The Development of the Management Accountant's Role Revisited: An Example from the Swedish Social Insurance Agency

This study traces the development of the management accountant (MA) role at the Swedish Social Insurance Agency (SIA). In 2012, the agency began a reformation by implementing the Lean management system in hopes of increasing customer trust. The results of this study show that the authority of the MA rests on decentralization and the proximity of MAs to managers, as previous research has shown, and more specifically on a definitional and a moral prerogative that may or may not be awarded to MAs enabling them to act as de facto managers. The study shows how the role of the SIA's operative level MAs changed into a helpdesk function with the role of assisting other groups to help themselves, in this case operative&#8208;level teams that had begun performing management accounting tasks. Thus, this study bears witness not to the expansion and hybridization of existing MA roles, but to the reduction in authority and de&#8208;hybridization of the MA role, from business partner to a pedagogical role on a consultative basis.

. For example, the introduction of activity-based accounting (Friedman & Lyne, 1997) and enterprise resource planning systems (Granlund & Malmi, 2002;Hyvönen, 2003;Rom & Rohde, 2006;Scapens & Jazayeri, 2003) were shown to influence the image and role of the MA. In the public sector, changes in the role of the MA have been primarily attributed to New Public Management (NPM) reforms and the ensuing attempts to create internal markets, elaborate performance measurement structures, and professionalized management. Consequently, the MA's role has been shown to have developed from focusing on operational issues to a wider scope of responsibility that includes participation in decision-making (Lapsley & Oldfield, 2000;Moll & Humphrey, 2007;Nilsson, Olve, & Parment, 2011;Paulsson, 2012).
Lately, in what some define as the "post-NPM" era (Christensen, 2014), newer private sector management concepts have been adopted by public sector organizations in an effort to curb the excesses and unintended consequences of earlier NPM reforms (Diefenbach, 2009). Among these concepts, Lean (Womack, Jones, & Roos, 1990) in particular appears to be gaining popularity (Radnor & Boaden, 2008;Radnor & Walley, 2008). Lean differs from traditional NPM practices in some important respects. It is focused on external effectiveness and value for customers, rather than internal efficiency and value for money (Maskell, 2009). Moreover, it generally implies a focus on flows that bring business units closer together, and the reorganization of employees in charge of production into Lean teams with relatively higher degrees of autonomy (Procter & Radnor, 2014).
Thus, the implementation of Lean in the public sector involves considerable changes to an organization and should presumably entail important implications for the development of the MA's role. In effect, as Fullerton, Kennedy, and Widener (2013) show, organizations tend to simplify their internal accounting systems as the implementation of Lean progresses. Moreover, as Burns, Ezzamel, and Scapens (1999), Lapsley and Oldfield (2000), Kurunmäki (2004), Jacobs (2005, and Burns and Baldvinsdottir (2007) write, the management accounting function may be appropriated by other actors in the organization. From such a vantage point, the introduction of teams with relatively higher degrees of autonomy may then induce competition over the management accounting function from yet another group. However, there is little empirical evidence on the integration of management accounting and control practices with a Lean strategy (Fullerton et al., 2013), and to date there are no studies on the role of the MA in a Lean setting, aside from the proposition, based on a hypothetical case study, that MAs are best suited to become team leaders or key team members in such a setting (Scheidt, Thibadoux, & Rosener, 2009).
In the following, we study the MA's role at the Swedish Social Insurance Agency (SIA), and how it was influenced by the current reform wave, which in the SIA's case involves the implementation of Lean. In a sense, this study follows in the footsteps of Paulsson (2012), who investigated the role of the MA at the SIA between 2005 and 2011. He asserts that although MAs perform several roles at the SIA, the inception of NPM practices fostered a coaching businesspartner role that places a premium on soft skills and thus on the MA's ability to communicate and work in teams, rather than on his or her knowledge of accounting. To increase customer trust and curb unintended consequences of earlier NPM-influenced practices, in 2012 the SIA began to reform itself again by implementing Lean as its new operational strategy. Given this new development, it is of interest to follow up on Paulsson's work by tracking the development of the MA's role at the SIA.
