1 Amsterdam Business School
Benefits of co-developing management control systems on the
ability for double loop learning
Name: Alex Martinius Student number: 10467963 Date: 18 June 2015
Word count: 12024
MSc Accountancy & Control, Control track
Faculty of Economics and Business, University of Amsterdam Supervisor: Dr. ir. B.A.C. Groen
2 Statement of Originality
This document is written by student Alex Martinius who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3 Abstract
Purpose To add to the knowledge on the feasibility of the developmental approach to developing a new management control system. In particular to determine the effects of adopting the developmental approach to the ability of the organization to apply double loop learning when confronted with changing circumstances.
Design/methodology/approach I use an exploratory case study within a medium sized health care organization located in the Netherlands. The organization had developed a management control system using the developmental approach. When confronted with a transition in the Dutch health care system, the organization was able to critically assess some of their norms and values to be able to cope with the new contingencies. The project data supplied by the project leader is used to determine the nature of the project, focusing on the main characteristics of the developmental approach. To gain insight into the ability for double loop learning, I conducted interviews with five participants in the project.
Findings I found some evidence that the adoption of the developmental approach did affect the ability to apply double loop learning. Especially the experimentation during the development process helped managers to gain insight into the norms used by the organization. These insights allowed them to discuss the norms and in some cases led to the organization adjusting the norms.
Research implications/limitations The results of this study suggest the developmental approach might also have advantages on organizational learning. However since it is a single case study the results should not be generalized to all organizations.
Practical implications Organizations that are confronted with changing environments might benefit from adopting the developmental approach to developing a MCS.
Originality/value This is, to my knowledge, the first study about the connection between the developmental approach to MCS and double loop learning, and one of few studies about the advantages of the developmental approach.
Keywords Developmental approach, double loop learning, organizational learning, management control systems
4 Acknowledgements
I would like to thank my thesis supervisor, Dr. ir. B.A.C. Groen, for her advice during my research, this has been very helpful, and for her enthusiasm during the process, which helped me to stay motivated. The support of my friends and family has also been invaluable. I would also like to thank all participants in this research for their cooperation, the interviewees, the project leader and the organization that allowed me to conduct this research.
5 Contents
1 INTRODUCTION ... 6
2 LITERATURE ... 8
2.1 DEVELOPMENTAL APPROACH TO DEVELOPING MANAGEMENT CONTROL SYSTEMS ... 8
2.2 DOUBLE LOOP LEARNING ... 10
2.3 DEVELOPMENTAL APPROACH AND DOUBLE LOOP LEARNING ... 13
3 METHODOLOGY ... 15 4 CASE DESCRIPTION ... 17 4.1 THE DEVELOPMENT PROCESS ... 18 4.2 IMPLEMENTING THE MCS ... 20 5 ANALYSIS ... 22 5.1 THE DEVELOPMENTAL APPROACH ... 22
5.2 EXPERIENCE BASED DEVELOPMENT ... 23
5.3 EXPERIMENTATION ... 24
5.4 PROFESSIONALISM ... 25
5.5 DOUBLE LOOP LEARNING ... 26
5.6 DEVELOPMENTAL APPROACH AND DOUBLE LOOP LEARNING ... 29
6 DISCUSSION AND CONCLUSIONS ... 31
LITERATURE ... 34
6
1 Introduction
This paper aims to add to the knowledge of a developmental approach to management control systems (MCS) (Wouters & Wilderom, 2008). Previous research has focused on the enabling nature of Performance Measurement Systems (PMS) developed with a developmental approach (Wouters & Wilderom, 2008) and the increase of initiatives employees take after co-developing the PMS (Groen, Wouters, & Wilderom, 2012). Yet little is known about other advantages of this approach to MCS, for instance on employee and organizational learning (Wouters & Wilderom, 2008). As commented by Wouters & Wilderom (Wouters & Wilderom, 2008) feasibility of the developmental approach is an issue because it is a time consuming approach. It also assumes local experience of employees and the willingness to build on that experience to develop the MCS. Gaining insight into other advantages, such as learning processes that occur due to the usage of this approach, can help understanding in which conditions the developmental approach can be feasible.
