Cost allocation systems for inter-organizational
shared service centres:
An interpretative approach to the development and
the adoption
Kasper Kuiper
University of Groningen
Faculty of Economics and Business
MSc Business Administration
Organizational & Management Control
Supervisor: Drs. R. Minnaar
Second Assessor: Prof. Dr. J. van der Meer-Kooistra
June 2013
Word count: 16.923 (excl. appendices)
Friesestraatweg 115 9718 NE Groningen
(06) 48831820
1
Abstract
This thesis contributes to literature by providing insight into the development process of inter-organizational cost allocation systems and aims to reveal the motives of decision makers on why certain cost allocation systems are adopted. By means of the framework of Kwon and Zmud (1987) are the initiation of the development process and the adoption of the developed cost allocation system placed in a theoretical context, to illustrate that the scope of this thesis is only focused on a small fragment of the complete implementation process. Data is provided by a single case study, conducted in the role of developer of a cost allocation system for three cooperating municipalities. Using an inductive line of reasoning is the empirical data, in the role of researcher, translated into theory. This thesis demonstrates that the development and the adoption are not two separable processes, since motives for adoption can be found already in the development process. It is concluded that involving the decision makers in the development process, results in a more positive stance from the decision makers towards the process and the outcome of the process, by creating a sense of ownership and a reduced anxiety towards the outcomes.
Keywords: cost allocation; shared service centre; inter-organizational cooperation; development;
2
1. Introduction
Organizations operating in the public sector are confronted with increasing institutional pressure. This pressure, stemming from political influential regulations (Ashworth, Boyne & Delbridge, 2009; D'Aunno, Succi & Alexander, 2000; Tsamenyi, Cullen & Gonzalez, 2006), legal requirements (Ashworth et al., 2009) and societal demands (Degadt & Herck, 2003), forces these organizations to effectuate better governance and better services with less resources (Stroobants & Bouckaert, 2011) and to create more efficient management control and governance structures (Orazem & Vodopivec, 2009; Roberts & Greenwood, 1997; Townsend & Wilkinson 2010). The establishment of shared service centres (SSC) is one way organizations choose to accomplish this (Bergeron, 2003; Janssen & Joha, 2006; Schulman, Dunleavy, Harmer & Lusk, 1999; Strikwerda, 2005).
In extant literature, different definitions for SSCs can be found, but they all share the same basic elements. In this thesis, a combination of the definitions used by Bergeron (2003, p. 3), Minnaar and
Vosselman (2013, p. 76), Quinn, Cooke and Kris (2000, p. 7) and Schulman et al. (1999, p. 9) is used: “An SSC is an autonomous organizational entity which provides noncore/facilitative services for
-more than one unit (can be divisions, business units or even organizations).”
Transient to the ex-post advantages of establishing SSCs, like cost savings, improved service levels and an increase in efficiency (Bergeron, 2003; Schulman et al., 1999), SSCs and especially the phase of setting up the SSC entails some difficulties. A priori the establishment of an SSC, were each entity has its own service centres, all costs of the regarded service centre can directly be allocated to the concerned entity. When organizations make use of SSCs, the costs this centre makes has to be allocated among more than one entity. ”In these situations the problem arises how to divide among the participants the joint costs, which result from the cooperation” (Tijs & Driessen, 1986, p. 1016).
1.1 Accuracy of cost allocation systems
To be able to answer how to divide the joint costs, a digression to equity theory is made. This, because cost allocation can be seen as an application and tangible form of equity theory, which is demonstrated by the following comparison; Equity theory is concerned with an equitable distribution of benefits (Huppertz, Arenson & Evans, 1978) and cost allocation is concerned with an equitable distribution of costs. Where equity theory considers distributive justice, the desire of the involved members to have a fair and just distribution of profit (Adams, 1963), cost allocation considers the distributive justice of costs. Equity theory states that actors investing in a collaborative effort should benefit from this, relative to the investment they make (Burnstein & Katz, 1972). Cost allocation in its turn, must be designed in a way that the actors (applied to SSCs), who make use of the shared services, should be billed relative to their burden on total costs (Bhimani, Horngren, Datar & Foster 2008; Edmonds, Tsay & Olds, 2009). From these parallels can be derived that the criterion for a cost allocation, to be qualified as ‘accurate’, is equity preservation.
3
Inaccuracies in allocating costs for inter-organizational SSCs can apart from the aforementioned problems, also lead to over- or understated total costs and in turn to an unjust profit or loss to the entire company.
Accuracy of a cost allocation is thus essential, but also very difficult. As pointed out by Zimmerman (1979, p. 519): “It is unlikely that cost allocations are perfect.” Or more recently acknowledged by Bhimani et al. (2008, p. 137): “Finding answers to cost-allocation questions is difficult. The answers are seldom entirely right or clearly wrong.” In practice it is even harder, where also secondary considerations, such as easiness to understand or to calculate, play a role (Tijs & Driessen, 1986).
1.2 Best practice in allocating costs
The contrast between the high importance of a cost allocation being right and the difficulty of making it right, has shown to be very intriguing for scholars. This can be noticed from the variety of approaches towards this problem taken, the different models to allocate costs developed, and the broad timeframe in which research towards this topic is conducted.
For example, Shapley (1953) developed an allocation method based on marginal costs; this method sees the formation of a coalition of partners as a sequential process where the partners enter one by one in the coalition. This method determines what the marginal costs for the coalition are when a new partner enters the formation. The new partner is held responsible for these marginal costs (Tijs & Driessen, 1986). This method is further explored and improved by, among others, Dubey (1975), Young (1985) and Driessen (1985).
The cost gap method, developed by Tijs and Driessen (1986); this method, a game theoretic approach, calculates which partner in the alliance is responsible for what proportion of the separable costs and uses this factor to allocate the remaining non-separable costs over the partners.
Frisk et al. (2010) used yet another approach. They tackled the problem of allocating costs, by allocating not the costs themselves, but by allocating the savings arising from the coalition. The aim of this method is to divide the total savings of the cooperation equally over the partners and is therefore called the ‘equal profit method (EPM)’. The cost allocation is thus made, based on the initial costs (before entering the partnership) of the partners and the total savings of the coalition. Several more allocation methods tend to exist, from which the ‘direct method’, the ‘step-down method’, the ‘reciprocal method’, ‘methods that use only one cost driver’, ‘methods that use several cost drivers’ and ‘the determination of a service’s cost price’ are only a small portion (Bhimani et al., 2008; Cooper & Kaplan, 1988; Curien, 2003; Edmonds et al., 2009; Rogerson, 1997; Rubio-Odériz & Pérez-Arriaga, 2000).
