon is observed using one or multiple case studies.

3. An explanation regarding the business phenomenon is developed and the findings of the research are compared with existing theories.

4. The final step consists of forming propositions, which “are changes of or additions to existing theories” (Van Aken et al., 2012, p.4).

Van Aken et al. (2012, p.5) state that “in theory development, the trigger for the research is a business phenomenon that is generally recognized in companies”. The business phenomenon in this research is that the networks of management accountants are likely to differ between a management accountant operating in a bean counter role compared to those operating in a business partner role. In order to investigate this business phenomenon, two case studies will be used (more details are provided in the next paragraph). According to Eisenhardt (1989, p.534), “the case study is a research strategy which focuses on understanding the dynamics present within single settings”. Furthermore, Yin (1981) describes three types of case studies, which are descriptive, explanatory, and exploratory. This research makes use of two exploratory case studies, as the goal is to uncover how the organizational structure influences the networks of both types of management accountants

(11)

Master thesis University of Groningen Page 10 within a company. In the discussion and conclusion section, explanations and propositions with respect to the business phenomenon will be provided.

3.2 Sample

The selection of companies for this study was not random, which is consistent with the view of Eisenhardt (1989, p.537) who states that for case studies, “random selection is neither necessary, nor even preferable”. This study follows theoretical replication (Yin, 2009, p.54), which means that cases are selected so that it is likely to “predict contrasting results but for anticipatable reasons”. Therefore, two ‘polar type’ organizations are used for this study. These organizations satisfied the two selection criteria for inclusion in this study, which are as follows:

1. The organization must employ both a bean counter and a business partner working, full-time within the company.

2. The organization must lie either at the mechanistic or the organic side of the organic-mechanistic organizational structure continuum.

Based on these two criteria, two Dutch organizations were selected for this study and they agreed to be part of this research. The first of these is Philips Drachten, which is a sub-unit of the manufacturing company Philips, located in Drachten with approximately 2000 employees (www.icdrachten.nl). The interviews took place with management accountants located in the innovation department of this site. The second organization is the Universitair Medisch Centrum Groningen, which is abbreviated as UMCG. This hospital is located in Groningen and has 11.586 employees, of which 8326 are FTE (www.umcg.nl).

3.3 Data collection

(12)

Master thesis University of Groningen Page 11 influences these networks. The interviewees were asked to write down the contact and the duration and frequency of the relationships in the table shown in Appendix 5, which was done in order to save time. Within Philips Drachten, the interview with the business partner was also used to gain insight in the organizational structure. Therefore, only two interviews were necessary for this case organization. All interviewees gave their permission to record the interviews, which prevented losing any important information. This increases the reliability and validity of this research (Golafshani, 2003). The interviews were transcribed afterwards and these transcripts were sent to the interviewees in order to confirm that the notes captured the interviewees’ narratives.

3.3.1. Concepts

The three concepts used for this research are the organizational structure, the role of management accountants and the social network theory. With respect to the first concept, interview questions were designed using the model of Khandwalla (1977), which is consistent with previous research (e.g. Ambrose and Schminke, 2003). In order to develop questions about the roles of management accountants, the models of Nilsson et al. (2011) and Granlund and Lukka (1998) were used, which is consistent with the method used by Paulsson (2012). Indirect questions were asked about the daily activities and tasks in order to classify the management accountant as either a bean counter or business partner.