To add to the knowledge of effects of a developmental approach to MCS, I will research the effect of this approach on the ability of double loop learning (Argyris, 1977). Single and double loop learning are terms coined by Chris Argyris and commonly explained using an analogy to a thermostat. A thermostat set to 68 degrees can take corrective action when it learns the room is too hot or too cold, it will turn the heat on or off. This is called single loop learning. When the thermostat can question itself whether it should be 68 degrees, it would be able to not only detect errors, but also question the underlying policies and goals. This is called double loop learning (Argyris, 1977). Double loop learning in organizations means the organization is able to question the prevailing norms and values. In a changing environment this will help organizations to adjust to the new situations it faces, while single loop learning will prevent the organization from making changes. Over the past decades double loop learning has become one the most eminent theories on organizational learning.
The research question of this paper is: how does a developmental approach to developing management control systems affect the ability of double loop learning?
The subject of my research is a project within a medium sized health care provider in the Netherlands that provides care to the elderly in a large region in the north west of the Netherlands. This environment is interesting because the Dutch health care system is rapidly changing. Budgets are decreasing and during 2014 and 2015 a large transition in the care of the
7 elderly was started by the Dutch government. During this transition the organization was faced with new financing structures and a reduction in capacity. The transition also entailed an increasing uncertainty of future budgets.
I will use the documentation on a project in which an MCS was developed as well as my own experiences as part of this project team to gain insight in the nature of the project. Interviews with employees involved in the development of the MCS will be used to study the impact of the nature of this process on double loop learning within the organization.
In the next section I will review the literature on double loop learning as well as the current knowledge on co-developing MCS. In section three the methodology will be described. I have done a case study within the organization that has developed a MCS using the developmental approach. The case description of this project follows in section 4. In section 5 I will present the results of interviews with 5 managers employed by the organization, that were all in some way involved in the development of the MCS, and have been using the system for over a year. Conclusions and discussion of the results are given in section 6.
8
2 Literature
2.1 Developmental approach to developing management control systems
Management control systems and the development of these systems has been a popular subject of many studies for many years. These studies have, among other subjects, focused on the relation between the MCS design and the organizations’ strategy (Langfield-Smith, 1997), the effects of MCS on product innovation (Bisbe & Otley, 2004) and the role of MCS as a disciplinary regime (Carmona, Ezzamel, & Gutiérrez, 1997). These studies have led to several leading frameworks on MCS, such as Simons’ levers of control (Simons, 1995) and Ferreira and Otley’s performance management and control framework (Ferreira & Otley, 2005). Most of the available studies on MCS have one thing in common, they study MCS from the perspective of top management. In this view the MCS are used to control middle management, to monitor whether objectives have been achieved, to communicate the organizations’ strategy and to set boundaries for employees. Recently some researchers have shifted their focus from top management to the effect MCS have on middle management that is to be controlled by these systems (de Haas & Kleingeld, 1999; Derfuss, 2009; Luft & Shields, 2003). In particular recent research has focused on the benefits of co-developing MCS: the developmental approach (Groen e.a., 2012; Wouters & Wilderom, 2008).
The developmental approach to developing a MCS is an approach in which middle management is involved in the development and design of the MCS. This approach to the development is characterized by three key elements: (1) being experience based, (2) allowing experimentation and (3) building on employees’ professionalism (Wouters & Wilderom, 2008). Previous research has found some benefits of this approach. The MCS are perceived as enabling employees to do their jobs more effectively rather then as a device for top management to control their actions (Wouters & Wilderom, 2008). Groen et al. found an increase in employees’ initiatives to improve their performances after co-developing a MCS (Groen e.a., 2012). In the next sections the three key characteristics will be briefly described.
