Different roles of management accountants and their related influence on risk management

(1)

Different roles of management accountants and

their related influence on risk management

A quantitative study

Student: Lianne Wenker

Student number: S2648679

Study: MSc BA: Organizational and Management Control Supervisor: Prof. dr. ir. P.M.G. van Veen-Dirks

Co-assessor: Drs. D.P. Tavenier

Date: 25-6-2018

(2)

2

Abstract

The aim of this research was to examine the possible relationship between the role of the management accountant and the influence they exercise on risk management. This research tested this possible relationship. Although this relationship is expected by the previous written literature, it has not been tested before. So, I do add a test to the present literature. A survey is conducted with 51 respondents, who are working as management accountants (or in similar functions). The survey measured the role of the management accountants and their perceived influence in risk management. In this study, the role of the management accountant did not determine the influence in risk management. Some other factors did (partly) determine the influence. These factors are the presence of a risk manager, the function and the sector where the respondents are working in. Although the relationship is expected by other authors, the empirical evidence is still not found. So, according to this study, this direct relationship between the different roles and the influence in risk management does not exist.

(3)

3

1. Introduction

The recent financial crisis has increased the interest in risk management (Hayne and Free, 2014; Soin and Collier, 2013). This increased interest in risk management is also an answer to the different scandals that questioned the legitimacy of some organizations (Power, 2004). By detecting and managing risks, organizations can decrease the amount of unforeseen scandals. Furthermore, not all risks can be assigned to one organization, so it exceeds organizational borders. Because of these reasons, risk management has become an integral part of public and private sector organizations (Huber and Scheytt, 2013; Soin and Collier, 2013). The growing importance of risk management influences the discussion about enterprise controls and the work of management accountants is affected by this (Bhimani, 2009).

To make risk management an integral part of organizations, Enterprise Risk Management (ERM) has been implemented. ERM is a recent development, which incorporates the management of all possible risks that an organization can encounter. This is important to gain a better

understanding of all faced risks (Gates, 2006). Examples of possible risks are preventable inside risks, strategy risks and external risks (Kaplan and Mikes, 2012). So, there are a lot of risks to manage for.

Previous research on risk management is primarily focused on technical aspects, like the use of the COSO Framework and other technologies/skills. Since the late 90’s, risk management is seen as a management control practice across all industries (COSO, 2004). As it is part of the management control practice, it is possible that management accountants deal with this topic. When looking at skills, management accountants can use several skills in their interaction with managers. According to Hall, Mikes and Millo (2015), this use can be defined as toolmaking. With toolmaking, experts adjust and reconfigure tools that embody their knowledge (Hall et al., 2015). In this example, the used tool is interaction skills. The usage of these tools can divide the management accounting profession in two groups, namely engaged toolmakers and compliance experts. Compliance experts do not direct the behavior and decisions of other people, but they have an important role by meeting requirements which were set beforehand. Examples of these requirements are written reports. Engaged toolmakers use tools to effectively communicate their knowledge, but make their own role still necessary for a full understanding of the information. With these skills, they can direct the decisions of other managers (Hall et al., 2015).

(4)

4

organizational dimensions that are important for risk management and it shows the different situations in organizations where risks can take place. This can be used to assess the risks that an organization can face (Hayne and Free, 2014). The 2017 extension also includes strategy-setting into this framework (COSO, 2017), which indicates that risk management becomes more important for organizations.

However, these aspects do not cover the whole subject. A possible question is: who controls the risks? According to Soin and Collier (2013), risk management becomes an issue of management control. But it is not sure where the management accountant fits into this situation (Soin and Collier, 2013). The work of the management accountant can be seen as calculative work. However, risk management is not an issue of the calculative work of management accountants (Mikes, 2009). So, if management accounting is seen as calculative work, it does not fit in risk management. However, if the work of the management accountant is seen in the broader context of management control, risk management could be a part of the job. So, it can be said that not all roles of the management accountant do take risk into account. This means that the role of the management accountant determines the fitting in the risk management process. The shift to a broader function can be seen in the emerging business-oriented roles for management accountants. The focus is changed from score-keeping and corporate policing to business-integrated roles (Burns and Baldvinsdottir, 2005; Windeck et al, 2015). This change is directed by changes in technologies and the environment (Lambert and Sponem, 2012) or institutional changes (Goretzki et al., 2013).

The management accountant with a business-oriented role is more involved in business processes. They add value in the decision-making process and they are exercising control (Windeck et al, 2015). Instead, a management accountant with a score-keeping role/number-oriented role is primarily checking the performance of the organization (Hopper, 1980). With this work, the management accountant has not a lot of influence with respect to decision-making. So, more power is available for the business partner management accountant than for the score-keeping/number-oriented management accountant.

(5)

5

As stated above, previous research focused on the technical aspects of risk management, but other things also need to be taken into account. Examples are the wider social, institutional and organizational context (Soin and Collier, 2013). This includes the decision-making process and the needed actors. In this research, the emphasis will lie at the actor management accountant and it will include some contextual factors.

As discussed, it is clear that risk management can be a part of internal control and the work of management accountants. However, little attention has been paid to the interrelation of risk management and the role of management accountants (Bhimani, 2009; Soin and Collier, 2013) and, therefore, the impact of the work of management accountants in the topic risk management is not really clear (Soin and Collier, 2013). This raises questions about the work of the

management accountant.

This research aims to give a better understanding of the interrelationship between management accounting and risk management, by focusing on the roles of the management accountant with respect to the influence in risk management. The corresponding research question is:

To what extent do the different roles of management accountants determine the influence on risk management?

So, the contribution of this research is to investigate if this relationship really exists and to give empirical findings of this possible relationship. This is an addition to the current literature, in which the relationship is not clearly stated and where statistical evidence is missing. Answering this question is also practically relevant, because companies can use this outcome to determine how to shape the role of the management accountant to fit into the companies’ decision-making process.