4
Besides historical specificity, there is also the existence of ‘local specificity’ (Simon, 1976). According to local specificity, the inimitability is preserved because of the fact that methods of working are adjusted due to local (within firms or within firms their coalitions) learning processes and cultural differences (Simon, 1976).
Where historical specificity states that methods are modified to fit the changing environment, due to environmental, exogenous factors, local specificity argues that methods are adapted due to internal insights or events, such as unforeseen complications in using the cost allocation method (Anderson, 1995). In other organizations, were they may not have these complications or compete in another environment, these historical and local specific methods will not fit and will turn out unsatisfactory (Barney, 1991; Simon, 1976).
The variety of methods to allocate costs, in combination with the concepts of local and historical specificity recognized in literature, leads to the adoption of the statement that there is no such thing as a ‘best practice’ in developing cost allocation systems. We can only speak of ‘best for the situation’ (Shepard & Hougland, 1978) or ‘local best’ (Becker, 2004) solutions. This is endorsed by Tijs and Driessen (1986, p. 1016), who state the following: “One cannot expect a final answer to the question: What is the best cost allocation method? The choice of a method depends on the concrete situation where such a cost allocation problem arises, the ideas of the participants with respect to fairness, the power feelings of the participants, the difficulty of understanding and calculating a cost allocation proposal, and many other factors.”
That the absence of a best practice for cost allocations is cited, lies in the fact that the usage of a best practice in building a new system promotes the legitimization of this system (Roberts & Greenwood, 1997) and thereby the willingness to adopt it (Doorewaard & Bijsterveld, 2001; O’Leary, 2000; Wagner & Newell, 2004).
1.3 Development and adoption
This discussion emphasizes the importance of the accuracy of a cost allocation system, the impossibility of making it completely accurate and the absence of a best practice in making the cost allocation. Actors involved in a new cost allocation system will be very critical before adopting new systems. But the system might be unable to completely preserve equity and the rationale that a best practice is used in developing it, will not be the reason for legitimizing adoption either.
Extant literature provides knowledge about the factors which promote successful implementation of cost allocation systems (Anderson, 1995; Shields, 1995), but lacks understanding of what makes decision makers willing to implement it in the first place. It does not indicate how the process from development to adoption occurs.
5
To provide structure in exploring this pinpointed gap in extant literature, a research question is provided. This research question serves as a leitmotif and assessment framework, to test whether actions during the research are relevant to fill the void (i.e. will the considered action bring us closer
to finding an answer). Finding an answer to the research question:
How is the process towards adoption of an inter-organizational cost allocation system shaped and why is a specific cost allocation system adopted?
Will give insight into the rationales, the considerations of decision makers in adopting new systems.
This knowledge also supports future developers of cost allocation systems, in developing their system and getting their work appreciated through adoption.
1.4 Outline
This thesis does not follow the classic structure of research papers. In order to not give a predetermined fixed direction in conducting the research, there are no hypotheses or a priori constructs developed. “This allows the researcher to let concepts emerge from the data rather than being constrained by previous theory” (Randall & Mello, 2012, p. 866). Instead, this thesis uses an inductive line of reasoning which is also known as bottom up logic (Huitt, 2003), which “is begun as close as possible to the ideal of no theory under consideration and no hypotheses to test” (Eisenhardt, 1989, p. 536). Therefore, following the structure of Crane (1989, p. 564-5), this thesis is “afforded some initial direction by the research question set out previously, which is mainly in terms of providing some guidance on what kind of data to collect rather than representing any form of research proposition or hypotheses”.
The second chapter, therefore, differs from typical research papers. Where usually extant literature is used to develop propositions or hypotheses to guide the research, is in this thesis the extant literature used to provide a background and context to the empirical part only.
6
2. Research Placement
As stated in the introductory chapter, the research conducted in this thesis takes place between the initiation and the adoption phase of a new cost allocation system. In this chapter, an elaboration is made on these phases, in order to define them and to explain where these phases are situated in relation to the complete implementation process of a new system. This elaboration is conducted against the background of the framework of Kwon and Zmud (1987).
2.1 The three-stage model of organizational change
The chosen framework to provide this background and context, the Kwon and Zmud (1987) framework is one out of many frameworks based upon the three-stage model of organizational change by Lewin (1952). Thereby, the Framework of Kwon and Zmud (1987) is not even the first or the most recent among them. Though, there are several reasons why this framework is chosen to show were this thesis can be positioned in extant literature. In order for this reasons to become clear and to justify this choice, the reasoning starts at the very beginning; by explaining the three-stage model of organizational change (Lewin, 1952) and by providing an answer to the question why there are so many frameworks based upon the three-stage model of Lewin (1952). From there, the motives to choose for the framework of Kwon and Zmud (1987) are given.
According to Lewin, A successful change project, involves three steps (Burnes, 2004, p. 9-10):
Unfreezing: Lewin believed that the stability of human behaviour was based on a
quasi-stationary equilibrium. He argued that the equilibrium needs to be destabilized (unfrozen) before old behaviour can be discarded (unlearnt) and new behaviour successfully adopted.
Moving: Carry out the change.
Refreezing: Refreezing seeks to stabilize the group at a new quasi-stationary equilibrium in
order to ensure that the new behaviours are relatively safe from regression.
The three steps of Lewin’s three-stage model (1952) were at first instance based upon changes in human behaviour. In extant literature, the models that came forth out of this model are not based solely on the change process of human behaviour anymore (Weick & Quinn, 1999).
To show how big the influence of Lewin’s three-stage model (1952) in today’s extant literature actually is and for what processes of organizational development it is used, an overview of the several models, for which the three-stage model is used as a basis, is provided (table 1).
7
Table 1: overview of three-stage model of organizational change (Lewin, 1952) based frameworks
Why the three-stage model (Lewin, 1952) had and has such an impact on organizational development literature is clarified by Hendry (1996, p. 4): “Scratch any account of creating and managing change and the idea that change is a three-stage process which necessarily begins with a process of unfreezing will not be far below the surface."