Nonetheless, this study is intended for a larger audience interested in the development of the MA's role and contributes to the field in two main ways. First, it contributes to research in the development of the MA's role in the public sector, given the widespread impact of NPM-driven reforms and the role MAs played in the enactment of NPMinformed practices as communicators and implementers of performance measurement and management. Second, it contributes to the conceptual understanding of the development of the MA's role, which until now has been implicitly theorized as an expansive movement of hybridization. As we attempt to show, such development builds not only on decentralization and the proximity of MAs to managers, but more specifically on a definitional and a moral prerogative that may or may not be awarded to MAs enabling them to act as managers. Moreover, we argue that the development of the role of the MA should conceptually include not only expansion and hybridization, but also contraction and de-hybridization.
In general, literature on the role of the MA describes changes in the MA's role as developing from being a provider of information or "bean counter, " to a supporter of management in its decision-making--a "business partner" (Burns & Baldvinsdottir, 2005, 2007Byrne & Pierce, 2007;Granlund & Lukka, 1997, 1998Scapens & Jazayeri, 2003). Such a development is described as a movement from a narrow scope of responsibility to produce correct accounting information about the past in a timely manner, to a wider scope of responsibility that encompasses the production and application of relevant accounting information to business decisions, and hence a focus on the future (Granlund & Lukka, 1997). Since the MA becomes an active member of the management team rather than a mere collector and processor of information, the movement has been described as one of hybridization (Burns & Baldvinsdottir, 2005, 2007. Similarly, in a study of the SIA, Paulsson (2012) shows that MAs perform a set of tasks that seem to be diversifying and expanding, thus conceptualizing them as "hybrid accountants. " Although MAs play many different roles at the SIA, Paulsson (2012) concludes that today's MAs need a broad set of knowledge and skills that include (hard) accounting and (soft) oral communication skills and the ability to work in mixed teams. He concludes that there is evidence that the influence from NPM practices at the SIA promotes a role that can be characterized as partner-like rather than counter-like. This is because the extensive responsibilities of managers under such a system require support and "it is often the MAs who provide that support" (Paulsson, 2012, p. 392). In that sense, even though they are not formally decision-makers, MAs do indeed participate in the decision-making process as members of management teams and preparers of the bases for decision-making. These results are in line with research on the MA's role, which states that organizational restructuring such as decentralization (Järvenpää, 2007) or the positioning of MAs in close proximity to managers (Burns & Baldvinsdottir, 2005) plays an important role in their development as business partners (El-Sayed & Youssef, 2015).
It is important to bear in mind that the descriptions of the different MA roles are typified ones, and what we really have is a spectrum between two stereotypes: the "bean counter" and the "business partner." A more nuanced typology has been proposed by Nilsson et al. (2011) who suggest four different roles-the accountant, the analyst, the educator, and the coach-depending on whether the MA deals with internal or external information and whether the MA's role is to analyze or to influence others. Similarly, Lambert and Sponem (2012) propose four different styles of the management accounting function-discrete, safeguarding, partner, and omnipotent-based on whether the authority of the MA is weak or strong and whether their client is local or Headquarters (HQ). Such typologies are welcome additions to the understanding of the MA's role, insofar as they enrich the spectrum of possibilities and open the door for a plural understanding of the MA's role that allows for the coexistence of different ones (Paulsson, 2012). Lambert and Sponem's (2012) findings show that the business partner role is not universal and that the "myth" of the MA as a business partner may not be widely believed. Therefore, the development of the MA's role may be better viewed as an expansion of (Burns & Baldvinsdottir, 2005, 2007Burns et al., 1999;Graham, Davey-Evans, & Toon, 2012) rather than a transformation to (Granlund & Lukka, 1998) a [conceptual] role with local variations. Nevertheless, the descriptions to date of the development of the role of the MA remain predominantly incremental and unidirectional, from "bean counter" to "business partner" (El-Sayed & Youssef, 2015), perhaps largely because empirical evidence reflecting any fundamental changes from the business partner role is still scarce (Hyvönen, Järvinen, & Pellinen, 2015).