The data collection with respect to the social network focused on the ego-centered network of the management accountant within the organization. It describes “individuals (ego’s) and all of the people (alters) whom these individuals identify as being associated with them” (Trotter, 1999, p.5). This means that the research focuses on the network ties of the management accountants (the ego) with other people within the organization (the alters). According to Marsden and Campbell (2012), egocentric network data can be used to inspect aspects of individual ties, which matches the goal of this study. In order to gain insight in the ego-centric network, name generators, which have the goal to “identify respondent’s alters”, and name interpreters, which are intended to “obtain information on the alters and their relationships” (Marsden, 2005, p.11), were used. Interviewees were asked with whom they have ties and subsequent questions were asked in order to provide more in-depth insights with respect to the strength of these ties (Burt, 1984; Marsden, 1990). However, for confidentiality reasons, the interviewees did not mention the name, but rather the function of the persons with whom they have ties. The strength of the ties was measured using 1) the concept of closeness (Marsden and Campbell, 1984), 2) the frequency of contact, and 3) the duration of the existence of the relationship. The degree of closeness was measured by asking the interviewees how close they feel with each of the contacts they mentioned. Furthermore, I relied on the description that the interviewees gave of their relationships in order to classify a tie as either strong or weak. This method is consistent with the approach Pettigrew (1997) used for her research. An upper limit of eight alters was used, while the minimum was set at three alters, which is consistent with the suggestions provided by Burt (1984). With respect to the frequency of contact, Krackhardt (1992) states that weak ties are those “who interact more than once per year and less than twice per week”, while “strong ties are those who interact at least twice per week” (Krackhardt, 1992, p.221). This classification is also used for this research. A remark made by Marin (2004, p.289) with respect to using name generators is that “respondents presented with a name generator are more likely to name alters with whom they share stronger ties, alters who are more connected within the network, and alters with whom they interact in more settings”. However, Battilana and Casciaro (2013, p.825) use the work of McEvily (1997) to state that “ego-network data have been shown to correlate well with data where each dyad is constructed based on information gathered from both members for the pair”. Therefore, using name generators is deemed to be an appropriate approach for structuring the social networks of management accountants.

3.4 Data analysis

(13)

Master thesis University of Groningen Page 12 pattern coding. This increases the internal validity of this research, as it provides a structured approach for analyzing the data and forming conclusions (Yin, 2009). According to Miles and Huberman (1994), there are two levels of coding. The first one is first-level coding, which is mainly focused on summarizing the segments of data. I made use of two approaches to create these codes. Deductive codes were developed using literature and professional insights. Next to that, I made use of inductive codes by looking at the transcribed interviews in order to find emerging important topics. The second level is pattern coding, which “is a way of grouping those summaries into a smaller number of sets, themes, or constructs” (Miles and Huberman, 1994, p.69). By using pattern coding, I grouped the first-level codes by finding a code which applied to all first level codes within the group. This coding process assisted in making sense of the large amount of collected data. An overview of all codes is provided in the codebook, which is shown in Appendix 2. Furthermore, the interview questions and an example of a coded interview are presented in Appendices 5 and 6.

4.Results

This section provides an overview of the empirical findings of this research. The results are presented for Philips Drachten and the UMCG separately. The results for each case are structured according to the three sub-questions mentioned in the introduction.

4.1 Philips

Both management accountants interviewed at Philips Drachten operate within the Financial Planning & Analysis (FP&A) department. The bean counter is dedicated to the personal care division, while the business partner oversees all the innovation divisions at Philips Drachten. Next to FP&A, two additional finance departments exist at Philips. The first one is an external sourcing partner, who performs the traditional bookkeeping tasks. The second one is financial operations, which is the linking pin between the sourcing partner and FP&A. People from this department are on the payroll of Philips. So, in general, this means that if FP&A wants certain entries to be made or requests a certain report, it contacts financial operations, who then tells the sourcing partners what to do.

4.1.1 The organizational structure of Philips

The organizational structure of Philips Drachten can be described as organic. The business partner at Philips stated that:

‘We are by our very nature organic. In other words, we can do a lot.’

People within the innovation department need to have flexible on-job behavior. The job description describes the boundaries within which they can act. The business partner describes this clearly by stating that:

‘You must show entrepreneurship and manage the company as if it was your business. You cannot say: I cannot do that because it is not listed in my job description.’