Model STAGES/PHASES
Change Model (Lewin, 1952)
Unfreezing Change Refreezing
Stages of Innovation Adoption (Becker & Whisler, 1967)
Stimulus Conception Proposal Adoption decision
Two stage Innovation Adoption Model (Zaltman, Duncan & Holbeck,
1973)
Primary adoption firm level decision
Secondary adoption actual implementation
Organizational Innovation Model
(Pierce & Delbecq, 1977)
Initiation Adoption Implementation
IT Diffusion-Procedural Model (McFarlan & McKeeney, 1982)
Identification and investment
Experimentation, learning and adaptation
Rationalization and management control Widespread technology transfer
Six-stage model for innovation implementation (Kwon & Zmud, 1987)
Initiation Adoption Adaptation Acceptance Use Incorporation
Technological Diffusion Perspective (Cooper & Zmud, 1990)
Initiation Adoption Adaptation Acceptance Routinization Infusion
Innovation Adoption (Rogers, 1995) Knowledge of innovation Attitude towards innovation Adoption Decision Implementing Confirmation of decision
Four Phase Innovation Adoption Process (Darmawan, 2001)
Initiation Adoption Implementation Evaluation
Innovation Adoption and Implementation (Gallivan, 2001) Primary authority adoption decision Organizational assimilation
Organizational acceptance and consequences
Organization Innovation Adoption
(Frambach & Schillewaert, 2002)
Awareness Consideration Intention Adoption decision
Continuous use
User acceptance
Six-stage diffusion process (Xu &Quaddus, 2005)
Initiation Adoption Pilot implementation Organic growth Organizational implementation Diffusion Technology Implementation Process
(Kim & Garrison, 2010)
Agenda Setting
Matching Adaptation Redefining/ Restructuring
8
The simplicity and the accuracy of Lewin’s model (1952) thus led to the generic characteristic of the model. It can be applied, explicated and modified to fit all kinds of organizational change, such as the implementation of a new innovation (Becker & Whisler, 1967; Darmawan, 2001; Gallivan, 2001), new technology (Cooper & Zmud, 1990; Kim & Garrison, 2010), new IT systems (McFarlan & McKeeney, 1982) and even knowledge management systems (Xu & Quaddus, 2005).
2.2 Kwon & Zmud framework
Of all these frameworks in extant literature is specifically chosen for the Kwon and Zmud framework (1987) because of two reasons. The first is that they provide a model which consists out of six steps, which makes it among the other six-step models, the most detailed implementation model. Using such a detailed implementation model indicates most specifically which part of the complete process this thesis is focused upon and that it is only a small fragment of the entire process. The second reason to choose for the Kwon and Zmud framework (1987), the choice over the other six-step models, is because they were the first; the studies which provided other six-step models of organizational development (Cooper & Zmud, 1990; Frambach & Schillewaert, 2002; Kim & Garrison, 2010; Xu & Quaddus, 2005) only created slightly different terminology for the same phases. At first sight these models are the same as the Kwon and Zmud framework (1987). It is appropriate to take the first created framework, since this framework proved itself a base for further research.
In the framework of Kwon and Zmud (1987) they linked the three-stage model of organizational change (Lewin, 1952) to the traditional criteria of IT implementation success (acceptance and usage), to model the implementation process as six sequential stages (Anderson, 1995: 5).
The model of Kwon and Zmud (1987) is comprised of the following six stages (Anderson, 1995; Xu & Quaddus, 2005):
Initiation: Internal needs, institutional or competitive pressures initiate the change and a
search for solutions begins.
Adoption: A proposed solution is chosen and the decision to invest resources is made. Adaptation: Unforeseen needs or shortcomings of the system are identified and resolved. Acceptance: This stage represents the minimal level of use and maintenance that the new
system requires to be sustained.
Use: The complete set of work practices is replaced by the new system; the new system
becomes a part of normal activities.
Incorporation: The system is infused when it is used in unanticipated ways. To improve work
effectiveness and is seamlessly integrated with other organizational systems.
2.3 Scope
9
By focusing on the phase between the initiation and the adoption - From the start of the search for solutions, until a solution is chosen - this paper attempts to provide that depth. The process that takes place between the first two milestones is identical to the gap in literature as indicated in the introduction. By focusing on this one particular phase, a rich description about the development of a cost allocation system can be provided and except from providing answers on what influences the development process and reaching the milestone of adoption, a thorough description can be made also on the ‘how’ and the ‘why’ of these influences.
In figure 1, a model is provided which summarizes the foregoing elaboration of the focus of this thesis and the comparison with extant literature.
F O R E G O I N G L I T E R A T U R E Initiation Adoption Adaptation Acceptance routinization infusion T H I S T H E S I S E X T A N T L I T E R A T U R E Initiation Adoption Development Adaptation Acceptance Use Incorporation T H I S P A P E R
Figure 1. Theoretical position of research
10
3. Methodology
For this thesis, with the aim to shed light on what happens before the actual implementation of a cost allocation system, a single case study at the heart of the development of an SSC for inter-municipal cooperation is conducted. This chapter describes why this research method is chosen and how it is executed in the empirical part of the research. Thereby, a description of the case, the Dutch municipalities ‘Ooststellingwerf’, ‘Weststellingwerf’ and ‘Opsterland’, is provided.
3.1 Research methods
Conducting this research in the form of a case study is appropriate since this paper attempts to explain the research object in depth (Bryman, 2012). As stated by Flyvbjerg (2006, p. 236): “The proximity to reality, which a case study entails, will constitute a prerequisite for advanced understanding.” Thereby, the objective of this research is learning, building theory, rather than proving existing literature, this makes a case study the obvious choice for execution of the research (Eisenhardt, 1989; Eisenhardt & Graebner, 2007; Flyvbjerg, 2006; Yin, 1994).
Since this research builds on uncivilized ground (i.e. not studied before), a single case is the preferred way to conduct the research (Yin, 1994). The use of a single case enables the empirical part to be described very richly (Weick, 2007), which matches the goal of the research; to give detailed insight into what happens between the initiation and the adoption phase of a cost allocation system and to expose the decision makers their rationales in adopting such a system. This rich description contributes to the persuasive power of the research (Siggelkow, 2007).