2. Literature review

(6)

6

2.1 Risk management

Recently, a lot of attention is paid to risk management (Gendron, Brivot and Guenin-Paracini, 2016; Woods, 2009) and new risk management concepts have evolved (COSO, 2004). Risk management can include several forms of risks, like preventable risks, strategy risks and external risks (Kaplan and Mikes, 2012). As risk management includes many facets, it has emerged as a wide practice in organizations (Mikes, 2009). Preventable risks can be controlled by rule-based management (Kaplan and Mikes, 2012), but that is not possible for strategy risks. These risks are more and more discussed in the boardroom (Arena, Arnaboldi and Azzone, 2010), what indicates the importance of risk management.

A form of risk management is ERM, to control for uncertainty by linking risk management with strategy-setting. This became extremely important in the 1990’s (Arena et al, 2010). Few companies have implemented ERM completely, so little research has been done about the impacts of full implementation and the incorporation of strategy-setting and strategic risks (Gates, 2006). A reason for no full implementation can be to see risk management as a

discipline of everything and nothing. This means that the willingness to control all risks ends in the control of nothing (Power, 2009). The willingness is dependent on the risk appetite of the company. Risk appetite means the amount of risk a company wants to take (Power, 2009). However, risk appetite can exist in all organizations (Power, 2009), so not only in companies that implemented ERM. Every organization organizes risk management different, because of the calculative cultures (Mikes, 2009) and the choice to only count risk or to make risk count (Mikes, 2011). Another reason is the determined focus on risk management; companies can focus on only financial and legal risks, but also on company-wide risks.

The tools for risk management are mostly highly analytical and data-driven technologies. As an example, I will discuss the COSO framework. From the 2004 COSO-framework, a cube is derived. This cube shows three different dimensions: objectives, place in the company and the kind of activity (Hayne and Free, 2014). In this way, a company can assess the risk practices in all possible ways. So, this is a framework to analyze risk. In 2017, an extension of this

framework is published. The aim of this new framework is to provide a better insight into the relation between strategy, risk and performance. With this framework, managers will gain more insight into the impact of different risks on the strategy-setting and how this influences the

(7)

7

knowledge to the managers, while keeping themselves important by having tacit knowledge. So, communication is also a skill that is important in risk management.

Although these methods are used, there are different ‘calculative cultures’ that influence the attitudes towards thesemethods. These cultures are ‘calculative idealism’ and ‘calculative pragmatism’. Calculative idealism prefers risk quantification, while calculative pragmatism sees numbers as predictions and it encourages envisionment (Mikes, 2009). This ends in two ways to manage risk, risk measurement and risk envisionment. While risk measurement works with first order measures and then aggregates it in second-order measurements, risk envisionment is more based on judgment of the measurements to provide different scenarios (Mikes, 2011). Risk management is evolving in the direction of envisionment, but it is still undergoing

(8)

8

2.2 Roles of management accountants

There is a lot of discussion about the changing roles of management accountants during recent years (Järvenpää, 2007). Several distinctions are made. A ‘number-oriented’, ‘administrating’, ‘controlling’, ‘banking’ and ‘consulting’ role is defined by Mouritsen (1996). Bollecker (2007) separates the roles as ‘surveillance’, ‘consulting’, ‘coordination’, ‘change’ and ‘information’. However, these different roles can be divided between a bookkeeping role and a business-oriented role (Janin, 2017; Järvenpää, 2007; Morales and Lambert, 2013). Furthermore, this division is used a lot in management accounting research (for example Hartmann and Maas, 2011; Janin, 2017; Lambert and Sponem, 2012). Because of this widespread use, this division will be used in this research. I have changed the name from bookkeeping to number-oriented, because bookkeeping is found too strict. The meaning is the same, so these two names can be seen as synonyms.

The change in roles is mainly the transition from a number and control orientation to a business-oriented role (Burns and Baldvinsdottir, 2005; Janin, 2017; Järvenpää 2007). However, both roles are still discussed in other papers. This showed that both roles are still present and that not all organizations have the same role for management accountants. Too, the described change cannot be found in all organizations. Some organizations still deal with number-oriented

management accountants. The number-oriented role primarily contains data-treatment and the use of calculative techniques (Morales and Lambert, 2013), while the business-oriented role is defined as the willingness and ability of the management accountant to provide more added value to the management of the companies (Järvenpää, 2007).

But how are these roles fulfilled in organizations? The work of management accountants takes place in the whole organization, not only in the accounting section (Jöhnsson, 1996; Scapens and Jazayeri, 2003). In this way, the work of the management accountants is not isolated, but interdependent on the whole organization. This means that the role of the management accountant can be changed by cultural needs and pressure to realign with the wider

organizational culture. Examples are changes in organizational structures, new organizational values and continuous improvement (Järvenpää, 2007). With this change, other competencies are necessary for the management accountant. Strategic, social, communication and

interpersonal competences are important competencies when having a business-oriented role (Järvenpää, 2007; Vaivio and Kokko, 2006), instead of only technical and analytical

(9)

business-9

oriented approach is the independence dilemma. This means that it is possible that the management accountant will lose a part of their independence when they get more influence (Lambert and Sponem, 2012).

2.3 Interrelationship between risk management and the roles of

management accountants

For high-quality risk management, it is important that the responsible person has an impact in the organization (Lambert and Sponem, 2012). This means that a possible influence of a

management accountant in risk management is determined by the impact this person exercises. To have influence, a management accountant needs to persuade managers. So, the

effectiveness of the management accountant is dependent on the reactions of the managers in the organization with respect to the actions of the management accountant in the area of risk management (Van Veen-Dirks and Tillema, 2017). However, both described roles still exist in different organizations and these roles do also affect the influence of the management

accountant (Lambert and Sponem, 2012).