Also, decision-making within Philips Drachten is mainly focused on consensus. However, sometimes authority is used by managers on both a departmental and companywide level. For example the top management of Philips might obligate Philips Drachten to achieve a certain target. An example mentioned by the business partner was a reduction of five percent of the travel costs. However, he also stated that the top management gives Philips Drachten freedom with respect to the method used to achieve this reduction; they do not state which actions and decisions Philips Drachten should take in order to accomplish this goal. This is not uncommon. The business partner stated that:

(14)

Master thesis University of Groningen Page 13 This again shows that Philips Drachten can be described as having an organic organizational structure. Furthermore, because the department of Philips Drachten included for this research is focused on innovation, using a tight control system is not feasible. Both the bean counter and business partner stated that innovation is associated with uncertainty. When starting on an innovation project, the duration and the costs of this project are not known in advance. However, this does not mean that innovation projects are out of control. Philips Drachten gained a lot of experience throughout the years on how to handle these projects. Furthermore, each project is divided into phases and budgets are used for each project. Also, a control system is in place; however, this system is focused mainly on the costs of the project. According to the business partner:

‘I know at every moment precisely what the costs are of a certain project.’

With respect to communication channels, Philips Drachten has an open structure. This openness is made possible by the messenger system, which allows voice and video communication over the intranet. Every employee of Philips worldwide can be contacted using this system. For example, the bean counter stated that she has contact with colleagues in Hong Kong and China. Furthermore, she stated that the innovation department does not have a tight hierarchy. According to her:

‘At this department, we have an innovation director, who is an important person. However, we often chat with him and drink coffee.’

However, Philips is moving towards are more mechanistic model. RACI (roles & responsibility chartering) are implemented, which describe the roles and responsibilities of employees within Philips Drachten and make clear who is allowed to do what. According to the business partner:

‘This organization is changing towards are more mechanistic structure. This means that my responsibilities are clearly defined for me as a business partner.’

‘When looking at the management style, I still have a high degree of freedom, however, this becomes more and more restricted.’

So, while the innovation department of Philips Drachten can currently be described as being organic, it is moving towards a more mechanistically-orientated organizational structure.

4.1.2 The social network of the management accountant within Philips

With respect to the management accountants operating within Philips Drachten, it must be noted that the classic bean counter is not a job that exists anymore at Drachten. According to the business partner, two types of accounting exists. The first one is transactional accounting, for which precise descriptions and procedures can be developed. These classic bookkeeping activities are outsourced, which is more cost-effective. The second type of accounting is judgmental accounting, which implies that a judgment has to be made. An example of this type mentioned during the interviews are impairments. With respect to these two types, the business partner stated that:

‘Transactional accounting is outsourced, judgmental accounting is still done by us. However, we need to make sure that we do as little judgmental accounting as possible. We want to automatize as much

as we can.’

(15)

Master thesis University of Groningen Page 14 bean counter stated that this is mainly financial. The business partner also stated that he focuses mainly on financial information. For example, he noted that:

‘I don’t look at quality, that is the responsibility of a quality manager. I don’t feel responsible for that.’ However, he also mentioned the use of non-financial information. This focuses mainly on internal control; for example, making sure that the right people sign for a purchasing order. According to him:

‘The role of the business partner today involves making sure that the internal control framework works, which means making sure that people comply with rules and policies.’

Both management accountants also stated that they focus mainly on the future. The bean counter stated with respect to her work that:

‘We make budgets, which is about the future. However, these budgets are based on experiences from the past.’

The business partner also acknowledged that, while he focuses mainly on the future, he still does some analytical tasks using information from the past. However, he noted that these practices should actually be executed by financial operations and the external sourcing partner. But representatives of these functions are not available on the location in Drachten, which is why people ask him to perform these analyses. The business partner also stated that this is one of the main drawbacks of centralization of these bookkeeping tasks.

‘In principle, you can centralize everything. However, if things are executed centrally, people do not have the connection with the local business. You need someone who understands what is going on in

the organization.’

He also stated that as a business partner, you need to understand the business.

‘A business partner needs to understand what is going on, which processes there are. This is crucial for him while performing his job. If you do not have this knowledge, you cannot function properly.’ This same point is made by the bean counter.

‘You cannot perform my function with your eyes closed; knowing how to make debit and credit entries simply is not enough.’

The felt responsibility can be classified as wide; it is not simply generating accounting reports. As a matter of fact, both the bean counter and business partner mentioned that they are not producing accounting reports; this is done by the sourcing partner and financial operations. The business partner stated that:

‘The classic bean counter produces reports and the business partner makes sure that they are applied in the daily business. With the new RACI model, Philips wants to make sure that business partners are

not putting time in generating these reports.’