2.1.1 Experience based development
The first main characteristic of the developmental approach is that it is experience based. Experience based development means using experience and existing knowledge in the development process. Most approaches to MCS development focus on the goals that are set at the top of the organization and the actions that are needed to achieve these goals. The systems
9 are usually developed by support departments, such as the finance department, without consulting employees that are to be controlled by the system. Therefore the knowledge and experience of these employees is not used in the design of the MCS. These top-down mandated systems are found to be less successful (Cavalluzzo & Ittner, 2004; de Haas & Algera, 2002; Scott & Tiessen, 1999). Cavaluzzo & Ittner found performance measurement and accountability are hindered by factors such as the inability of information systems to deliver accurate and timely data (Cavalluzzo & Ittner, 2004). De Haas & Algera found people will disregard information from formal MCS if this information does not confirm their view of the world, and argue the importance of linking the design of MCS to peoples’ mental models (de Haas & Algera, 2002), the strategic dialogue. In the developmental approach employees are involved in the development of the MCS, therefore their knowledge and experience on capturing relevant data for the MCS can be used in the design. The process involves identification, appreciation, documentation and consolidation of this local knowledge (Wouters & Wilderom, 2008)
2.1.2 Experimentation
The second key characteristic of the developmental approach is experimentation. This involves developing the MCS and its subsequent testing and refining to arrive at a valid and reliable system. Employees are best suited to judge that the outcomes of the system reflect their work efforts. Since the MCS is unlikely to be “right” after the first phase, experimentation with employees is needed to improve the data that is captured, as well as a focus on the interpretation of this data and the way it is presented. Experimentation is also needed to identify unintended and undesirable effects and behavior that might occur in response to the MCS. For instance a MCS that focuses solely on measuring costs might result in cost reduction, however it may also cause quality dropping below standards. It is important to anticipate these side effects and adjust the MCS if undesired behavior occurs. Involvement of employees also contributes to an effective organizational change process (Bourne, Neely, Mills, & Platts, 2003)
2.1.3 Professionalism
Professionalism involves employees’ orientation on learning for the purpose of improving their work processes. It refers to the degree to which individual employees show commitment to their profession and the current organization. Professionalism is likely to be stimulated if self-involvement into improvements is made possible.
10 Existing literature on developing MCS has focused on the increase in initiatives after co-developing MCS and the enabling nature of MCS as perceived by employees after co-co-developing it. Yet little is known about the benefits of this approach on employee and organizational learning. To add to this knowledge, the benefits of a developmental approach to MCS will be studied in relation to the ability of double loop learning in this paper.
2.2 Double loop learning
Double loop learning is a term coined by Chris Argyris and Donald Schön. According to Argyris and Schön learning involves the detection and correction of error (Argyris & Schön, 1978). When something goes wrong, many people do not question given or chosen goals, values and rules, but try to solve the problem within these variables. For instance when a department exceeds budgeted costs and tries to solve this problem by cutting costs without questioning the norms that led to the budget. This is called single loop learning. The focus in single loop learning is on the techniques and making these techniques more efficient (Usher, Bryant, & Johnston, 1997). When the focus of learning and problem solving lies on the underlying norms and policies, this is called double loop learning. In this case the department exceeding the budget would also be able to question the norms underlying the budget and the budget itself.
Argyris argues that double loop learning is important, because without it individuals are not capable of re-examining their values and assumptions (Argyris, 1976), and organizations capable of double loop learning are more efficient (Argyris, 1977). Especially in changing environments organizations will benefit from the capability of questioning their norms and policies. The changing environments could render old values, norms and policies obsolete. Being incapable of re-examining the budgets would prevent the organizations adapting to the changing environment. Based on single loop learning and double loop learning, Argyris and Schön developed 2 models, Model I and Model II (Argyris, 1977), shown in exhibit I.
2.2.1 Model I
Argyris and Schön argue that people have mental maps that guide their actions in response to certain situations (Argyris & Schon, 1974). Although mental maps are determining people’s actions even more than the theories people espouse, most people are unaware of the mental maps. Therefore Argyris and Schön argue there are two theories of action: espoused theories and theories in use. Theories in use are the theories that are implicit in our actions, these theories govern what we do. Espoused theories are theories we call upon to convey our actions to others.
11 Model I assumes peoples espoused theories are not the same as the theories they actually use. People are not even aware of the theories they use. The validity of these theories is tested by their effectiveness in achieving the values people hold. As shown in exhibit I, Argyris and Schön identify 4 governing variables that people who operate by Model I strive to satisfy: (1) define in their own terms the purpose of the situation in which they find themselves, (2) to win, (3) to suppress their own and others’ feelings, and (4) to emphasize intellectual and deemphasize emotional aspects of problems (Argyris, 1977).
The defensive and unilateral actions that are the result of these governing variables, result in self-sealing, single loop learning and little or no testing of theories. This prevents an organization from learning from these actions and publicly discussing and scrutinizing the theories in place.