Although Soin and Collier (2013) state that this relationship is not really clear, some studies explained the role of the management accountant and their influence in risk management. The case study by Giovannoni, Quarchioni and Riccaboni (2016) showed that when management accountants primarily have a number-oriented role, their part in risk management will be the communication to top management and performance measures, while the risk managers established measures for risk management that the management accountant can measure. These two groups consisted of different people, so the management accountant didn’t have influence on risk management. They are dependent on the measures that the risk managers do establish. However, as stated above, management accountants communicate performance data to the management. So, their influence is defined by the way they communicate with top

management and how they present the performance of the risk adjusted measures to the top managers (Giovannoni, Quarchioni and Riccaboni, 2016). This means that the number-oriented role determines that they only pass data to the management and that they do not help with decision-making.

(10)

10

participant in decision-making (Burns and Baldvinsdottir, 2005). Furthermore, a business-oriented management accountant advises other managers (Janin, 2017), what also increases their influence in the decision-making process. It is also possible that management accountants deal with both roles equally. This situation is called as “hybrid accountants” (Burns and

Baldvinsdottir, 2005). This leads, similar to the business-oriented role, to broader roles and more involvement in the organizational processes (Byrne and Pierce, 2007). Because of these

similarities, hybrid accountants are seen as business-oriented management accountants. From this, I can derive that more business-oriented roles lead to more influence in the decision-making, including decisions about risk management. Contradictory, management accountants with a number-oriented role can be highly dependent on managers according to the above described case study by Giovanni et al (2016). This lowers their influence. So, it can be argued that a more business-oriented role leads to more influence in risk management and that a number-oriented role leads to less influence. To test these assumptions, the following hypothesis is tested:

Hypotheses 1: A dominant business-oriented role leads to more influence in risk management

than a dominant number-oriented role.

Furthermore, Giovannoni, Quarchioni and Riccaboni (2016) stated that the dependence on managers (in this research risk managers) has an impact on the influence of the management accountant. The risk manager establishes the measures and makes the decisions, while the management accountants needs to apply them (Giovannoni, Quarchioni and Riccaboni, 2016). So, if the management accountant is highly dependent on a risk manager, their influence is lower than when they are independent of this manager. From this, I derived that if there is no risk manager, the influence in risk management will be higher than when there is a risk manager present. The reason is that it is expected that the risk manager has a lot of knowledge about risk management, what makes the knowledge of the management accountant less important and the opinion of the risk manager more important.

Hypotheses 2: In case there is a risk manager in the company, the influence in risk management

is lower for all roles than when there is no risk manager.

(11)

11

Figure 1 Conceptual model

3. Method

The theory testing approach is a good approach when there is literature available about the topics. However there should be a literature gap (Van Aken, Berends and Van der Bij, 2012). As can be derived from the literature above, there is literature available that indicates the possible relationships. However, there is not a lot of evidence for these hypotheses. This makes the theory testing approach appropriate. A possible data collection method is a survey, which is used in this study.

Below, the sample selection and procedure are discussed. Afterwards, the measures and their validity and reliability are discussed.

3.1 Sample selection and procedure

To disperse the survey, I have used my LinkedIn network and a contact from a university of applied sciences. On LinkedIn, I contacted 149 persons personally and I shared an

(12)

12

and responded or ignored it. So, I cannot really analyze the non-response bias. Possible respondents are professionals with functions that deal with management accounting or controlling, like the function management accountant, controller, head economic and

administrative service, financial manager and manager controlling. I chose to add more functions than only the management accountant, because these functions have different names across organizations in the Netherlands and these professionals should be included if the work is similar.

3.2 The instrument

First, I asked a few questions about function, work place and the importance of risk management in the organization to be sure the answers could be used in this research.

To rate the importance of risk management, the instrument of Sax and Torp (2015) is used. They based their instrument on the COSO Framework 2004. In this way, I was able to measure how a company deals with risk management and if risk management is important for the

organization.

With respect to the questions about general risk management in the company, the confirmatory factor analysis (CFA) showed that all 6 variables did load into one factor

(Chi-square(15)=261.182; p=0.000). The eigenvalue of this one factor is >1 and it determined

76.663% of the total variance. The variables for general risk management showed a reliability of α=0.936 when the answers are combined into one variable. So, both the validity and reliability are good. The tables are showed in the appendix.

3.2.1 The role of the management accountant

To measure the role of the management accountant, the survey instrument of Maas and Matĕjka (2009) is used. These measures refer to the work of business-oriented management

accountants (Maas and Matĕjka, 2009). I also used the instrument of Hartmann and Maas (2011). They used questions to refer to the work of the number-oriented management accountant (Hartmann and Maas, 2011).

(13)

13

score on these questions means that they have a business-oriented role. For the second five questions, the opposite is true. In this research, I used the proxy as an equally weighted average of all 5 scores per subject. This proxy is also used in the research of Maas and Matĕjka (2009). The scales for the different roles were appropriate in the study of Hartmann and Maas (2011), but the CFA computed for this sample showed that for both kind of roles

(BUS_ORIENTED_ROLE and BOOK_ROLE) two factors were needed (respectively

Chi-square(10)=67.397; p=0.000 and Chi-square(10)=40.418; p=0.000). This test showed that there are two factors with an eigenvalue>1 per role and those 2 factors determine respectively

71.257% (BUS_ORIENTED_ROLE) and 64.191% (BOOK_ROLE). However, almost all

questions loaded on both factors. This made it difficult to interpret. I chose to establish 1 factor per role, as described in the study of Hartmann and Maas (2011). After this, a reliability test is conducted. For the business-oriented role variables, the test showed α=0.638. This is high enough. If one item is kept out, the reliability decreased. In this way, all variables about the business-oriented role are combined. The variables for the number-oriented role showed α=0.662. This is also high enough and this combination raised the highest reliability too. Because of these outcomes, I kept the answers of the questions into one new variable.

3.2.2 Influence on risk management

There is no instrument available that surveyed the influence in risk management. So, I made an instrument in the same way as Maas and Matĕjka (2009) did. They searched for common practices. I did the same with respect to the influence in risk management.