With respect to communication, both management accountants mentioned that they do not only communicate with other management accountants. The bean counter stated that:

(16)

Master thesis University of Groningen Page 15 The fact that both the bean counter and business partner communicate with other departments also becomes clear from their social networks. These social networks are shown in Figures 5 and 6. Each

box represents a function with which the management accountant has a relationship. Furthermore, each tie is classified as being either strong or weak. As can be seen, both the bean counter and business partner have three strong ties and five weak ties. The duration and frequency of both social networks is shown in Appendix 3. With respect to the network of the bean counter, the three strong ties represent her ‘home base’; she reported a high degree of closeness with these three actors. The bean counter is dedicated to a division of personal care innovation. She supports the people (mostly innovation teams) within this division. Furthermore, she also communicates with the innovation director of personal care; however, this contact is less frequent. The other five actors are more supporting or controlling for her as a management accountant. According to her:

‘We are part of the finance & accounting team and we support the innovation teams and the head of innovation. These people (the other five actors) support us or control us.’

The duration of all these relations, except for accounting and internal control, is 2,5 years, which is the time she has been working within Philips. The business partner reported a high degree of closeness with the center of expertise, innovation leads and head of innovation personal care.

‘These are really close to me. We are all innovation. And then you have the outside world.’ Contact with the head of innovation personal care and the center of expertise is mainly through the messenger system. He also stated that, for him operating in a business partner role, the other five actors are more relationships on an informal basis. For example, he noted that:

‘I communicate with the general site manager because I want to know what is going on within the business. I also have no formal relationship with finance business. Internal control & audit control me

and I have a relationship with the other two actors because there was no one else.’

The duration of these relationships have all been, except for one, longer than five years, which is the time the business partner has been operating in his current function.

4.1.3 The influence of the organizational structure on the social network of the management accountant

Of the seven characteristics included in the framework of Khandwalla (1977), which were described in Section 2.1, two are found to influence the social network of the management accountant. These

Figure 5: Social network bean counter Figure 6: Social network business partner

Business partner Weak General manager _site Finance business Weak

Innovation leads Strong

Suppliers

Weak _{Internal control &} audit Weak Center of expertise Strong Head of innovation Personal Care Strong Government subsidies Weak

Bean counter Weak _OperationsFinancial Sourcing partner Weak

(17)

Master thesis University of Groningen Page 16 include the communication channel and emphasis. With respect to the communication channel, both the bean counter and business partner acknowledged that the messenger system makes it possible to have a wide range of relationships within Philips. For example, the bean counter noted that she is able to contact the CFO of Philips with the system. Also, this system makes it possible to have connections all around the world. However, the business partner also stated that it depends on how you as a management accountant make use of the system.

‘As a business partner, you can make yourself invisible within a company by staying in your room. However, I do not think that you can function properly in this way. As a business partner, you need to

be approachable and visible.’

This is something which the messenger system makes possible. However, restricting effects of the organizational structure were also noted by both management accountants. These have to do with the emphasis characteristic. The business partner is not really satisfied with the new RACI model. ‘I understand the reasoning behind the model, however I don’t agree with the implementation speed.’ Furthermore, because of this new model, some activities are outsourced and centralized. However, the business partner thinks that some of these activities will eventually come back to Drachten. ‘Of course you can centralize everything, which makes it possible to make processes efficient and gain economies of scale. However, after some time it will become an island which has lost contact with the

outside world. Therefore, you always need someone who knows what is going on within an organization.’

According to the business partner, this new RACI model will have an influence on his network. ‘Because of this model, I will have less contact with for example the suppliers. However, the core of

my network remains the same.’

Also the bean counter mentioned that she feels restrained by the organizational structure. However, she does not feel the need to change anything.