2.2.2 Model II
In Model II the emphasis is on publicly testing hypotheses and assumptions behind ideas or policies. Contrary to Model I, in which theories in action might not be espoused and actions are considered defensive, in Model II governing variables are valid information and free choice. This results in joint control of tasks, less defensive action and frequent public testing of theories. Model II includes the ability to call upon valid information in order to make informed choices. It encourages open communication and is seen as dialogical. Contrary to Model I learning, Model II learning also entails participation in design and implementation of action and the monitoring of implementation.
The governing variables in Model II enable employees to critically assess organizations’ norms and values and to do so publicly and based on valid information. Interactions are less defensive allowing employees to admit errors or point to errors made by others. This encourages the organization to learn from errors, while in Model I errors are likely to be concealed and remain under the surface. If errors remain unknown the organization will not be able to learn from them.
12 Exhibit 1
Governing variables for
action Action strategies for actor Consequences on actor and his associates Consequences on learning Effectiveness Model I
1 Achieve the purposes as I perceive them
1 Design and manage environment so that actor is in control over factors relevant to me
1 Actor seen as defensive 1 Self-sealing
2 Maximize winning and
minimize losing 2 Own and control task 2 Defensive interpersonal and group relationships 1 Single loop learning Decreased 3 Minimize eliciting
negative feelings
3 Unilaterally protect self 3 Defensive norms 3 Little testing of theories publicly
4 Be rational and minimize
emotionality 4 Unilaterally protect others from being hurt 4 Low freedom of choice, internal commitment, and risk taking
Model II
1 Valid information 1 Design situations or encounters where participants can be origins and experience high personal causation
1 Actors seen as minimally
defensive 1 Testable processes
2 Free and informed
choice 2 Task is controlled jointly 2 Minimally defensive interpersonal relations and group dynamics
2 Double loop learning Increased
3 Internal commitment to the choice and constant monitoring of the implementation
3 Protection of self is a joint enterprises and oriented to growth
3 Learning-oriented norms 3 Frequent testing of theories
publicly
4 Bilateral protection of
others 4 High freedom of choice, internal commitment, and risk taking
13 2.3 Developmental approach and double loop learning
As explained in the previous sections on the developmental approach and on double loop learning, both of these theories bring advantages to organizations. The developmental approach to MCS brings advantages to departments that are involved in developing a MCS. Employees consider the system to be enabling (Wouters & Wilderom, 2008) and take more initiatives to improve their work practices after co-developing the system (Groen e.a., 2012). On the other hand the double loop learning process brings advantages to the organization as a whole. The organization is able to question its norms and values and is able to adjust these norms and values if needed to adjust to changing circumstances.
The developmental approach involves a multi-disciplinary process for developing the MCS. During meetings, the ideas of employees about the system are gathered. Because of the knowledge and experience of the capturing of relevant data that employees have, their input is essential in this process. In many organizations local reports are made on the departmental level that are not part of a centralized MCS. People outside the department might not even know about these reports (Wouters & Wilderom, 2008). By involving employees from these departments themselves local experience can be used in the development process.
The ideas gathered during the initial meetings involve the data to be captured in the MCS, as well as how to interpret the data and how to present data. Based on these ideas an initial design for the MCS can be made. During subsequent meetings the initial design will be discussed, tested and may be altered to better represent the actual work efforts of employees. Alterations can be made to the data captured, the IT systems that are used for gathering the data, the interpretations of data and the presentation.
These characteristics of the developmental approach show considerable similarities to the double loop leaning model discussed in section 2.2.2. This model is characterized by jointly controlled tasks, situations or encounters where participants can be origins and experience high personal causation and frequent testing of theories publicly. These characteristics are found in the developmental approach as multi-disciplinary project group meetings, building on local knowledge and experience and testing of MCS within these groups.
The involvement of employees in the development of the MCS also indicates learning-oriented norms within the organization. Involving employees allows them to learn from the development process, but also allows the organization to learn from the tacit knowledge these employees have.
14 The joint task of developing the MCS requires minimally defensive interpersonal relations and group dynamics and will benefit from valid information and free and informed choice as governing variables, these are also characteristics of Argyris’ Model II.
Action strategies in model I are quite contrary to the characteristics of the developmental approach. These strategies involve unilateral control over tasks and protecting oneself, where co-developing a MCS involves cooperation between people. Defensive relationships within the project would much rather harm the process then contribute to a successful result. Consequences of single loop learning are little public testing of theories, while experimentation and testing is an important aspect of the developmental approach.