As management accountants deal with risk management, the following work is exercised by them: contributing to frameworks and practices for identifying, measuring, managing and reporting risks to the achievement of the objectives of the organization (Ahid and Augustine, 2012). I used this definition, to raise questions about the influence. The definition is divided into contributing to framework/practices for identifying; for measuring; for managing and for reporting. In this way, it is possible to check if management accountants are dealing with parts of this total package.

Further tasks of the management accountants in risk management can be: providing expertise on the process of risk management; assist in the quantification of risk; analyzing risk

(14)

14

appropriate) and quantifying alternative risk mitigating strategies (Shenkhir and Walker, 2007). All these tasks are asked in the survey. To combine the answers, I used the same approach as described for the roles. So, I made an equally weighted average. The variables about influence in risk management did load into one factor (Chi-square(36)=311.208; p=0.000) when looking at the CFA. One factor had an eigenvalue >1 and this factor explained 63.317%. The reliability for the combined variable ‘INFLUENCE’ showed α=0.924.

3.2.3 Presence of a risk manager

To test the second hypothesis, it was only necessary to know if there is a risk manager present in the company. This is asked with a yes or no question. The other necessary data is asked in the questions above.

3.2.4 Control Variables

Shown expertise According to Kotter (1997), a person can get power to influence other people by showing that they are experts in the topic. So, if a management accountant has shown that he/she has expertise in risk management, they get more influence in risk management. Because this is a survey for the management accountant, I could only see his/her view on this variable. Questions were about the experience with the management of risks and how the management accountants think they are seen in the organization. According to the case study of Goretzki et al (2013), management accountants think they are seen as experts if other people ask them for help. The CFA for expertise showed that the answers on these 2 questions were not able to establish one factor (Chi-square(1)=0.006; p=0.937), so these items are used separately. Change in organizational culture A change in organizational values can influence the

(15)

15

Other control variables in this study are age, function and sector. In this way, it was possible to analyze if these factors (partly) determine their influence in risk management.

3.3 Analysis

3.3.1 Data analysis

The dependent variable in this study is the INFLUENCE. This item is measured on an ordinal scale. The independent variables are ROLE and RISK_MANAGER. The variable ROLE is converted to a categorical variable (0=business-oriented role; 1=number-oriented role). This division is made by looking at the average score on the role questions. If the business-oriented role scored equally or higher than the number-oriented role, the role is labeled as business-oriented. Otherwise, it is labeled as the number-oriented role. The variable RISK_MANAGER is also measured on a categorical scale (1=yes; 2=no). Furthermore, the control variables are categorical and scale variables.

The answers of the dependent variable are not normally distributed (see figure 2). I have tested if this distribution is raised because of outliers, but there are no outliers present in the sample (see figure 3). Because of the absence of the normal distribution, a normal regression analysis was not possible (Statistics.leard.com, n.d.). However, an ordinal regression analysis does not deal with this assumption and can be used with this kind of data (DeCarlo, 2003;

Statistics.leard.com, n.d.). This resulted in the use of an ordinal regression analysis.

(16)

16

Figure 3 Possible outliers

So, to conduct this analysis, I used the PLUM procedure. This is a method for conducting an ordinal regression analysis. Before I could start this procedure, it was necessary to check two assumptions: there should be presence of proportional odds and no existence of multicollinearity (Statistics.Leard.com, n.d.). According to Kleinbaum, Kupper and Muller (1988), there are two criteria for multicollinearity. The tolerance should be higher than 0.2 and the VIF should be lower than 10 (Kleinbaum et al., 1988). In this project, the lowest tolerance is 0.177 and the VIF is as highest 5.643. The second criterion is met, but the first criterion is not completely met. The culture questions scored a little below 0.2 (respectively 0.177 and 0.188). However, because of the good results on the second criterion, this was no real problem. Another test that is executed to make sure that the ordinal regression analysis could be used is the test of parallel lines. This test must show no significant results. When this test is not significant, it means that there are proportional odds (Strand, Cadwallader and Firth, 2018). As stated above, this needs to be present in an ordinal regression analysis. My results showed no significant results

(17)

17

Table 1 Multicollinearity

Table 2 Test of parallel lines

For the first hypothesis, I looked at the scores for the ROLE variable. The second hypothesis is answered by looking at the variable RISK_MANAGER.

4. Results

(18)

18

contacted persons. However, it is also possible that not everybody saw the message or that people who were not contacted responded. This made it not possible to give an exact response rate. However, I assumed that the response rate is lower than the computed 34.2%.

All respondents are working in the Netherlands. Furthermore, all questions are answered, so there are no missing values in the data set. In this way, it was possible to use all the data in the analysis.

4.1 Descriptive data

To give a view of the sample, I computed the descriptive data for age, sector and function of the respondents. Furthermore, the extent of general risk management in companies is described. Afterwards, the division of roles is computed and the influence on risk management is described too. Below are the descriptives for the variables that are discussed in this section.

Table 3 Descriptives

4.1.1 Age

(19)

19

4.1.2 Sector

A lot of sectors are included in this study, but most respondents are working in the sectors Industry; Financial Services and Non-Profit (Cumulative 56.9%). The category ‘Other’ came from the survey, where answers were given as Marketing, Maritime Service and Semi-non-profit. For a complete overview, see the table below.

As all sectors are included in this study, it is a representation of the Dutch work environment and the results are not completely biased with respect to a single sector.

Table 4 Sector

4.1.3 Function

Amount % Controller 46 90.19% Financial manager 5 9.81% 51 100% Table 5 Function

(20)

20

4.1.4 General risk management

Risk management is an important topic in most organizations, what is showed by the relatively high mean on the combined variable RISK_GENERAL_TOTAL in table 3. The standard deviation is also high, so the answers are spread. So, it is clear that risk management is not unimportant anymore. However, at least one respondent scored the lowest possible score. So, risk management is not important in all organizations.