‘If you have rules, then you are restrained. However, I do not feel that I want something to change.’ ‘Some of the rules are financial rules, which we simply cannot change. We also operate within the IFRS framework. So of course we are restrained, but we comply with the law. But that is life. You also

cannot drive through a red light.’ 4.2 UMCG

The organizational chart of the UMCG is shown in Appendix 7. According to this chart, the UMCG is divided into six sectors. The business partner who is included for this research operates as a management accountant (sectorcontroller) in sector A, which is about providing long-term care to patients. Along with these six sectors, four more supporting divisions exist within the UMCG. One of these is called ICT, Functional & Data Management and Medical Technics, which is abbreviated as IFM. The bean counter operates within this division as a management accountant and she is responsible for the ICT element.

4.2.1 The organizational structure of the UMCG

The organizational structure at the UMCG can be classified as more on the organic side of the continuum. According to the concerncontroller, his management style is allowed to vary.

(18)

Master thesis University of Groningen Page 17 However, he also noted that people lower in the hierarchy within the UMCG often have more extensive job descriptions.

‘Within the UMCG it is often the case that the more operational the job, the more extensive the job description is.’

For example, nurses need to use protocols if certain events happen. They also need to register a lot. Nevertheless, on-job behavior also needs to be flexible at times.

‘It is not completely standardized work, but that is because of its nature. Nurses work with people, which involves a certain degree of unpredictability.’

This also has consequences for the used control system within the UMCG. Planning is used and certain norms are in place with respect to, for example, the number of nurses attached to a certain section. However, although norms can be developed, unpredictability means that these norms cannot capture everything. For example, the concerncontroller stated:

‘It is difficult to predict if a patient will get side effects.’ Therefore, according to the concerncontroller, this will influence the control system.

‘You can plan the work of nurses on the things that they need to do. However, if something unpredicted happens, you will have a problem. Therefore, buffers need to be in place.’

The procedures also state that nurses need to work in couples. In fact, the concerncontroller noted that much of the work on the operational level is executed in teams. Therefore, it might be that informal control can also be seen as a certain form of control system which is present within the UMCG. However, the concerncontroller is not convinced if this form of control system is present within the organization.

‘You hope that this (informal control) happens. It would be a healthy culture, it is something that you want. However, I’m not sure if it exists within the UMCG.’

With respect to the decision-making authority, the concerncontroller described the UMCG as using consensus in making decisions.

‘At the UMCG, we have very transparent decision-making processes.’

Furthermore, he also stated that in some cases more authoritarian decisions are preferred in his opinion. At the UMCG, there are six sectors, which involve, for example, long-term care and cancer. Each of these sectors has a sector director, who reports to the board of directors. Furthermore, each of these sectors is divided into departments and each department has a head of department. However, these department heads are also reporting to the board of directors and are thereby circumventing the directors of the sector. The concerncontroller is not really satisfied with this structure.

‘It creates a certain area of tension. I think that it would be better for the decision making processes if all the power is on the sector level.’

(19)

Master thesis University of Groningen Page 18 ‘I think that the Philips messenger system is great. But we do not have such a system at the UMCG.

We do have an intranet, which provides news and announcements. But this is one-way communication.’

What the UMCG does have, next to a formal controller meeting, is a stand up meeting, during which controllers tell what they are currently working on. According to the concerncontroller:

‘This is a nice method to inform the others about what you are doing at the moment.’

4.2.2 The social network of the management accountant within the UMCG

Within the UMCG it was possible to identify a management accountant who operates more in a bean counter role and one who operates more as a business partner. However, the bean counter had some characteristics which, according to the models of Nilsson et al. (2011) and Granlund and Lukka (1998), characterize a business partner. Both management accountants stated that the information that they use consists of both financial and non-financial elements. According to the bean counter:

‘We use both types of information. For example, we control if the realized hours match the budgeted hours within the ICT department. However, in the end we will attach a euro sign to this.’ This point is also made clear by the business partner. According to him, the basics of the work of a controller consists of issues about money. With respect to his work, he described the Balanced Scorecard of the UCMG, which consists of five perspectives.

‘And that is not only about money. Of course it is also about quality. However, we have other people in my sector who are responsible for that.’

Both the bean counter and business partner also noted that they focus not only on the past, but also on the current situation and the future. According to the bean counter:

‘Yes, we analyze the budget and check if the budget is achieved. However, we are also working on the budget of 2015 and the multiannual planning for the years 2016, 2017 and 2018.’