Based on these similarities between double loop learning and the developmental approach, and the inconsistencies between the developmental approach and single loop learning, it seems likely that the developmental approach and double loop learning would be complementary to each other. Either an organization capable of double loop learning would be more likely to adopt the developmental approach to a new MCS, or adopting this approach would increase the ability of the organization to apply double loop learning. Since the subject of this study is an organization in which norms and values were subject of discussion after adopting the developmental approach, my proposition is as follows.
Proposition: Adopting a developmental approach to developing a management control system increases the organization’s ability to apply double loop learning
15
3 Methodology
This research is a case study. As proposed by Yin, a case study is the preferred research method for “how” questions (Yin, 2013). A case study allows the researcher to study why a phenomenon occurs. This case study is an exploratory case study. It seeks to add to the knowledge of a developmental approach to MCS and the benefits of this method. The relation to organizational learning that is studied has not been studied before. At this stage a single case study is the most common research method (Spicer, 1992). Although I was involved in the development of the management control system, this case study can not be seen as action research. The project that is studied in this paper took place in 2013, while the research started in February of 2015. Therefore I tried to rely mainly on documents provided by the project leader instead of my own recollection of events. Information was gathered on the project that is the subject of this study, for instance minutes, reports and correspondence. This will explain the nature of the project as a developmental approach to the development of a MCS. The project leader provided the project archive. This project archive contains all official documents, draft versions of the MCS, correspondence with participants and top management. These documents were coded using Atlas.ti. The case description was sent to the project leader to ensure the description of the development process is a good and complete representation of the project. Since I was also a participant in the project as a business controller for the organization, I will also use my own experiences during the project. The case description, which will be based on the documents and experiences, will be reviewed by the external project leader to ensure it is a truthful representation of the project and no key aspects are overlooked.
To gain insight into the potentiality for double loop learning in the organization, interviews were conducted with participants in the project as well as managers that joined the organization during the project or after developing the MCS. By interviewing participants as well as managers that joined after the initial development and design of the MCS, I have gained insight in the different views of these managers. This could lead to a better understanding of the key aspects of the developmental approach that influenced the ability for double loop learning. To assure availability of key participants, board approval was given and employees were encouraged by the board to participate. All interviews were recorded and coded using Atlas.ti software shortly afterwards and before the next interview to be able to incorporate any new insights into the next interviews. Interviewees were informed the interview would concern the development of the MCS, but not informed of the research question to prevent they would be guided toward desired
16 answers. Interviewees were selected to ensure managers of all main nursing homes were included. Anonymity was ensured to all interviewees.
The focus of the interview guide (appendix 1) was first the development of the MCS. The role of the interviewee was explored, as well as the occurrence of the key aspects of the developmental approach, experience based development, experimenting and professionalism. These were separate points of interest. After exploring the development process the changes in the organization were discussed, mainly focusing on changes that originated from within the organization, for instance changes in strategy, norms and values. This led to exploring the existence of a relation between the (development of the) MCS and the changes that were discussed. Interviewees were encouraged to speak freely, to ensure all relevant aspects were covered. The conclusions drawn from the interview were sent to the interviewee to ensure it would represent their meaning in a correct way. All interviewees confirmed the conclusions based on their comments are correct.
During the analysis of the interviews, all recordings were listened to several times, as well as after all interviews were analyzed. This way the first interviews were also analyzed again with any new insights from the later interviews. It also ensures nothing was overlooked during the process of writing.
The research design will be based on the work of Yin (Yin, 2013) and follow the procedure outlined in exhibit 2.
Exhibit 2
17
4 Case description
The subject of this case study is a project within a medium sized Dutch health care provider. It provides health care services to approximately 750 elderly in several large and smaller nursing homes. The organization strives to deliver excellent services to their clients. The organization is run by the board of directors and the management team consisting of a 2 managers of care services, a manager of Finance and Control, an IT manager, and an HRM manager. The organization manages three large nursing homes with capacity for over 100 residents and some smaller nursing homes with a capacity of 25 to 50. Two of the large nursing homes are run by a location managers and department managers for the departments within that nursing home. The smaller nursing homes are run by just a location manager. Besides the intramural care in the nursing home, the organization also provides home care and treatment to clients in the region of the nursing homes.