4.1.5 Division of roles

For this study, I made two categories: ‘Dominant business-oriented role’ and ‘Dominant number-oriented role’. As can be seen in table 1, the mean on both variables are quite close. So, it seemed that both roles are still present in the Netherlands and it seemed that all respondents do partly work of both roles. As described earlier, to determine which role is the dominant one, I looked at which average on the role questions was bigger. If the variable AVERAGE_NUMBER-ORIENTED was bigger than the variable AVERAGE_BUSINESS_AVERAGE_NUMBER-ORIENTED, than the number-oriented role was dominant and vice versa. If both averages had the exact same answer, the respondent is added to the dominant business-oriented role.

Amount Percentage

Dominant business-oriented role 27 52.94%

Dominant number-oriented role 24 47.06%

Total 51 100%

Table 6 Division of roles

4.1.6 Influence in risk management

(21)

21

give help in the process of risk management. The computations and decisions about risk correlations and risk tolerances are seemingly less important in the work of the management accountant. It is possible that this work is done by other people. So, some work is carried out more often by management accountants than other risk management practices, what middles the combined variable.

4.2 Empirical data

At first, the correlation matrix is discussed.

(22)

22

Table 7 Correlation

4.2.1 The analysis

First, I checked the model. This is described below. After this description, I interpreted and described the results of the ordinal regression.

Table 8 Model fit

(23)

23

significant difference (Chi-square(16)=30.998; p=0.013). So, the established model with all explanatory variables fits better with the data than a model without these variables. This made the established model appropriate to use.

Table 9 Goodness-of-fit

Second, I checked if the expectations of the model matched the data. A test to do this is the Goodness-of Fit test. In this test, the null-hypothesis says that the fit is good. So, an insignificant result means that the model matches the data (Strand et al., 2018). This test showed at both the Pearson test and the Deviance test an insignificant result (Chi-square(1084)=1112.314; p=0.269 and Chi-square(1048)=267.097; p=1.000). So, the expectations of the model did fit with the data.

Table 10 Explained variance

Third, I looked at the explained variance of the dependent variable. This model explained approximately 45.7%, according the the Nagelkerke R-Square. This means that approximately 54,3% of the variance must be explained by other factors which were not included in this study. Forth, I was able to start with analyzing the results of the PLUM procedure. As can be seen in the table below, a few variables are significant. These variables are: RISK_MANAGER;

FUNCTION_1 (Financial manager); SECTOR 2 and 9 (construction and non-profit). The control variables AGE, EXPERTISE and CULTURE CHANGE are not significant, same as the

(24)

24

Table 11 PLUM-procedure

4.2.2 The relationship between role and influence

The formula of an ordinal regression is y=a-bx (Strand et al, 2018), where b is the estimate. As table 12 shows, the estimate of the business-oriented role is -0.170. But when this number is included into the formula, the –b changed into a +0.170. So, a business-oriented role adds more to the variable INFLUENCE, what indicates that a business-oriented role leads to more influence than the number-oriented role (b=0). However, the result is not significant (OR = 0.834;

p=0.801). This means that the business-oriented role does not significantly lead to more influence than the number-oriented role. So, the first hypothesis is rejected.

4.2.2 Presence of a risk manager and influence

(25)

25

4.2.3 Control variables

Age The estimate of this variable is 0.01, what is really low. However, this variable showed no significant results (OR=1.011; p=0.698). So, age is not important in this model.

Function The estimate of the function financial manager is 2.895. This means that the influence in risk management is lower for a financial manager than for a management accountant. This result is significant (OR=18.087; p=0.012). So, the variable function is important in this model. Sector Most sectors showed no significant results, except for the construction and non-profit sector. The construction sector had an estimate of 3.911 and showed to be significant

(OR=49.937; p=0.016). The non-profit sector had an estimate of 2.256 and this sector showed significant results too (OR=9.545; p=0.016). So, these sectors showed significant negative differences with respect to the influence when compared to the other sectors.

Expertise This variable is tested by two questions. The estimate for the variable previous jobs is -0.047, but this result is not significant (OR=0.954; p=0.878). The other variable asked for help has an estimate of 0.355, but is also not significant (OR=1.426; p=0.344). This means that the expertise has no significant influence on the dependent variable.

Culture change This variable is also tested by two questions. The question about culture shift has an estimate of -0.124, but the result is not significant (OR=0.884; p=0.732). The question about the influence of the culture shift had an estimate of 0.260, but this result is also not significant (OR=1.296; p=0.425). This means that a culture change has no significant influence on the dependent variable.

(26)

26

5. Discussion, conclusion and limitations

5.1 Discussion

My main aim of this research was to examine the relationship between the different roles of management accountants and their influence in risk management. Although there were no statistical studies available about this relationship, other researchers expected that this relationship exists. However, this study gave some interesting results with respect to these expectations.

According to the role of the management accountant, my study showed no significant differences between the different roles. This is contradictory to the papers of Burns and

Baldvinsdottir (2005), Giovanni et al. (2016) and Janin (2017) described in the literature review. Instead of finding that the business-oriented role has more influence than the number-oriented role, there seemed to be no significant relationship. Even the statement of Soin and Collier (2013) is not found; they expected some kind of relationship. It is possible that the researchers were looking at the wrong kind of roles. As described in the literature review, it is also possible to make the division between engaged toolmakers and compliance experts (Hall et al., 2015), what maybe better divides the management accountants with respect to influence. Another reason can be that roles are not important at all, but that risk management becomes part of the work of all management accountants. Furthermore, it can also be important what kind of ‘calculative culture’ is present. As discussed in the literature review, ‘calculative idealism’ differs from ‘calculative pragmatism’ when dealing with data (Mikes, 2009).