Opponents argue that the lack of generalizability is a weakness of single case studies (Bryman, 2012; Flyvbjerg, 2006). Indeed, there is no formal generalizability with a predetermined sample size, statistically calculated to be large enough to enable generalization of the findings to the entire population. However, this thesis aims to contribute to (first) insights in the diversity of situations and problems. The knowledge about this specific situation can in its turn contribute to the overall knowledge about this topic (Jansen, 2005). This phenomenon is called analytic generalization (Yin, 1994) and is praised as “the force of example” (Flyvbjerg, 2006, p. 228).
11
3.2 Data collection
Data is collected by the author during four months of intensive fieldwork between February and June 2013. The author is during this period employed by all three of the municipalities, it is part of the job to develop a cost allocation system and to get it adopted. This dual role of employee and researcher enables analysis of the negotiations by being in the centre of the negotiations that lead to the forthcoming cost allocation system. This closeness leads to an intimate sense of things, "how they feel, smell, seem" (Mintzberg, 1979, p. 587). The strength of this closeness is that “the intimate interaction with actual evidence often produces theory which closely mirrors reality” (Eisenhardt, 1989, p. 547).
To execute the job of developing the cost allocation system, three other developers of cost allocation systems are consulted. These persons developed a cost allocation system in the past for inter-municipal SSCs and were asked for advice. To intertwine this consultation in the role of developer with the role of researcher, this advice was given in the form of a narrative interview. A narrative interview is a form of interviewing which is open and unstructured, this moves the interviewee to speak freely. Since there is no interference of the interviewer, this often leads to stories, for which they are called narrative interviews. In this way is the interviewee able to give an image of the event as concrete and detailed as possible (Arend, van der, 2007).
The advantage of a narrative interview is that statements are made in a detailed description of the context. This provides except from the statements itself, also the motives for giving the answer. “A story in a narrative interview is not just a listing of events, but an attempt to link them both in time and meaning. It is the plot that gives coherence and meaning to the narrative, as well as providing the context in which we understand each of the events, actors, descriptions, goals, morals and relationships that usually form a story” (Jovchelovitch & Bauer, 2000, p. 59). A narrative interview thus provides a comprehensive and coherent picture of the view of the respondent.
Since gathering data for the development of the cost allocation system for the OWO SSC takes four months and these narrative interviews only between one and one and a half hour; the information gathered from the narrative interviews will stand out meagre against the data from the four month during fieldwork. Therefore is chosen to not handle them as separate cases, but to incorporate them in the single case. This is appropriate because of the fact that these interviews, their advice and their experiences also contributed to the forming of the cost allocation system and are actually part of the single case. These narratives are handled as so called ‘embedded cases’ (Perry, 2000).
3.3 Case description
The case study started in February 2013 and evolved around the development of a cost allocation system for an SSC which has to deliver back office tasks to the Dutch municipalities
‘Ooststellingwerf’, ‘Weststellingwerf’ and ‘Opsterland’ in the near future.
12
The municipalities Ooststellingwerf, Weststellingwerf and Opsterland lay in the southeast corner of the province of Friesland in the north of the Netherlands. The population and the size of the three municipalities are provided by table 2.
The emphasis of this case study lies on neither one of the three municipalities in particular, but rather on the cooperative bonds between them. At the time the research is conducted, the municipalities are developing an SSC for the back office tasks. More specifically for the department ‘business operations’, which consists of the financial administration, planning & control, computerization, automation, personnel and salary administration, internal control, procurement, taxes and the insurance activities. For a visual representation of the organizational change within and among the municipalities, see appendix 2.
The development of this SSC is accompanied by the development of a cost allocation system. The development of this cost allocation system is where this research is focused upon. Other information than regarding the development of the cost allocation system, given in the case narrative (chapter 4), is cited to create context and to support understanding only.
3.4 Data presentation
The development of this cost allocation system is executed by (among others) me, in the role of employee rather than in the role of researcher. Therefore, the information gathered about the development process is not only first-hand, I am the first-hand source. To enable the reader to experience the development and the adoption of the cost allocation system as I experienced it, the subsequent chapter (the case narrative) is written in the I-form. The entire listing of events, occurrences and experiences that shaped the system and led to adoption are given from my (developer) perspective. This follows Eisenhardt & Graebner (2007, p. 29) on how to write up empirical evidence in a single case study: “By presenting a relatively complete rendering of the story within the text.”
The fact that the results of the case study mainly represent my own interpretation of the events, entails a weakness. This weakness is already reduced by the aforementioned triangulation of research methods, however not completely overcome. Therefore a regular member check, already during the phase of gathering notes and during the writing of the narrative, is applied to “decrease the chance of misinterpretation” (Krefting, 1991, p. 219) and to “establish the validity of researchers’ interpretations of the collected data” (Sandelowski, 1993, p. 4).
Municipality Population Surface (km²) Number of villages
Ooststellingwerf 25.847 226,64 13
Weststellingwerf 25.643 228,36 26
Opsterland 29.865 227,68 16
13
3.5 Data analysis
Subsequently is the narrative, which can be seen as a summary of all the raw data, arranged into groups. Following Marshall (1999), is this done by studying the narrative repeatedly and by considering possible meanings and how these fit the developed group. The distinguished groups are held against the raw data to verify whether no evidence is overlooked, if the groups embrace all relevant data and if there is no presence of possible conflicting evidence. Based on this process is determined whether refinement or supplementation of the narrative or the groups is desirable. In the discussion chapter (5), the narrative is fragmented and the distinguished groups are held against extant literature which is sought for convergent and conflicting findings in order to abstract the results and enable this thesis to contribute to extant literature. This intertwining of the narrative with literature, “demonstrates the close connection between empirical evidence and emergent theory” (Eisenhardt & Graebner, 2007, p. 29).
Due to the inductive line of reasoning applied in this thesis, there is no theoretical frame provided to which the findings can be held to conceptualize them. In effect, literature provided in the discussion chapter may not be mentioned before in this thesis, but is applied nonetheless if it fits the findings of the research retrospectively.