Due to the rising cost of health care services in the Netherlands the Dutch government started reforming the health care system in 2013. Because of the new government policies, the organization faced a decrease in budgets and an increase in vacancies in their nursing homes. These changes were part of a transition in which the government restricted access to intramural health care to clients with a high demand of health care. For the organization this meant some of the smaller nursing homes that focused on elderly with a lower demand of health care had to downsize. In 2014 and 2015 some tasks were also delegated from the national government to municipalities and health care insurance companies. This delegation was accompanied by a reduction of the budgets of up to 40%. The delegated tasks concerned the home care and treatment and rehabilitation at the nursing homes.
These changes urged the organization to be more efficient and to cut back their costs. Direct personnel expenses are approximately 60% of total costs, other expenses consist of overhead expenses (19%), cost of capital (11%) and material expenses concerning meals, nursing materials and cleaning and laundry services (10%). The organization employs approximately 1500 employees, 800 full time equivalents (fte).
Due to a decrease in budgets and a varying occupancy of the capacity in the nursing homes and uncertainty about future budgets, more attention was given to efficient staffing of the care departments, in accordance with budgets as well as the needs of the residents. In 2013 budgets were based on an estimate of the population in the homes in terms of the expected weight of
18 support (ZZP)1. Based on the norms on which the ZZP rates are based, published annually by the Dutch health care authority, personnel expenses for care departments were budgeted. Budgets were calculated by the finance department, while underlying norms were set by top management. The organization used a top-down budgeting process with little consultation of middle management.
4.1 The development process
In the final quarter of 2012 a project group started developing a tool to assist middle management in efficiently deploying their human resources based on the norms adopted in the budgeting process. This project group consisted of an external project leader, several middle managers from different nursing homes, an HRM advisor and a policy advisor. After several project group meetings, the researcher (business controller in the organization) was added to the project group. The steering committee overlooking the project consisted of the finance manager and one member of the management team responsible for the nursing homes. The project group reported progress to the steering committee. The steering committee was responsible for the results and reported to the board.
One of the main aims of the project was to develop a pool of standby employees that could be called upon in case absence of an employee, for instance due to illness. This so called flex pool would enable managers to fill gaps in their schedules and provide flexibility in case of unanticipated increase in demand for health care at their department. Managers would be able to keep the scheduled shifts to a minimum, preventing overstaffing their departments in case the demand would suddenly decrease due to decreasing occupancy.
The development process lasted several months. During this period the project group first developed norms for composing the time schedules for their departments. Time schedules were based on the needs of the residents of the nursing homes. This process also involved norms for hours of training per employee per year, in part mandatory training to retain their license as health care workers, and sick leave. Another point of interest was the minimum requirement of education level of the employees and the need to have highly qualified personnel available throughout the day. These norms were based on the experience of middle management and the norms used by the HR department and the finance department for setting the yearly budgets.
1 In The Netherlands, financing of care of the elderly is based on classes of weight of support and treatment needed, so called Zorgzwaartepakketten (ZZP).
19 The norms on availability of qualified personnel were also based on national regulations and requirements made by the health care insurers. The norms for the organization were then incorporated in an existing tool the project leader had developed during former projects in other organizations.
The next stage of the development process was formulating best practices on deploying human resources based on the different populations at the departments, for instance somatic and psychogeriatric illnesses, or specialized departments for people who suffer from Korsakoff’s syndrome of for rehabilitation after surgical treatment. During this process, the managers of these departments were able to share their experience, question each other’s practices and learn from each other. Best practices were developed in several sessions which each lasted two hours. The participants were the project leader, who was responsible for guiding the process, managers from middle management, the HRM advisor and the business controller.
During these sessions everyone was encouraged to come up with new ideas to be tested. The HRM advisor and project leader were available for advice and to critically assess ideas. These assessments included comparing ideas to the norms for availability of personnel and the required level of education. The business controller then calculated the costs of the time schedules. During this phase many of the time schedules were found to be too expensive due to the use of highly educated personnel. This was a result of only using measurement of human resources (fte) required. While the time schedules were within the norms on fte, the managers did not take into account the costs of these resources and this information was not available in the MCS. Other schedules were found to be below the minimum quality requirements the organization or health care regulations required. The business controller and project leader decided to add the financial parameters, consisting of budgets as well as projected costs of the time schedule, to the tool. This enabled the users of the tool to estimate the cost of their time schedules and compare these to the budgets. It also allowed managers to calculate budgets and costs for other scenario’s, such as an increase or decrease of ZZP, leading to an increase or decrease of budgets. Managers were also able to test new time schedules and compare the needed human resources with the available resources.