Not only has the role of the MA expanded and become a hybrid between accounting and management, as shown in previous research, but the management accounting function itself has also been appropriated by other professions such as managers and others that are also becoming hybrids (Burns & Baldvinsdottir, 2007;Burns et al., 1999;Jacobs, 2005;Kurunmäki, 2004;Lapsley & Oldfield, 2000). For example, Kurunmäki (2004) shows how the introduction of accounting practices in the medical profession brought on the hybridization of medical expertise. In turn, Jacobs (2005) argues that instead of hybridization, which would include all medical experts, it may be better to understand the appropriation of management accounting tasks by professionals other than accountants as a process of polarization. There, a separate and limited "sub-group emerges to manage financial and administrative responsibilities, leaving the fundamental values and practices of the wider profession unchanged" (Jacobs, 2005, p. 135).
Research into the role of MAs has also shown that new management accounting instruments and techniques play an important role in the development of MA functions and responsibilities (Ezzamel & Burns, 2005;Hyvönen, 2003;Järvenpää, 2007;Rom & Rohde, 2006;Scapens & Jazayeri, 2003). For example, Newman and Westrup (2005) observed that the relationship between MAs and technologies such as enterprise resource planning systems are increasingly intertwined and that MAs use their position to exert and retain control of information technologies, thus reshaping their professional expertise. They also noted that when MAs neglect their responsibility to provide information/data for developing new systems, there's a risk that vendors, management consultants, and shareholders will do so instead, thus making the systems more designed for their needs than the MA's needs. Ezzamel and Burns (2005) illustrate in turn the presence of competition among various professions, including MAs, over the management accounting function and the failure of MAs to retain control over the studied firm's procurement/purchasing functions.
In sum, the expansion of the role of the MA into a business partner role involves participation in management teams and, therefore, the privilege of participating in decision-making. Since MAs prepare information as bases for decisionmaking, such a privilege can be characterized as definitional. The hybridization of other groups in relation to MAs' tasks and the competition between groups/professions that seems to ensue over the MA function suggest, however, at least the possibility that expansion is not the only avenue of development for the role of the MA.

METHOD
This article draws on a 2012 to 2015 study of the SIA and traces its latest reform, which includes Lean-inspired changes such as a renewed customer focus, the inception of self-managing teams, and the implementation of Lean as the SIA's operational strategy. Therefore, it is reasonable to expect that these changes will have had an impact on the role of MAs that research on Lean and management accounting has not previously studied, other than in hypothetical terms, and to argue that MAs are best fitted to be team leaders or key team members (Scheidt et al., 2009).
Although we welcome Paulsson's (2012) study of the SIA and its contribution to the discussion about the increasingly complex role of the MA, specifically in the NPM context, we see limitations in such findings due to the methodological approach used. The material was mainly collected through a web-based questionnaire and complemented by a number of interviews with MAs and managers in a time of relative stability. In contrast, we chose to study the MA's role mainly through qualitative semi-structured interviews in a time of organizational change, enabling us to pose questions and collect reflections about the social order of both the new emerging work situation and how it differs from the previous one (cf. Czarniawska & Sevón, 1996). This design enables us to nuance the analysis of the MA's role at the SIA both before and after the introduction of Lean as the overall business strategy of the SIA. Secondly, while Paulsson (2012) restricted the choice of interviewees to MAs and managers at different organizational levels, we included managers, specialists and, in particular in this article, caseworkers.
Seventy-seven interviews were conducted between 2012 and 2015: 27 interviews with managers from top management, support functions, and the three customer divisions, as well as MAs, representatives from the Lean department and central projects, and 50 interviews with operative-level employees. The interviews were based on discussion themes of a general nature rather than specific questions. A typical interview would begin by asking the interviewee to describe his or her job in general terms and, more specifically, how the management accounting function is conducted within the SIA. Subsequently, questions were posed about changes in daily work and whether or not they had experienced any challenges. The interviews were conducted in Swedish, lasted between 45 and 90 minutes, and were recorded and transcribed. In addition, documents such as the letter of appropriation, annual reports, the SIA's staff magazine, and internal documents were used as empirical material.