14
4. Case narrative: Developing the OWO cost allocation system
4.1 Initiation
The first day of the project started with a kick-off meeting, where documentation and an explanation were provided about the reason, the structure and the goal of the project. This kick-off meeting was held to get to know each other and to exchange expectations, rather than to get some work done. The documents and conversations about the project provided a clear description of why this project was launched:
Municipalities are on the move and have to cope with large governing tasks in the coming years. Such as the decentralization of government tasks, governmental cuts, ageing and dejuvenation of the population, shrinkage of the economy and the economic crisis. Also increasing demands on the conditions of service and digitalization play an important role. Cooperation and up scaling are widely seen as a way to achieve governing power and to cope with these challenges. For this purpose, the municipalities Ooststellingwerf, Weststellingwerf and Opsterland have the joint ambition to seek more cooperation at the strategic level. (Project plan, appendix 3)
For this cooperation we need to divide the costs between the three municipalities. Dividing cost is in itself easy, you can just split the cost in three and let all three of the municipalities pay their share. However, we have the least amount of employees and we are not going to pay for theirs! There will be more differences between the municipalities like that, which we have to take into account when we divide the costs. (Team leader BO, Ooststellingwerf)
The internal needs, a cost allocation system which takes the differences between the municipalities into account, were presented. The search for solutions, the development of the cost allocation system, has begun. The milestone of initiation, depicted from the framework of Cooper & Zmud (1987), is reached.
4.2 Development
4.2.1 Entering the field
After the kick-off meeting, I was introduced to all employees of the business operations (BO) department of Ooststellingwerf separately. This was done in a very elaborate manner and towards every employee in the same way:
This is Kasper Kuiper, a top talent from Groningen; he supports us in allocating the cost for the OWO-cooperation. He is attracted to provide a fresh perspective, new ideas and a critical view
on how we do things in here. (Sr. financial specialist, Ooststellingwerf)
Following this introduction, I told something about myself, my education and my working experience. The reactions of the employees were all very positive, a grasp:
15
During the rest of the day, which was mostly spend to read project documentation, the head of the department came by. After we started talking, the team leader came to join the conversation. This is in itself not startling, but the way how this was done was indeed. As if he wanted to make sure that I was introduced properly, since again the terms ‘top talent’, ‘fresh perspective’ and ‘new ideas’ were used.
From that moment on I was intrigued by the term ‘top talent’. It is too much of a coincidence that the two persons used this high expectation creating term and emphasized it separately from each other, for a person they actually do not know well. I asked about this reputation after I found the term again in written form:
It is planned to make use of a so-called Top Talent to provide support for this project (Project plan)
The status of top talent is a separate track, these kinds of externals are hired for the more challenging projects. The status is based on background and the recruitment for these kind of externals is done by a specialized agency, to make sure the best are selected. (Team leader BO, Ooststellingwerf)
The project documentation also provided an overview of the persons who are involved in the project (table 3). In this structure the working group does the research, produces information and provides the first recommendations on how to allocate the cost towards the project group. The project group in their turn discusses these findings and give either the working group the task to further work out specific questions or develop a set of recommendations themselves, based on the recommendations of the working group and their discussion. This set of recommendations moves up to the steering committee, which decides to adopt the recommendations on the cost allocation system given by the project group or let them adjust the cost allocation system according to their comments.
Steering Committee (decision makers)
Department head Business Operations, Opsterland (delegated client)
Department head Business Operations, Ooststellingwerf Department head Business Operations, Weststellingwerf
Project Team (developers)
Project Group Team leader Planning, Control & Administration (PCA), Weststellingwerf (project leader)
Team leader Business Operations, Ooststellingwerf Department controller Business Operations, Opsterland
Working Group Senior financial specialist, Ooststellingwerf Policymaker planning & control, Weststellingwerf Financial counselor, Opsterland
16
This overview made immediately clear why the team leader BO of Ooststellingwerf put so much effort in introducing me with the vigorous terms towards the department head; she was our decision maker.
4.2.2 Gathering information
After spending the first week in the office of the municipality Ooststellingwerf, I was transferred to Weststellingwerf for further work on the project:
“You need to work in all three of the municipalities during the project. Otherwise, if you will only stay here, the others will think that you are employed by us. They will prejudge the eventual cost allocation system, by feeling that the chosen methods will be in favour of Ooststellingwerf. We cannot have these prejudgments, it will bias the decision making. Thereby, you work for all three of us, so it is not even the case.
(Team leader, BO Ooststellingwerf)
Except from the fact that they wanted me to work from all three of the municipalities to:
Get to know the faces behind the numbers (Department controller, Opsterland)
Show that you really work for all three of the municipalities (Team leader, Ooststellingwerf)
Get all the information you need (Team leader BO, Ooststellingwerf)
Better get to know the team members (Team leader PCA, Weststellingwerf)
It provided also great opportunities. When I was stationed at one of the three municipalities for the first time, all the team leaders of the BO departments made sure that I had a proper introduction at the department. Everyone responded well to my presence, they all offered help or in the case I would be in the need of information:
Just come by / I am happy to give it / Let me know. (Employees BO)
All the people who might had some useful information, to accommodate the building of the cost allocation system, knew who I was, what I did and where they could find me. In the beginning this provided a massive flow of information, employees which were not even in the project team gave it intentionally and unintentionally without me even asking for it. Intentionally:
You might want to look into the tax differences, I know for a fact that there are differences in the dog tax. (Employee BO, Ooststellingwerf)
Unintentionally given information came from so called ‘water cooler conversations’:
The advantage of becoming stationed in Weststellingwerf is the hypermodern building over there. (Employee BO, Opsterland).
17
Sure, I will look into that right away / I will bring it at your desk in about ten minutes / I will call ….. for you, she knows this better than me. (Employees BO)
However, after these two weeks it was like people lost their interest or motivation in providing me with information from their own initiative. So I had to ask for it more often, this was in itself not a problem, but I noticed a shift in reactions when asking for information as well. Now, when I needed something, the responses were along the lines of:
I am kind of busy at the moment / Next week I have some time to do that / I think someone else can help you better. (Employees BO)
When this phenomenon occurred again for the third time in my third week in Opsterland, it could not be a coincidence anymore. It probably was not the culture of the municipalities, since these differed quite well between the municipalities (a special working group for the identification of the differences in culture between the municipalities is established). It was neither the case that my information was saturated, I was in the need of a lot more. The department controller of Opsterland was asked if he had a clue why this happened, he joked:
It is already normal to have you around here, maybe you integrated too well. (Department controller, Opsterland)
A lot of events immediately crossed my mind, even though he was not serious, he was completely right; in Opsterland I was stationed in the office of the planning & control team and most of the times I was the earliest in. I turned on the lights in the office, except for one day. In the second week he (Department controller) was already settled when I came in, he said:
I was confused this morning, it was all dark and was wondering where you were. Normally you are already working when I come in. (Department controller, Opsterland)
In Weststellingwerf, when I came in for an appointment one and a half week after I worked there for four weeks, they asked:
Where were you all this time? We missed you.