Furthermore, I expected that the presence of a risk manager lowers the influence of the

management accountant (Giovanni et al, 2016). This study did show that the presence of a risk manager leads to less influence in all kind of roles. So, similar to the expected influence, it does matter if there is a risk manager present. Reasons are already mentioned in the literature review, for example that the risk manager makes the decisions and raises measurements instead of the management accountant.

(27)

27

that these companies deal less with risk management than profit companies. For the construction sector, it is possible that the work is more number-oriented and that risk management is done by other people.

The perceived expertise of management accountants by other people does not significantly determine the influence in risk management. This is contradictory to the conclusion of Kotter (1997), who said that if people are asked for help they are seen as experts and that this increased their influence. It is possible that this is not important in the subject of risk management, because the subject is relatively new. Furthermore, it is also possible that

management accountants are just not seen as experts in this topic. Although Järvenpäa (2007) stated that a change in the culture can lead to more influence, this is not found in this research. Reasons can be that there were no culture changes or no changes that had influence on the work of the management accountant and how he/she is seen in this sample.

To conclude, the expected relationship in previous papers between the roles of management accountants and their influence in risk management is not found in this research. The expected relationship about the lower influence when a risk manager is present is found. So, according to my sample, the role of the management accountant does not determine the influence on risk management. However, the presence/absence of a risk manager does.

The research question “To what extent do the different roles of management accountants

determine the influence on risk management?” can be answered. The different roles of

management accountants do not determine the influence in risk management. However, other factors determine the influence. These factors are the presence of a risk manager, the function and the sector.

5.2 Theoretical and managerial implications

The theoretical implications are that the expected relationship between the roles and the

influence is not found and that researchers need to find other factors or other divisions to explain the influence. The expected relationship with respect to the presence of a risk manager is found, what can be added to the existing theory.

The managerial implications are that managers do not have to change the role of the management accountant to give this person more influence in the topic. Both kinds of

(28)

28

give the management accountant more influence when there is a risk manager in the

organization, they need to make good arrangements with these two functions about the tasks.

5.3 Future research and Limitations

Future research should further examine this topic in an exploratory way. It is possible that there are underlying processes/factors that determine this relationship, like the way of communicating by the management accountant. Further research about the different roles of management accountants is necessary; there is not a lot of literature present that explores the different roles. Researches also need to look at another division, what possibly explains the differences better. Also, the calculative culture needs to be taken into account.

Furthermore, the factors that determine the influence in risk management do also need to be examined in more detail. Most research is quite broad about influence in decision-making and not specific about decision-making in risk management. This can also be derived from this project, because 54.3% is not explained by the factors in this study. So, there should be more factors that determine the influence in risk management. If these two concepts are discussed more in-depth, researchers can look again at possible relationships.

My research has a few limitations. The first limitation is the sample of this project; it is relatively small for a quantitative study. However, because of time constraints and the specific target group, it was not possible to get more respondents. The second limitation is the sample strategy. I asked people to fill in the survey, but I also shared the project on LinkedIn. In this way, it was not sure who answered the survey and the non-response could not be measured. Next to these two restrictions, the instrument itself had some shortcomings. The survey was partly

(29)

29

References

Ahid, M. and Augustine, A. (2012). The roles and responsibilities of management accountants in the era of globalization. Global journal of management and business research, 12(15), pp. 43-53.

Arena, M., Arnaboldi, M. and Azzone, G., (2010). The organizational dynamics of Enterprise Risk Management. Accounting, Organizations and Society, 35, pp. 659-675

Bhimani, A. (2009). Risk Management, Corporate governance and management accounting: Emerging interdependencies. Management Accounting Research, 20, pp. 2-5.

Bollecker, M. (2007). La recherché sur les controleurs de gestion: êtat de l’art et perspectives.

Comptab. Controle Audit, 13(1), pp. 87-106.

Burns, J. and Baldvinsdottir, G. (2005). An institutional perspective of accountants' new roles – the interplay of contradictions and praxis. European Accounting Review, 14(4), pp. 725-757.

Byrne, S. N. and Pierce, B. (2007). Towards a more comprehensive understanding of the roles of management accountants. European Accounting Review, 16(3), pp. 469-498.

COSO, (2004). Enterprise Risk Management Framework. American Institute of Certified Public Accountants, New York NY

COSO (2017). Enterprise Risk Management – Integrating with Strategy and Performance. American Institute of Certified Public Accountants, New York NY

DeCarlo, L. T. (2003). Using the PLUM procedure of SPSS to fit unequal variance and

generalized signal detection models. Behavior Research Methods, Instruments, & Computers, 35(1), pp. 49-56

Gates, S. (2000). Incorporating strategic risk into enterprise risk management: A survey of current corporate practice. Journal of Applied Corporate Finance, 18(4), pp. 81-90

(30)

30

Goretzki, L, Strauss, E. and Weber, J. (2013). An institutional perspective on the changes in management accountants’ professional role. Management Accounting Research, 24, pp. 41-63. Gover, L., Halinski, M. and Duxbury, L. (2016). Is it Just Me? Exploring Perceptions of

Organizational Culture Change. British Journal of Management, 27, pp. 567–582

Hall, M., Mikes, A. and Millo, Y. (2015). How do risk managers become influential? A field study of toolmaking in two financial institutions. Management Accounting Research, 26, pp. 3-22

Hartmann, F. G. H. and Maas, V. S. (2011). The effects of uncertainty on the roles of controllers and budgets: an exploratory study. Accounting and Business Research 41(5), pp. 439-458

Hayne, C. and Free, C., (2014). Hybridized professional groups and institutional work: COSO and the rise of enterprise risk management. Accounting, Organizations and Society, 39, pp. 309-330

Hopper, T. (1980). Role conﬂicts of management accountants and their position within organisation structures. Accounting, Organizations and Society, 5(4), pp. 401–411. Huber, C. and Scheytt, T. (2013). The dispositiv of risk management: Reconstructing risk management after the financial crisis. Management Accounting Research, 24, pp. 88-99. Janin, F., (2017). When being a partner means more: the external role of football club management accountants. Management Accounting Research, 35, pp. 5-19

Järvenpää, M. (2007). Making Business Partners: A Case Study on how Management Accounting Culture was Changed. European Accounting Review, 16(1), pp. 99–142. Jöhnsson, S. (1996). Accounting for Improvement. London: Pergamon.