The business partner mentioned that if a budget is not met, you only can make changes in the processes by looking back at what caused these deviations in the past. The main difference between the business partner and bean counter within the UMCG is with respect to the required knowledge of the business. The bean counter stated that this is not required for her.

‘In my current role as ICT controller, I do not need knowledge of the business. Other people from ICT do need this knowledge, as they are working together with medical specialist to create new systems.’ She also noted that if she compared herself with the business partner, she expected that the business partner needs more knowledge of the business. To a certain degree, she was right. The business partner stated that he is not a physician.

‘So, do you need the knowledge… I think to a certain level. I will never be a doctor.’

According to him, it is important to have the knowledge and skills to analyze problems in a structured manner. He stated that this will make every problem less complex.

‘You need knowledge of conducting business. And it is nice if you understand which type of care is done by a medical department. But if you have knowledge of conducting business in general, you can

(20)

Master thesis University of Groningen Page 19 The felt responsibility of both the bean counter and business partner within the UMCG can be classified as wide; it is not simply generating accounting reports. As a matter of fact, both management accountants stated that they barely produce these reports.

‘If we talk about reports, those are primarily established by controllers within my team.’ (Business partner)

‘The assistant controller within this departments establishes the reports within this department.’ (Bean counter)

With respect to the contacts of both management accountants within the UMCG, both stated that they do not only communicate with other management accountants. Both the bean counter and business partner have a wide spectrum of contacts with much variation. This also becomes clear from their social networks, which are shown in Figures 7 and 8. Each box represents a function with

which the management accountant has a relationship and each tie is classified as being either strong or weak. As can be seen, the bean counter has five strong ties and three weak ties, while the business partner has seven strong ties and one weak tie with department heads within sector A. These department heads are medical people. The business partner explained that, with respect to his felt degree of closeness, two domains can be found with the UMCG.

‘You have the medical domain and the business domain. I have a good relation with the department heads within sector A. But it is a totally different world.’

The other seven actors within his network are strong ties, as the business partner stated that they operate in the business domain. These actors are not only operating within his sector. He noted that he does not feel a higher degree of closeness with people within his sector then with outsiders.

‘That is not how I am. I feel at home everywhere.’

Contact with the concerncontroller is specifically about sector A, while contact with the Board of Directors and the director financing is more about UMCG wide subjects. The actor external financing negotiates with insurers about how much money the UMCG can ask for its care products. This is something in which the business partner can assist. Furthermore, the three remaining strong ties operate within sector A. The business partner gives advice to the managers of sector A and his boss is the director of sector A. Lastly, the controllers of sector A report to the business partner. The network of the bean counter also includes a strong tie with the concerncontrollers. Furthermore, it

Figure 7: Social network bean counter Figure 8: Social network business partner

Business partner StrongDirector sector A Director financing Strong

Managers sector A Strong Controllers sector A Strong Board of Directors Strong External financing Strong Concerncontrollers Strong Department heads sector A Weak

Bean counter Strong _{controllers IFM}Assistent Project leaders Strong

(21)

Master thesis University of Groningen Page 20 also includes strong ties with the head of departments, project leaders, the business manager IFM and the assistant controllers of IFM.

My rooting goes mainly with people who run the business within IFM. And much information also goes through the concerncontroller.’

According to the bean counter, the concerncontroller and the business manager of IFM are her main information providers. Furthermore, she also receives reports which are made by the assistant controllers. With this information, she can help and advise the other actors of her network. She also stated that all of the included relations are formal.

‘With all these actors I have weekly, monthly or once per quarter consults. These are recurring appointments.’

The duration and frequency of both social networks are shown in Appendix 4.

4.2.3 The influence of the organizational structure on the social network of the management accountant

Of the seven characteristics included in the framework of Khandwalla (1977), which were described in Section 2.1, two are found to influence the social network of the management accountant. These include emphasis and the communication channel. Both the bean counter and business partner stated that the emphasis is focused on getting the job done. They note that the organizational structure allows them to have a broad and varied spectrum of relationships. According to the business partner:

‘I feel free to get in contact with the people that I want. I do not feel restricted in that. If I am working on something and I want to include a certain stakeholder, I do not feel obstructed in doing that.’ He also stated that within the UMCG you are more or less free to go your own way. The bean counter felt the same about this issue.