Yet another event that crossed my mind was the image of myself inflating balloons for a ‘colleague’ her birthday. After around two weeks they saw me no longer as the ‘top talent’ that came in, but as a colleague, part of the team and it was normal for them to have me around. The water cooler conversations were no longer a source of information either; they only provided me with information about people their weekends and evenings. Were my physical presence at all of the municipalities worked as a great advantage in the beginning, after a while it became a drawback.
This problem was overcome by having a very dedicated project team (table 3), consisting of employees with formal leadership in the organizations. They urged their colleagues to provide the information when needed.
4.2.3 Project structure
18
The formal structure, as sketched in table 3 and the paragraph above the table, differed from what the actual structure looked like after a few weeks; two changes caused this shift:
1. I had risen in the hierarchy
- In the minutes of the first official meeting with the working group and the project group is
stated the following:
The project group has to make clear agreements with the steering committee at what moments reporting on the progress is desirable .
However, I suggested to the steering committee how this could be done. - In the beginning, as already mentioned, I was introduced in the following way:
This is Kasper Kuiper, he ‘supports us’ in allocating the costs.
Later, when I joined a meeting with persons involved I had not met before, I was introduced as following:
This is Kasper Kuiper, he supports us in allocating the costs of the OWO cooperation.
Well, actually, he leads the project. (Policy office P&C Weststellingwerf)
- Together with the Project leader, I was the delegate from the project team in the meetings
with the steering committee and in the meetings with the client. 2. The project group merged with the working group:
- Due to the full agendas of all of the working and project group members it was nearly impossible to gather the teams for a meeting on short notice. To overcome this, the project group and working group were merged. We kept meetings if every municipality had at least one representative, in this way meetings could be scheduled even if one or two members were not able to fit it in their schedules. This merger was endorsed by the project leader (member check), from where she continued:
That is being innovative within a project, you should never pin yourself down on a structure. There is something written, but this is something you have to see as a starting point; you have to develop it yourselves in order to make it work.
(Team leader PCA, Weststellingwerf)
The actual structure of the people involved in the project looked like follows: Steering
Committee
Department head BO, Opsterland (delegated client) Department head BO, Ooststellingwerf
Department head BO, Weststellingwerf Project
Team
Team leader PCA, Weststellingwerf (Formal project leader) Top talent intern (Informal/ deputy project leader)
Team leader BO, Ooststellingwerf Department controller BO, Opsterland Senior financial specialist, Ooststellingwerf Policymaker planning & control, Weststellingwerf Financial counselor, Opsterland
19
An advantage of climbing the ladder was that I now was contacted directly by the steering committee. Where in the beginning I was only aware of the information the project leader provided me, now I was able to hear their concerns, thoughts or their agreement first hand. This gave a more complete and richer picture of this information. Metaphorically speaking, at first I only saw the tip of the iceberg, now I was able to see (partly) under water.
Also the merging of the groups had an advantage, except from the fact that planning of meetings became easier. Due to this arrangement, the working and the project group came to a set of recommendations together, in contrast to the formal structure where the working group had a developer/decision maker relation with the project group. We got rid of a hierarchical layer in the project, which made the development less viscous and more direct.
4.2.4 Current system
In developing the cost allocation system, the client had given us the following guideline:
The new system is presented and explained in terms of improvements to the current system. (Project plan)
This ‘current system’, is used for an already established (small scale) SSC for the municipalities of Ooststellingwerf and Weststellingwerf (OW system). In order to live up to this guideline it was vital to know the OW system very well. A detailed analysis of invoices, contracts, and annual overviews occurred for that purpose. Thereby, an (narrative) interview was conducted with the developer of the OW system:
When I asked him the question whether he could tell the story of how the development of the cost allocation system for the OW cooperation occurred, he started with a reprimand:
I make an important distinction between the concept of allocating and equalizing costs. With equalizing is meant to balance the costs, to make an approximation of the division of the costs. Allocating costs means that everything is calculated exactly to the last comma. These terms are perceived different, it reflects the method used.
I am persistent in the statement: No, it is not an allocation it is an equalization of costs, allocation is much too mathematically.
(Deputy director, Weststellingwerf / Developer OW system)
He continued:
In the development of the system I sat together with the deputy director of Ooststellingwerf. We oriented us on how other municipalities handled the distribution of the costs and asked two consultancy firms and the ADM (association Dutch municipalities) to inform us about what these kind of systems look like. We noticed that there were very extensive systems, while our own guideline was to keep it simple. Our directors therefore allowed us to experiment.
20
One of the specific principles of the system we came up with, was the establishment of an equalization reserve. We did this so we would not have to go to the municipal councilfor assent on every small deviation from the budgeted costs.
When we eventually presented our OW system, we got a lot of negative comments from administrative personnel who are rather precise. They said: ‘We can’t do this, we need to monitor and administrate precisely to whom we charge the actual costs.’ This was too simple for them, people who are financially literate said that this system will not work properly. But still we said ‘let us try it anyway, see if this can work’.
When we showed the overviews of 2012 to the managers of Opsterland, they were pleasantly surprised. It is only a small piece, though there are no big differences between the budgeted and the actual costs; so we do not need a new system for the OWO cooperation necessarily. I wanted to keep it that way, but new game, new chances. (Deputy director, Weststellingwerf)
The experiences with the current system were good, though there are nine persons involved in the development of a new system; I asked him why?
The directors of the municipality of Opsterland said: ‘We do not want the feeling that we are zipped in on what Ooststellingwerf and Weststellingwerf constructed, we want to look at it on our own as well. We take note of the fact that the OW system works out well, but we actually want to start with a clean sheet.’
We understand that you do not want to take everything for granted when you step in, you want to look at it yourselves. However, we said that they have to keep the experience we gained in mind. I need to become convinced of changes! (Deputy director Weststellingwerf)
Since the developer of the OW system, as temporary department head BO of Weststellingwerf, was one of the decision makers in our cost allocation system, we called the project ‘project equalization’ to not raise preliminary resistance against our eventual cost allocation system. However, it raised questions later on with another member of the steering committee:
Why do we call it equalization now, I thought we called it allocation. When I see the term equalization, this feels to me like retroactive allocation of cost (with Dutch meaning of the terms), but this is not what we are doing. We look forward! (Department head Opsterland)
Ergo, it was a cost ‘allocation’ system again.