After these sessions the MCS was designed based on the outcomes of the meetings. The project leader designed the tool, with assistance from the business controller for incorporating the financial norms. An instruction for the use of the tool was also composed. Several managers were then asked to experiment with this new version of the tool.
20 In this stage of experimentation the outcomes of the tool were compared to actual costs for the departments. This uncovered large differences in expected costs and actual costs. These differences were partly due to discrepancies in the used time schedules and the desired time schedules as they were calculated. After testing the tool for several departments at different locations and discussing the results with several managers the project leader and business controller started questioning the results. The project leader reviewed the calculations and discovered an error in the calculation of projected costs. This error was corrected and a new version was released for testing.
The project documents clearly indicate two of the characteristics of the developmental approach discussed in section 2 were present during this process. For instance the participating managers were asked to give input for the norms that should be used, and an employee from the planning department was asked to assist in developing rules on scheduling personnel. This indicates local knowledge was used in the development of the tool. Experimentation was included in the action lists the project leader formulated after each project group meeting. The third characteristic, professionalism, was most apparent during the meetings in which managers developed best practices for composing the time schedules.
4.2 Implementing the MCS
After testing the MCS, the system was implemented during the second half 2013 for all departments in the organization. Managers use it to compose their time schedules, compare estimated costs with budgets and to compare the needed personnel to cover the shifts in the time schedule with the personnel available at their department. The results of the calculations are used in discussing the needs of the departments to recruit new personnel. The calculations were also used to experiment with new ways of deploying personnel, and find ways to schedule shifts more efficiently. For instance the qualified nurse’s (level 4 or 5) night shifts were discussed and have in some of the nursing home been assigned to specific departments, where they used to service the entire nursing home with a lower level health care worker (level 3) assigned to the departments. This change is assumed to lead to a higher quality of care and a reduction in costs as well.
During the budgeting process for 2015 the tool was used to set budgets for one of the nursing homes. This nursing home had trouble staying within budget for several years due to the inefficient construction and capacity per department as well as varying occupancy. In the
21 budgeting process the budgets for the departments of this location were based on the costs of the minimum of personnel needed to guarantee good care for the residents. Other departments were still budgeted according to the norms the organization had used for the past years, with some adjustments made to reflect changes in the organization (consisting of lower overhead costs and a new organizational structure).
After the first year of using the MCS, the tool was redesigned to be less complicated after several managers had complained the tool was too complex. The new version contained only one excel sheet with a few input cells. Managers would be able to adjust their input quickly and directly see the results of the calculations.
The goal of forming a pool of standby employees, one of the main reasons of starting the project, was not accomplished. During the project, the focus had shifted from this aim to that of developing a tool that would enable middle management to experiment with new time schedules and keep personnel costs within budgets in the changing circumstances they were confronted with.
22
5 Analysis
To gain insight in the way de development of the management control system affected the organizations ability for double loop learning interviews were conducted with five employees within the organization. The employees interviewed were all in some way involved in developing and implementing the tool and most of the have used the tool for at least one year. Interviewees were selected based on the nursing home they work in, so all larger nursing homes are represented by at least one interviewee, as was one of the smaller nursing homes with a specialized department. To get a good insight in the way the development phase attributed to the ability for double loop learning, some interviewees were selected that had been involved from the start, as well as participants that joined the project in later stages, such as the implementation phase. One of the members of the steering committee was also interviewed. This employee had been responsible for the large nursing homes as a member of the organizations management team and had also been involved in several discussions about the prevailing norms of the organization.
The selected interviewees also had different backgrounds. Some were managers with many years of experience in health care management, while some had only been in that position for a few years. Two managers had worked for the organization for over 5 years. Three managers actually joined the organization at the start or during the project. Four of the managers were either managing one single department within a nursing home or were managing an entire location. The fifth manager was the manager that was part of the management team.