‘Sometimes, I feel restricted by the organizational structure. However, if I need someone for doing my job, I will search for the people I need.’

Furthermore, with respect to the communication channel, the UMCG has an open-door policy. According to both the bean counter and business partner, communication is mainly face-to-face, by telephone, or by mail. The UMCG does have an intranet, which makes communication possible. However, both management accountants did not find it useful for one-on-one contact.

‘It is useful for sharing documents with a large group of people. I do not see it as an useful addition for my role.’ (Bean counter)

‘The intranet is useful to reach a large audience. But is not applicable to one-on-one contact. There is no need to organize a session on the intranet for that.’ (Business partner)

With respect to suggestions for improvements of the organizational structure, the business partner did not mention any. However, the bean counter did mention one. According to her, she sometimes would like to have more transparency.

‘For example, I do not know what the financial director and the director IFM discuss. And I do not always need to know this. But sometimes decisions are made and I am involved in my role too late. In

(22)

Master thesis University of Groningen Page 21 The existence of a medical and a business domain also influences the advice that management accountants give within the UMCG. According to both the bean counter and business partner, their advice is only about business issues; advice about medical issues is not their domain. The bean counter gave a clear description of this distinction:

‘If I get a request for a certain medical project, I can only give advice about for example the financing of this project. I cannot say anything about the medical content.’

5.Discussion

After presenting the results for each case individually, this section will provide a cross-case analysis by comparing the two cases. Furthermore, the results will be analyzed and compared using relevant literature and the section ends with theoretical and managerial implications.

5.1 Influence of organizational structure

The main goal of this research was to investigate how the organizational structure influences the network of the bean counter and business partner both working within the same company. After analyzing the results, it can be stated that the organizational structure does influence the content and size of the social networks of both types of management accountants. In fact, this research uncovered that this effect on the network did not differ between the bean counter and the business partner. Furthermore, the network of both types of management accountants did not differ much from each other, as both have networks which include contacts from outside their own department. This results in an updated research framework, which is depicted in Figure 9. The fact that the social

Organizational Structure

Social Network Business Partner/

Bean Counter

networks of both the bean counter and the business partner include people from outside their own departments is in contrast with the expectation stated in Section 2.4, which is that the business partner was expected to have more contacts outside his department compared to the bean counter because his role includes giving advice to other actors. However, this contrasting finding can be explained by the fact that the bean counters included for this research also have characteristics which are associated with a business partner. The bean counters at both Philips Drachten and the UMCG are hybrid accountants, who are described by Paulsson (2012, p.383) as “participating not only in accounting oriented tasks but also in issues that are related to the operations in the organization”. This makes it likely for them to have contact with people outside their financial orientated department in order to become involved in the operations of the organization, which is confirmed by this research.

Within Philips Drachten, two elements of the organizational structure proved to be of influence on the social networks. The first element consists of the communication channels available within the organization, which was found to be open and can therefore be classified as organic according to Khandwalla (1977). Philips Drachten makes use of an extensive messenger system next to face-to-face communication. This allows management accountants to not only get in contact and share knowledge with people located in Drachten and other locations in the Netherlands, but also with people working for Philips across the globe. With respect to communication channels, Gupta and Govindarajan (2000) found that knowledge flows in multinational companies can only happen if there are adequate transmission channels. In the case of Philips Drachten, these transmission channels proved to be adequate, as the messenger system enables employees within Philips to have a wide social network. In contrast with Philips Drachten, the management accountants did not find such an extensive messenger system useful at the UMCG. This is reasonable, as the units of the

(23)

Master thesis University of Groningen Page 22 UMCG are not dispersed across the globe or even across the Netherlands. As the business partner at the UMCG stated:

‘It is just not needed here. Everything is present here at this “village”. If you walk a few minutes you can reach everyone.’