4.2.5 Other inter-municipal cost allocation systems
To broaden our perspective on possible cost allocation systems, rather than using only the OW system as starting point, developers of already established cost allocation systems for other inter-municipal SSCs were consulted. Following, a summary of how this process went for the cost allocations systems of the ISZF1 and the DEAL2 is given. The focus lies on the most signatory and the most remarkable comments the developers had about their processes.
1
SSC ICT cooperation agreement South west Friesland (Netherlands)
2
21
ISZF
We first agreed upon the fact that we wanted to cooperate, because we were too small and we foresaw an increase in costs. The choice for an SSC came only afterwards. We started working together without having a cost allocation system, it was developed ‘on the fly’. I cannot recommend that, you will make mistakes. On the other hand that, is how I work/ who I am.
The bigger consultancy firms produce cost allocation systems from a theory book, they want to implement the most precise systems available. But we as cooperating municipalities are in learning paths, it makes no sense to achieve the most ambitious system immediately. This will not fit in the development strategy of a municipality, it does not fit in their perception of what is right. In my opinion you can better consult another municipality. We did things wrong and other municipalities did things wrong ánd right, therefore there is a lot of knowledge at neighbouring municipalities who already made a cost allocation system.
There was no such thing as adoption, I just told that this is how we are going to do it. You can calculate everything to the penny, but keep looking each other in the eyes.
(Interim director, ISZF)
DEAL
The cost allocation system for the DEAL municipalities was developed very objectively. In the preparation for the development of a cost allocation system I received from the directors information about other municipalities’ cost allocation systems. On basis of that information and my own expertise, I developed five scenarios, ranging from very easy to really complicated. These five scenarios were discussed and fine-tuned with employees from the participating municipalities.
Both the scenarios were already worked out, so the fact that they did not chose the scenario I suggested did not lead to more work for me. When I looked back at it, I developed a multiple choice question, which they answered; every answer was fine with me.
The most precise and also most complicated method, cost-plus pricing, does not seem the right choice at the moment.(Consultant, expertise: financial support/ planning & control)
In addition, there were three other developers of inter-municipal cost allocation systems consulted. They were willing to provide information about the content of their systems only, not specifically for the development and adoption processes (time constraints). Oftentimes this provided information about the process as well:
On parts of the development of the distribution variables and the cost allocation system, we hired an external agency. Eventually the steering committee decides and that is what you have to deal with. (Department head resources, BMW)
The municipalities chose to bear the cost of the startup phase equally, every municipality one third of the initial cost. (Department head finance & control, BEL-combination3)
3
22
We make use of a distribution of cost according to the proportion of hours worked for the municipalities. In the beginning we brought the overhead separately into account, but later we incorporated it into the hourly rate we charged the municipalities for the services. This was more complicated, but also more transparent. (Adviser planning & control, Bel-combination)
At the moment we allocate the costs in proportion to the population count of the municipalities. It is conceivable to charge per product or program, this leads however to a lot of bureaucracy.(Interim director, ISMH4)
We chose for charging in proportion to the population count, since this is a simple manner to allocate the costs of a cooperation. It also prevents cherry picking by individual municipalities. (Interim director, ISMH)
We searched in our network for how other municipalities arranged it, but our cooperation consists of a lot more (10) municipalities; you cannot compare that with others. (Interim director, ISMH)
When we discussed the aforementioned inter-municipal cost allocation systems (on content) within the project team, we found that none of these fit into our thoughts on how to allocate the costs. Even though there were some interesting and useful findings:
The egg of Columbus is not among them. (Policymaker P&C, Weststellingwerf)
4.2.6 Product breakdown structure
We had the knowledge of how the costs are allocated between other municipalities, how it is done in the current (OW) system and how these systems work out in practice. In addition, there is a lot of financial expertise in the project team.
For the actual execution of the project we made a product breakdown structure (PBS), the costs of the SSC were split into different cost categories which we evaluated separately. When we eventually handled all the categories, we could bundle our conclusions and recommendations and we had the total product, the cost allocation system OWO (Appendix 3). This was also the way we suggested to report the steering committee on the progress, by category.
Good explanation, informing us and identify bottlenecks in between. (Department head, Ooststellingwerf)
Good approach to discuss conclusions and recommendations with us by theme. (Department head, Opsterland)
Except for the fact that the steering committee now knew how often and in what form they were kept up to date, i.e. they knew where they stand, it gave a grip on the development of the system for the project team as well. We first worked out the first cost category and discussed it with the steering committee before we continued to the second cost category. We figured that if the steering committee agreed with the recommendations in between, or gave us feedback which we processed in the system, it would be very unlikely for them to not adopt the bundled recommendations and the complete cost allocation system.
4
23
Again metaphorically speaking, we saw the cost allocation system as a book we wrote for the steering committee. If they liked every chapter separately, there is no reason to not like the book as a whole. It is much less inconvenient to rewrite a chapter, than the complete book.