Kaplan, R. S. and Mikes, A. (2012). Managing risks: a new framework. Harvard Business

Review, 90(6), pp. 48-60.

Kleinbaum, D. G., Kupper, L. L. and Muller, K. E. (1988). Variable Reduction and Factor

Analysis. Applied Regression Analysis and Other Multivariable Methods. PWS Kent Publishing

(31)

31

Kotter, J. P. (1977). Power, dependence and effective management. Harvard Business Review,

55(4), pp. 125-136

Lambert, C., and Sponem, S. (2012). Roles, authority and involvement of the management accounting function: a multiple case-study perspective. European Accounting Review, 21(3), pp. 565–589.

Maas, V. S. and Matĕjka, M. (2009). Balancing the dual responsibilities of business-unit controllers: field and survey evidence. The Accounting Review, 84(4), pp. 1233-1253. Mikes, A. (2009). Risk management and calculative cultures. Management Accounting

Research, 20(1), pp. 18-40.

Mikes, A. (2011). Counting risks and making risks count: boundary-work in risk management.

Accounting, Organizations and Society, 36(4-5), pp. 226-245

Morales, J. and Lambert, C. (2013). Dirty work and the construction of identity. An ethnographic study of management accounting practices. Accounting, Organizations and Society, 38, pp. 228-244

Mouritsen, J. (1996). Five aspects of accounting departments’ work. Management Accounting

Research, 7, pp. 283-303

Power, M. (2004). The risk management of everything: Rethinking the politics of uncertainty. Demos.

Power, M. (2009). The risk management of nothing. Accounting, Organizations and Society, 34(6-7), pp. 849-855

Sax, J. and Torp, S. S. (2015). "Speak up! Enhancing risk performance with enterprise risk management, leadership style and employee voice". Management Decision, 53(7), pp. 1452-1468

Scapens, R. and Jazayeri, M. (2003). ERP systems and management accounting change: opportunities or impacts? A research note. European Accounting Review, 12, pp. 201–233 Shenkhir, W. G. and Walker, P. L. (2007). Enterprise Risk Management: Tools and techniques

(32)

32

Soin, K., and Collier, P. (2013). Risk and risk management in management accounting and control. Management Accounting Research, 24(2), pp. 82-87

Strand, S., Cadwallader, S. and Firth, D. (2018). 5.4 Example 1 - Running an Ordinal

Regression on SPSS. [online] Restore.ac.uk. Available at:

http://www.restore.ac.uk/srme/www/fac/soc/wie/research-new/srme/modules/mod5/4/index.html

Accessed 22 May 2018].

Statistics.laerd.com. (n.d.). How to perform an Ordinal Regression in SPSS | Laerd Statistics. [online] Available at: https://statistics.laerd.com/spss-tutorials/ordinal-regression-using-spss-statistics.php [Accessed 22 May 2018]

Statistics.leard.com (n.d.). Linear Regression Analysis using SPSS Statistics | Leard Statistics

[online] Available at: https://statistics.laerd.com/spss-tutorials/linear-regression-using-spss-statistics.php [Accessed 18 June 2018]

Vaivio, J. and Kokko, T. (2006). Counting big: re-examining the concept of the bean counter controller. Finnish Journal of Business Economics, 55(1), p49-74

Van Aken, J., Berends, H., and van der Bij, H. (2012). Student Projects. In Problem solving in

organizations: A methodological handbook for business and management students. Cambridge

University Press.

Veen-Dirks, van P. and Tillema, S. (2017). Risk Management in the Banking sector: The

influence of personality traits on the impact of Management Accountants. London: CIMA. (CIMA

Executive Summary Report; vol. 13, no. 1)

Windeck, D., Weber, J. and Strauss, E. (2015). Enrolling managers to accept the business partner: the role of boundary objects. Journal of Management and Governance, 19(3), pp. 617-653.

(33)

33

Appendix

The questionnaire

General

What is your age? __________________________________

What is your function? __________________________________

Are you working in the Netherlands? O yes O no

In which sector are you working? 0 industry

0 construction 0 automotive 0 wholesale 0 retail trade

0 catering and recreation 0 transport and logistics 0 financial services 0 not-for-profit 0 other …

Rate the importance of described risk management activities (1=really unimportant, 7=really important) Our firm has a policy for handling major risks that could affect the firm’s

ability to reach its strategic objectives

1 2 3 4 5 6 7

In our firm, we have standard procedures in place for identifying major risks and opportunities

1 2 3 4 5 6 7

Risks and opportunities are analyzed as a basis for determining how they should be managed

1 2 3 4 5 6 7

We have standard procedures in place for launching risk-reducing measures

1 2 3 4 5 6 7

We regularly prepare risk reports for the top management and the board of directors

1 2 3 4 5 6 7

We have standard procedures in place for monitoring the developments in major risks and the risk-reducing measures launched

1 2 3 4 5 6 7

(34)

34

Role of the management accountant

Please indicate how important these activities are for you in your current job. 1 = very unimportant; 2 = relatively unimportant; 3 = averagely important; 4 = relatively important; or 5 = very important

Business-oriented role

Analyzing product and customer profitability 1 2 3 4 5 Development and evaluation of investment opportunities 1 2 3 4 5

Developing new business strategy 1 2 3 4 5

Finding new ways to meet business targets 1 2 3 4 5

Developing cost saving and revenue increasing plans for the business 1 2 3 4 5

(Source: Maas and Matĕjka, 2009)