Although all 77 interviews fed into our understanding, specific attention was paid to 31 interviews, conducted primarily in 2013 and 2014, for the analysis of the MA function at one of the customer divisions of the SIA, that is, the Local Insurance Centers (LIC), which are responsible for complex social insurance cases that require extensive contact with customers. As Paulsson (2012, p. 391) remarks in his study, the SIA is a large organization in which the role of the MA varies. MAs at the regional and local levels, he writes, "are often members of the management teams and, thus, participate in meetings where most decisions are made," while MAs at the central level work more as analysts and are not members of a management team. Therefore, this study concentrates on the regional and local levels by focusing on the MA's role at the LIC. Among the 31 selected interviews, eight were performed with MAs at different organizational levels, and five at the operational level; five interviews were with line managers, three with managers at the central HR department, and 15 interviews were with caseworkers (see Appendix). At the time of our interviews, all 31 interviewees had completed the Lean training program designed for all employees of the SIA and were organized into Lean teams, meaning that they were more or less in the same phase of Lean implementation.
Before we account for our findings, however, we provide a brief presentation of the studied organization, the SIA, and its latest reform.

CHANGES AT THE SIA
The SIA is a politically governed organization with around 13,000 employees. Since 2005 it has undergone extensive transformations-the merger of 21 regional social insurance agencies followed by another major reorganization-in order to create a modern, efficient, and service-oriented agency. The latest change in this succession of events is the implementation of the Lean-inspired "Common Journey" in 2012.
Lean was introduced at the SIA to curb the negative consequences attributed to preceding reforms, that is, low levels of customer trust and lack of employee empowerment due to a strong internal focus on efficiency and control, detailed and quantitative-oriented performance management as well as the neglect of the customers' perspective. In that vein, internal quantitative targets, previously set and broken down by MAs, were banned and replaced by three customer-value propositions, namely that the customer should always feel (1) that s/he is met with respect and understanding, (2) that the SIA makes his/her life safer, and (3) that the SIA makes things simple for him/her. The propositions were set up to guide the daily activities of the also newly implemented self-planning teams, whose customer focus was to be developed during Lean meetings and continuous improvement practices. Moreover, an extensive Lean training program was offered to all employees, the organization was reorganized into value-flow-oriented processes and the practice of continuous improvements was introduced and implemented by way of a coaching leadership.
One of the purposes of the measures was to loosen the previous authoritative, top-managed, centralized, and rulebased leadership and control structures that characterized the agency, in which MAs, who were responsible for elaborate performance measurements and reports heavily reliant on production data, had obtained a prominent and authoritative role. Both MAs and managers, who had developed a close relationship, were thus expected to take a step back and develop a "supporting role, " and allow caseworkers, now organized into teams, to plan and evaluate their own work to a greater extent. In the following we account for the role of the operative level MA and the role of caseworker teams in relation to management accounting practices. Since there are no concise, quantified goals, indicators, that is, customer satisfaction, customer values and customer-value propositions are combined to determine how performance could be improved. Measuring customerrelated values, as opposed to production statistics, is however no easy feat: several MAs felt the response rate was low and applying local knowledge to the whole was difficult.

The role of the MA
Other methods were tried to complement customer surveys. Focus groups, for example, were used locally in some cases in an attempt to understand customers. The problem, as one MA states, is that although the members of the focus group are usually interested in the discussion, which leads to a better and deeper dialogue, the number of individuals/groups that can be interviewed is limited, thus making it difficult to generalize from such a set of data. Another MA spoke of the possibility of auscultation by managers as a method of evaluating the work caseworkers perform in meeting the customer. The problem, of course, is that meetings with customers are sensitive situations and having a manager watching over the officer could be difficult for both the officer and the customer. Thus, measurement practices are still omnipresent, but how they are used and who the user is has changed with the implementation of Lean. In effect, as another MA pointed out: We stress the analytical side much more [now]. We talk about goals and results in a different way now and try to see what more we can learn from what we see… and to try to create insight and knowledge based on it. (MA 4, May 2014) In sum, the MAs' authority to provide clear and concise advice has changed as the demand for elaborate calculations of production data has decreased, along with an increased demand from top management for evaluations of customer satisfaction, which are more difficult to produce and more complex to analyze.