At the UMCG, the organizational structure of having an open door policy makes it possible to have a wide array of contacts. This means that management accountants at the UMCG are allowed to get in contact with the people they need in order to perform their job, which again can be classified as an organic, open communication channel. A second element of the organizational structure which influenced the social networks at Philips Drachten was the implementation of a new RACI model, which is considered to be part of the mechanistic emphasis characteristic, as mentioned in the framework of Khandwalla (1977). According to the business partner, the core of his network will remain the same, however other (weak tie) relationships will vanish because of this new RACI model, as these are not his responsibility anymore. This is in contrast with the UMCG, as the emphasis at this organization is on getting the job done, which is a characteristic of an organic organizational structure according to Khandwalla (1977). It means that both the bean counter and business partner are able and allowed to get in contact with the actors that they need in order to perform their job. Both types of management accountants at the UMCG stated that they do not feel restricted in getting in contact with the people they want. According to Podolny and Baron (1997), putting restrictions on the network of the business partner at Philips Drachten might have negative consequences. They note that weak ties provide a valuable source of information. Therefore, the removal of some weak ties from the network of the business partner at Philips Drachten might result in him having less information available about certain aspects of the business. This is something that Philips needs to keep in mind, especially because the business partner stated that knowledge of the business is essential for him in performing his job. Furthermore, differences between the advice that management accountants give in both organizations were found. At the UMCG, management accountants only give advice about business issues; medical issues simply are not their domain. Management accountants at Philips Drachten are able to give advice about almost all issues that are important within the organization.

With respect to the organizational structure of both organizations, it can be stated that these are more on the organic side of the organic-mechanistic continuum. These findings are consistent with the findings of Joyce et al. (1997) and Courtright et al. (1989), as they found that organic structures are more suited for companies dealing with uncertainty. Both the innovation department of Philips Drachten and the UMCG do have uncertainty in their environment, which makes having an organic organizational structure a logical choice according to the literature.

5.2 Network ties

(24)

Master thesis University of Groningen Page 23 better adapt to uncertainty and environmental changes (Krackhardt, 1992). As already mentioned in the previous section, both organizations face uncertainty in their environment. This means that, according to Krackhardt and Stern (1988), both organizations are better able to deal with this uncertainty because of these interdepartmental strong friendship ties. Another interesting result is that the business partner at Philips Drachten has strong ties with actors with whom he mainly has contact using the messenger system. This finding is consistent with the work of Wilson et al. (2006). They found that the level of trust at computer-mediated teams was comparable to face-to-face teams. Therefore, it can be stated that the strength of a tie can be strong even if the communication happens via a messenger system instead of face-to-face.

The measures used to classify the tie as either strong or weak include the duration of the relationship, the frequency of contact and the concept of closeness. The frequency and duration of the relationships are presented in Appendices 3 and 4. Overall, the concept of closeness proved to be most helpful in classifying the strength of the ties. The durations of the relationships have been either the length of time the interviewees are working in the organization or the time that they have been operating in their current function. For example, the bean counter at the UMCG is currently working 7 years at the UMCG, with 2 years in her current function. Therefore, the duration of her relationships are either 2 or 7 years. Furthermore, much variation was observed in the frequency of contact. According to Krackhardt (1992, p.221), “strong ties are those who interact at least twice per week”. However, the findings of this research present a contrasting view. Some of the relationships which are classified as strong ties have a frequency of more than twice a week. However, this does not hold for every relationship classified as a strong tie. This can be explained by the fact that Krackhardt (1992) based his classification on the research of Granovetter (1973), who studied how unemployed people use their contacts to find a new job. However, the network that people use to find a job probably differs from the network of contacts people have within their employing company. Overall, these findings support the view of Marsden and Campbell (1984, p.482), who state that of the techniques available for measuring tie strength, “a measure of closeness or intensity is the best indicator of strength”. They also studied the use of the frequency and duration as a technique for measuring tie strength, but conclude that “we would suggest that these indicators be avoided in favor of closeness” (Marsden and Campbell, 1984, p.499). This study used all three measures, which is consistent with Capaldo’s (2007) method. However, of these measures, the closeness measure proved to be most helpful in classifying the tie strength.

5.3 Theoretical and managerial implications

“The influence of the organizational structure on the network of management accountants within a company”