Number-oriented role

Developing internal controls and procedures 1 2 3 4 5

Ensuring that the business units observe all ﬁnancial reporting requirements

1 2 3 4 5

Developing performance reports for higher level managers 1 2 3 4 5 Assessing whether the business units observe agreements with corporate

headquarters and adheres to company regulations

1 2 3 4 5

Ensuring that business-unit managers do not spend more than strictly necessary from a corporate perspective

1 2 3 4 5

(Source: Hartmann and Maas, 2011) Influence on risk management

Please indicate to what extent you deal with these activities. 1 = never, 2 = little, 3 = somewhat, 4 = much, 5 = a great deal

(35)

35

Assist in the quantification of risk 1 2 3 4 5

Analyzing risk correlations 1 2 3 4 5

Setting and understanding risk tolerances and appetites (how much risk is appropriate)

1 2 3 4 5

Quantifying alternative risk mitigating strategies 1 2 3 4 5

Dependence (only the first question is used in the research project, the other questions seemed

to be unnecessary).

Is there a risk manager in your organization? O yes O no

In the coming questions management can be defined as a risk manager, but also as general managers if there is no risk manager.

There can be little action taken in risk management until management approves a decision;

1 2 3 4 5 6 7 1=not at all true, 7 = completely true

Controllers who want to make their own decisions in the topic of risk management would be quickly discouraged;

1 2 3 4 5 6 7

Even small matters in risk management must be referred to the managers for approval;

1 2 3 4 5 6 7

I must ask my managers before doing almost anything in this topic. 1 2 3 4 5 6 7 Any decisions I make in this topic must have the managers approval 1 2 3 4 5 6 7

Control variables

Did you deal with risk management in previous jobs? 1 2 3 4 5 Never, a little,

somewhat, often, always

Do managers ask you for help? 1 2 3 4 5

Completely disagree, disagree, not

(36)

36

Has there been a shift in the organizational culture the past years 0 yes – the culture now is better than before 0 yes – the culture is now worse than before

0 yes – the culture now is different but not better or worse than before

0 No- the culture has not changed

0 I don’t know/I do work here since a short term

If yes, do you feel that this change has positively influenced the importance of your opinion?

(37)

37

Confirmatory factor analysis

General risk management

KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .892

Bartlett's Test of Sphericity Approx. Chi-Square 261.182

Df 15

Sig. .000

Total Variance Explained

Component

Initial Eigenvalues Extraction Sums of Squared Loadings

Total % of Variance Cumulative % Total % of Variance Cumulative %

1 4.600 76.663 76.663 4.600 76.663 76.663 2 .534 8.899 85.562 3 .346 5.764 91.326 4 .229 3.809 95.135 5 .178 2.964 98.100 6 .114 1.900 100.000

(38)

38

Component Matrixa

Component 1 Our firm has a policy for

handling major risks that could affect the firm’s ability to reach its strategic objectives

.886

In our firm, we have

standard procedures in place for identifying major risks and opportunities

.941

Risks and opportunities are analysed as a basis for determining how they should be managed

.840

We have standard procedures in place for launching risk-reducing measures

.916

We regularly prepare risk reports for the top

management and the board of directors

.786

We have standard procedures in place for monitoring the developments in major risks and the risk-reducing measures launched

.875

Extraction Method: Principal Component Analysis.

(39)

39

Business-oriented role

Df 10

Sig. .000

Component

1 2.115 42.302 42.302 2.115 42.302 42.302

2 1.448 28.953 71.256

3 .720 14.394 85.650

4 .496 9.925 95.574

5 .221 4.426 100.000

Component 1 Analyzing product and

customer profitability

.396

Development and evaluation of investment opportunities

.440

Developing new business strategy

.751

Finding new ways to meet business targets

.764

Developing cost saving and revenue increasing plans for the business

.785

(40)

40

Rotated Component Matrixa

Component

1 2

Analyzing product and customer profitability

.837

Development and evaluation of investment opportunities

.870

Developing new business strategy

.733 .202

Finding new ways to meet business targets

.917 -.177

Developing cost saving and revenue increasing plans for the business

.784 .171

Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a a. Rotation converged in 3 iterations.

Number-oriented role

Total variance explained

Initial Eigenvalues

Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumulative

% Total % of Variance Cumulative %

1 2.149 42.971 42.971 2.149 42.971 42.971 1.718 34.363 34.363

2 1.061 21.220 64.191 1.061 21.220 64.191 1.491 29.828 64.191

3 .822 16.433 80.624

4 .566 11.319 91.943

5 .403 8.057 100.000

Df 10

(41)

41

Component

1 2

Developing internal controls and procedures

.521 .709

Ensuring that the business units observe all ﬁnancial reporting requirements

.683 -.344

Developing performance reports for higher level managers

.772 .346

Assessing whether the business units observe agreements with corporate headquarters and adheres to company regulations

.650 -.563

Ensuring that business-unit managers do not spend more than strictly necessary from a corporate perspective

.627

Extraction Method: Principal Component Analysis. a. 2 components extracted.

Influence in risk management

Df 36

(42)

42

Component

1 5.699 63.317 63.317 5.699 63.317 63.317 2 .954 10.598 73.915 3 .571 6.341 80.256 4 .499 5.547 85.803 5 .418 4.647 90.450 6 .286 3.182 93.632 7 .247 2.740 96.371 8 .192 2.133 98.505 9 .135 1.495 100.000

Component Matrixa Component 1 Contributing to frameworks/practices for identifying risks .828 Contributing to frameworks/practices for measuring risks .890 Contributing to frameworks/practices for managing risks .810 Contributing to frameworks/practices for reporting risks .836

Providing expertise on the process of risk management

.756

Assist in the quantification of risk .827

Analyzing risk correlations .602

Setting and understanding risk tolerances and appetites (how much risk is appropriate)

.756

Quantifying alternative risk mitigating strategies

.824

(43)

43

Expertise

Reliability

Α Questions

0.936 General risk management 0.638 Business-oriented role 0.662 Number-oriented role