As developed below, the management accounting function at the SIA is no longer solely in the hands of MAs. It has spread to other groups such as the Analysis and Prognosis department, which is responsible for the administration of large databases with statistics over the SIA's operations in relation to the customer. Additionally, it has also spread to the Insurance Processes department, which is responsible for the development and standardization of all control processes regarding the benefits provided by the SIA. The next section concentrates on the operative level and a group that has also begun to compete with MAs over the management accounting function: caseworker teams at the local offices of the LIC, that now also produce data on customer values and customer satisfaction.

Lean teams with increased responsibilities
As part of the Lean reform, caseworkers were reorganized into teams of eight to 12 people with a "collective task description and goals" (Försäkringskassan, 2012b(Försäkringskassan, , 2015. As such, the teams were awarded greater autonomy and thus also control on their own activities. Caseworkers were expected to distribute tasks within the team, decide what kind of information they need to perform their job, and to produce such information locally. This means they were also involved in management accounting. Moreover, as expressed below, they were expected to work with quality issues and competence development.

Through team-based work and leadership, the conditions are improved for the integration of production-oriented
and insurance-oriented management control as well as for ensuring that quality, equality and timely decisions are applied equitably. Our mission is visualized and becomes clearer. The teams are empowered and given more responsibility to actively work with quality issues and to develop their competence. (Försäkringskassan, 2012a) To produce and assess the information needed and perform their work as a team, teams made use of several Leaninspired tools in their day-to-day work. Lean boards, for example, were used as an accounting and control tool to visualize and support continuous improvement through the active participation of employees in weekly meetings, where improvement suggestions were discussed. In effect, a Lean board consists of three major sections: (1) goals, measurements, and trends, (2) daily control, and (3) prerequisites (Försäkringskassan, 2012b(Försäkringskassan, , 2015. The first section of the board, "goals, measurements, and trends, " builds on the agency's vision and mission. It is visualized by a matrix where goals and indicators for the specific team are illustrated. It includes graphical descriptions of the team's longterm development, expected in-flow of errands, capacity, and risks that might affect production. The second section of the board, "daily control, " is divided into two subcategories-results according to plan and also according to actual outcome-including in and outflow of errands and reviews of different quality aspects, such as results from customer surveys, bottlenecks, and waste. To keep track of the team's production, some teams create their own reporting systems. A team coordinator describes it thus: We have created a file to have a general view of everything we have so that we will be able to calculate if need be, because we have a production goal every month and we want to be able to follow it up and know that we're doing enough to reach it, that we have this or that much and that we'll have to do this or that many cases. (Caseworker

18, May 2014)
The third section of the Lean board visualizes the team members' presence and absence, a "traffic light" indicator showing the "health status" of the team and the work environment, an agenda for the daily Lean board meeting, an action plan illustrating what, who, and when an improvement suggestion should be made, the "rules of the game" for the team and an illustration of the need for competence development.
Teams at the SIA also use the "A3 template, " another Lean-based tool. The A3 is a specific form that teams are required to fill out suggesting and arguing for a specific improvement they wish to recommend. Put simply, the idea of the template is to ensure that all proposed suggestions are thought through in a rational way, with well-defined problems and solutions as well as calculations of possible gains following a suggested change. To prepare and fill out such a form the team has to perform systematic measurements and calculations to prepare an argument that justifies the proposed change. A caseworker, who is also a Lean coach, explains: Nowadays we write an A3. We write a detailed statement […] what the problem is and the solution […] All applications for changes have been accepted although we have removed some of them because we realized they weren't effective. But it's good to have tested them because now we know they don't work. (Caseworker 19, May 2014) Several interviewees confirmed that preparing an A3 that requires defining, calculating, and explaining the improvements a suggestion might achieve is difficult and time-consuming.