We also noticed that the meetings of the project team became highly efficient since they were scheduled according to the distinguished categories. In the first meetings we got on multiple side tracks, e.g.:
A cost baseline measurement / Cost containment / Variance analysis. (Minutes meeting)
After the establishment of the PBS, only one topic, one cost category, was discussed per meeting. Although this was not explicitly stated, all the members hold on to this structure. When someone was about to deviate, he or she was rebuffed:
We spoke about that last week, we agreed upon to do it like…… (Sr. financial specialist, Ooststellingwerf)
No, that are friction costs, we discuss that next week. (Department controller, Opsterland)
The structure of the meetings was further directed by the guideline of using the OW system as a starting point and the information gathered from the other municipalities. We started every meeting with a briefing of how it is done in the OW system and how other municipalities arranged it. Following, we discussed whether we agreed upon the current system, if we could use the information of the other municipalities or if we were able to come up with something better:
Most of the times, after discussing it, there was nothing left of the current system and we often made no use of how other municipalities did it. However, by using this information as a starting point, we had lift off for the discussion. Sometimes it is hard to answer the question ‘where do you start?’, we did not have that problem. (Financial counselor Opsterland)
However, when the emerged recommendations on how to allocate the costs were presented towards the steering committee, the existence of the current system made it harder. Especially, the developer of the current system had a hard time in accepting recommendations which deviated from the current system:
The current agreement (OW system) is …., I do not see why we would do it different now. (Department head, Weststellingwerf)
Actually, the project team did indeed:
I do not see the use in charging the municipalities by the route of an equalization reserve, it is only a supplementary action in the administration. (Financial counselor, Opsterland)
He does not see that it is not working as well as he thinks it does; upstairs, they only see the pieces of paper and not the misery we get from it.(Policymaker P&C, Weststellingwerf)
24
4.2.7 Practice vs. politics
However, as came forward later, this difference in perception did not arose from not willing to see the flaws of the current system, but from the political perspective the steering committee took in judging the recommendations concerning the cost allocation system (these were taken into account in the OW system). The project team wanted to keep the administrative burden as low as possible, since they had to eventually work with it. The steering committee however, wanted to develop the system in order to make it politically work as well, i.e. to not have to explain every choice they made to the council; they wanted to use the cost allocation system partly as an instrument to create boundaries in which they could autonomously do what they think is right for the functioning of the municipalities/SSC. They called this a “scratching area” (Department head Opsterland)
The project team knew that the steering committee could have political motives right from the beginning but ignored this at first:
We cannot take their political motives into account when we develop the system, we look at it from a financial perspective. We develop the allocation system from what is sensible and logical from a financial point of view; they can always choose to do it differently. (Department controller, Opsterland)
But after their motives for several disagreements on our recommendations came explicitly forward (the political motives), the project team figured that they could develop it from a financial perspective only, or make it work politically as well. Therefore the steering committee needed to cooperate; the steering committee stated for example that they wanted to charge the municipalities through an equalization reserve, this however was according to the project team not the right way to delegate:
Instead of goals (scratching area), they give us instruments (equalization reserve). They do not even tell immediately why they care so much about that particular instrument. If we know what they want to achieve with it, then maybe we can fine-tune the instrument or come up with a better instrument. They have to give us their motives and targets, so we can deliver a system which works for everyone. (Project leader)
They go too much into the content in our organizations, we are the people with financial expertise. They have to provide a frame only, we fill it in. I can come up with already two ways in which this scratching area can be created, which are less bureaucratic than the equalization reserve is. (Financial counselor, Opsterland)
On these proposed alternate solutions and also most of the other subjects, the members of the project team were very much on the same wavelength; in a way that it was almost too good to be true, so I asked about it:
25
We only need enough expertise to see what is feasible and what is not. It needs to be simple, but good. If you struggle over details, it will not benefit your future as colleagues” (Project leader)
When the steering committee noticed that we provided suggestions which solved their concerns, their approach towards us changed as well. From that moment on they became more transparent and clearer on what they actually wanted; the iceberg started floating. We started to work together and several more dilemmas on which we got stuck in discussion initially, were solved with for every party satisfying outcomes.
4.2.8 Dividing the costs
Eventually, apart from all the other topics addressed in the project, the actual distribution of the costs over the three municipalities is the very existence of the project.
The steering committee as well as the project team wanted to keep this simple “if appropriate”.
(Team leader BO, Ooststellingwerf)
We have to see if we can just divide the cost in three equal parts, or that it needs to be done differently. (Department head, Ooststellingwerf)
I just want to split the cost in three, that is the easiest way. (Policymaker P&C, Weststellingwerf)
Is that the real representation than? (Top talent intern)
We are not looking for the digits after the decimal. (Sr. financial specialist, Ooststellingwerf)
I think that dividing by three is the most convenient. (Project leader)
But what if it is actually 30/30/40? Or what if the steering committee asks how we came up with dividing by three, do we say that it is the simplest way then? Is that our
argument? (Top talent intern)
They agreed upon the fact that only simplicity as argument will never hold. They wanted to know the real proportions, but were anxious for the outcome:
I am curious to find out what will come out of the comparison. (Policymaker P&C, Weststellingwerf)
What if the proportions differ completely? (Sr. financial specialist, Ooststellingwerf)
Therefore we took “the starting point to divide the costs simply in three equal parts and to only
deviate from this distribution if the proportions were in reality completely different.” (Minutes meeting)
To get insight in the real proportions between the burden on the costs of the SSC, we had to go beyond simplicity; i.e. we had to do a complicated analysis in order to determine whether we could distribute the costs in a simple manner.
Eventually the result of the analysis, consisting of seven distribution variables with different weights,
26
All members of the project team stated that they wanted to divide the cost equally over the three municipalities:
It will probably cost us more if we have to determine the precise percentages every year over and over again, than the difference in cost between the municipalities actually is. (Department controller, Opsterland)
4.3 Adoption
Four months and one week after the project was initiated, in as it turned out the last meeting with the steering committee, the project team presented the bundled recommendations and the developed OWO-cost allocation system. After the presentation we prepared for another discussion round, but the only thing the steering committee said was:
Okay, I guess you can start work it out in excel and implement it into the financial system. (Department head BO, Opsterland)
The proposed solution on how to distribute the cost of the SSC for the municipalities of Ooststellingwerf, Weststellingwerf and Opsterland was chosen. The milestone of adoption, depicted from the framework of Cooper & Zmud (1987), is reached.
In my role as developer of the cost allocation system this was music to my ears, however as a researcher there still was a research question to answer. I frankly asked them if they could tell us their considerations in this decision:
Most parts of the system were already discussed, these were actually a formality at this point. It is fine that you presented everything, but in my opinion we already agreed on most things. Nonetheless was it good to summarise it, to get the complete picture. (Department head BO, Ooststellingwerf)
The only thing we had not heard from in an earlier stage, was the part of dividing the cost by three. However we already said that we ideally divide the cost by three, since, simply said, we are in this together and we will all make use of the services. That it is an informed recommendation from your side stems us favorable. (Department head BO, Ooststellingwerf)
We thought the plan was good, there were no odd things in it and you sticked to it. We knew already what dilemmas you faced during the discussions and we had the chance to give our opinion. We were in it together, you did the work and we gave some directions when necessary. (Department head BO, Opsterland)
As already stated, an insider-perspective is used to describe the process of the development and the adoption, i.e. the process is described from the perspective of the developer of the system. In the following chapter the role of researcher is taken in to provide a more objective stance towards the findings presented in this chapter.