According to both the interviews and internal documents (cf. Försäkringskassan, 2012aFörsäkringskassan, , 2012bFörsäkringskassan, , 2015, the introduction of Lean-inspired teams has also had an impact on the role of the MA. An MA describes it as follows:

Now that we are getting into Lean […] we do not need to support them [teams] that much. The idea of Lean is that local groups should do the work themselves. That is why it [not being stationed at local offices] has not been a problem. (MA 8, May 2014)
The role of MAs in relation to local offices, and specifically teams, is now to support the teams by providing the conditions and basis for them to manage, control, and develop. Thus, an important task for MAs is to help team members develop their skills and competencies in order to become self-managed and self-controlled, and develop their own indicators, in particular indicators regarding production planning. According to internal documents, the overall aim should be to "strive to involve the team as much as possible in the collection and visualization [of data] to promote the team's ownership of the information" (Försäkringskassan, 2012a(Försäkringskassan, , 2015. In practice, this means that the MA needs to work in ways that help each team reflect on what specific indicators are needed, how they are related to the national goals of the agency, and how relevant data could be collected and produced (Försäkringskassan, 2012b).

You can present numbers in different ways and what matters is how we talk about the numbers. We can't say
that ten customers had to wait more than 300 days and then say 'take care of it' as we did before. Rather, you can ask an open question […] about what the cause may be. (MA 2, May 2014) In that sense, the role of the MA could once again be described as more of a helpdesk technical advisor on a consultative basis, whose aim is to help teams release the knowledge they have within themselves.

DISCUSSION
As Paulsson (2012) shows, NPM practices promoted an increased degree of participation of MAs in the regional and local managerial decision-making process because the focus on managers and their responsibilities created a need for support of the kind MAs can provide. Before the introduction of Lean in 2012, the role of the MA could thus be characterized as partner-like rather than counter-like at the local and regional levels of the SIA (Granlund & Lukka, 1997, 1998Vaivio, 2006). Although MAs were not formal decision-makers, they did indeed participate in the decisionmaking process as members of various management teams. However, as our study shows, the role of the MA business partners at the SIA was not limited to participating in the decision-making process as members of the management team. They also played an operations management role insofar as they were enabled by the targets they set together with management to be the judges of performance. Several interviewees provided examples of situations where MAs, who continuously made reference to quantitative targets, had become highly authoritarian in the daily operations.
The business partner MA was thus, in effect, also an operations manager inasmuch as s/he could hold other groups accountable for failing to comply. The organization of the SIA's management accounting function at the regional and local levels, which led to the inclusion of the MA in the management team and consequently also the technocratization of management, can thus be said to have been based on two prerogatives awarded to MAs: a definitional prerogative in their role as members of the management team and preparers of data for decision-making purposes, and a moral prerogative in their role as assessors of unit and individual performance. These two prerogatives provided authority and consequently the basis for the MA's role to be hybridized (Burns & Baldvinsdottir, 2005, 2007 with the management function.
The introduction of Lean as the operational strategy of the SIA, along with other changes referred to as the SIA's Common Journey, have since changed the role of the operational level MA. In effect, MAs, as directed explicitly by the SIA's Director General, are not expected to manage, but only to provide support. MAs at the SIA no longer form part of management teams, thus seeing a decrease in their ability to influence future action. Neither do MAs enjoy the means to assess performance, as specific goal targets have been replaced by vague customer-value propositions and, therefore, cannot function as the means to ascertain right from wrong and evaluate organizational and individual action as good or bad. In other words, MAs at the SIA lost the definitional and moral prerogatives that enabled them to participate in directing future action and manage performance. Instead, operational level MAs act as a supportive unit called upon on a consultative basis to serve as discussion partners who possess information about the state of the SIA at a higher level of abstraction than the everyday practice of the operational unit.
As Nilsson et al. (2011) write, the MA's role can be more bean-counter-oriented as an accountant or an educator who clarifies strategies by formulating targets, or more business partner-oriented as an analyst or a coach who provides business advice. However, in this study teams appear to have taken over the role of the accountant and perhaps even that of the analyst at the local level. Moreover, because they are self-managed, teams have been awarded the definitional prerogative to define problems, to take action within the confines of the team, and to exercise the moral prerogative of judging their own performance. MAs, in turn, are educators; they do not formulate targets but help teams find the answers among themselves in order to learn to perform on their own. They are also coaches, but almost as outsiders, because they are no longer part of management teams and are located apart from regional and local offices.
This distances them even further from both managers and teams; they are organized as a helpdesk that provides consultative services on demand, but they do so without having clear answers to give, that is, without the definitional prerogative to define the problem at hand by breaking down the strategy into numbers, or the moral prerogative to assess right and wrong. What is left for the MA in this "democratization" of the management accounting function are the technical and pedagogical prerogatives to support on a consultative basis from afar.
The introduction of Lean as the operational strategy of the SIA can be understood as an impulse to change the MA's role as business partner, which was considered by many to be too authoritarian, into a new consultative helpdesk role that is customer-oriented in the sense that the MA's task is to support its customers (other groups at the SIA) in their work with end-customers. In doing so, the customer perspective that was introduced with Lean, when translated to the role of the MA, means that local and regional units become the customers of the MA unit almost as if the management accounting function had been outsourced. Such an impulse appears to involve a de-hybridization of the MA's role through the expropriation or re-appropriation by management of the management function, and a contraction of the MA's role through the appropriation of the management accounting function by other groups at the SIA, in this case teams. In effect, as decentralization develops further and customer orientation is internalized and permeates the organization, techniques such as customer surveys and focus group interviews will be used more and more to manage and control the operations of the SIA through teams. MAs at the regional and local level find themselves in a situation where they have to compete with other groups in the organization over the performance of the management accounting function (Burns & Baldvinsdottir, 2007;Burns et al., 1999;Jacobs, 2005;Kurunmäki, 2004;Lapsley & Oldfield, 2000). This follows the efforts of the SIA in its Common Journey to regain the trust of the customer by way of implementing Lean as its operational strategy. In effect, the absence of clear-cut goals limits the ability of MAs to judge and evaluate because there is no longer an index to good and bad performance. Instead, action must be grounded in reflection, given the overall goal of the SIA and its customer value propositions. Without the possibility of defining specific goals with which to stipulate and communicate what is right or wrong, the MA is left with an open-ended task that consists of helping operational units and teams to help themselves, and ultimately to learn how to do the job themselves.
In sum, the authority of the MA at the SIA is not only limited by the reorganization of the management accounting function away from management teams and decentralization to teams now endowed with the prerogatives once held by MAs, but perhaps more importantly, by the whole business framework provided by Lean. The Lean framework includes customer orientation and general goals that constitute an obstacle to the qualification of good and bad organizational performance by MAs. This change in the role of the MA shows not only, as previous research (Friedman & Lyne, 1997;Granlund & Malmi, 2002;Hyvönen, 2003;Rom & Rohde, 2006;Scapens & Jazayeri, 2003) has asserted, that reform is an important element in the changes to the MA's tasks as is decentralization (Järvenpää, 2007) and positioning of the MA in relation to management (Burns & Baldvinsdottir, 2005). It shows also that the repertoire of available roles for the MA to play is expanding, and, again more importantly, that the development of the MA's role need not be unidirectional from "bean counter" to "business partner" but can also fluctuate back and forth across the spectrum.

CONCLUSION
The results of this study show that the authority of the MA rests not only on decentralization (Järvenpää, 2007) and the proximity of MAs to managers (Burns & Baldvinsdottir, 2005), but more specifically on a definitional and a moral prerogative that may or may not be awarded to MAs enabling or disabling them to act as managers. Moreover, our study provides an example of how decentralization, if taken far enough, can lead to loss of authority among MAs. Finally, the results of this study show that the development of the MA's role need not necessarily be incremental or unidirectional, because the authority bestowed upon regional and local MAs at the SIA, which increased during NPM-influenced reforms (Paulsson, 2012), has now decreased as a consequence of the Lean-influenced reform. A deterministic development of the role of the MA from bean counter to business partner to helpdesk advisor is, however, not proposed.
Instead, the role of the MA is viewed as a malleable concept, an idea that is already latent in the notions of expansion and hybridization (Burns & Baldvinsdottir, 2005, 2007Burns et al., 1999) if contraction and de-hybridization are included as implicit conceptual possibilities of such developments. Consequently, the number of roles that an MA can take on is multiple and variable in the sense that, in time, the MA's role may not only expand and hybridize, but also contract and de-hybridize, through the appropriation and expropriation of tasks.