To ensure anonymity, the interviewees will be named manager 1 through manager 5, so comments can not be traced back to individuals.
5.1 The developmental approach
During the interviews, some managers commented in general on the development of the MCS. Views on the project, that was as described in the previous section a process that took several months to a year, differed strongly. Some managers felt it had taken too long to develop the tool. Manager 5 commented:
“… the process could have been much easier. It was a lot of “blabla” and not to the point. And after a year and a half, we didn’t gain much from it.”
23 It seems the many group meetings, that actually took several hours per meeting, had a lot of influence on her opinion on the development process. Especially for those participants that were managing specialized departments, as this manager did, these meetings might not have been useful. They could not gain as much from the discussions as managers that managed more generic departments.
Manager 1 also commented on the changing aim of the project, that started out as a project to form a flexible pool of employees:
“To be honest I think it did not go well. It was laborious, unnecessarily expensive and badly communicated. I also feel we could have gotten more out of it, for instance the flex pool.”
This manager also comments on the lengthy process of with many project group meetings. A similar remark was made by manager 2, who thought the implementation could have been much quicker. This indicates these managers would rather spend less time on designing and discussing the tool. It could have been implemented sooner, allowing the experimentation to take place after implementation. Both manager 1 and 2 were involved in this experimentation and found this to be helpful when it was discussed in smaller meetings, mostly with the business controller and/or the project leader.
One manager however found the process to be very helpful. When asked if the implementation process had been essential in giving managers insight into the ways they could influence their profitability, she said this process had been pivotal. She felt that before the implementation process had taken place, managers could only put in their time schedules and see the results, but were unable to act on these results.
5.2 Experience based development
The first part of the interviews concerned interviewees’ experiences in the development of the MCS. It focused on the characteristics of the developmental approach discussed in previous sections. Some interviewees confirmed their own experience was used in developing the tool. As one of the interviewees, manager 1, stated:
“Because I was closely involved and have been consulted often, I do recognize my input in the tool. We met several times to discuss the tool and changes were made. We also discovered an error in the tool that was corrected. I see this input reflected in the tool”.
24 Another interviewee, manager 3, had commented on the tool not representing the way the organization worked, in terms of roles of employees and their education level. These basic assumptions used in the tool were then adjusted accordingly to better represent the reality in the organization. For instance the health care worker that was initially listed as having a level 2 function, but had actually a level 3 function. She also commented on the calculation error that was found during the later stage of development, which was also corrected.
Interviewees that did not feel their experience was reflected in the tool were the employees that did not participate in the first stages of the project. Since most of the design work was done in the first stages of the project, it does seem logical these participants did not feel their experience was used. Although one participant, manager 2, who had only participated in the implementation of the MCS and as a user of the system commented:
“In the current version I see this especially concerning the smaller nursing homes where the tool takes into account the problems of these locations’ profitability … I did give this input, not in project groups or the steering committee, but through the finance and control department”
He also commented he did see this input reflected in later versions of the MCS. This indicates that even in the stages following development of the MCS, adjustments still had been made based on the experience the users had with composing time schedules and using qualitative and financial norms. In this stage of the process, the project leader, the business controller and two of the location managers had worked together very closely to make some final adjustments, mostly based on the outcomes of the calculations made by their department managers.
5.3 Experimentation
It is obvious that some experimentation took place during the development and the implementation of the tool. As described in the previous section participants were encouraged to test the tool after which the project leader and the business controller reviewed the outcomes and if needed adjusted the tool.
As commented by manager 3:
“The results in earlier versions were hard to read. What does this mean? What conclusions can we draw from this? It was meant to be a tool that would enable and give insight to managers … adjustments were also made to the presentation of the tool.”
25 And manager 1 stated, when asked how he had experienced the development and implementation of the MCS:
“It has proven to be a good way to clear up some things that were not clear. Together with one of the other managers we discovered some errors and were able to correct these. In the current version, on one A4 sized sheet, and with the calculations that are used in the background, it is a more clarifying tool”
These remarks emphasize the importance of the experimentation during the project. This experimentation has resulted in a tool that is easy to use for managers and gives them the insight they needed. It enables managers to quickly draw conclusions from the calculations done. As found by Wouters and Wilderom (Wouters & Wilderom, 2008) in previous research, in this project experimentation during the development also improved the enabling nature of MCS. The experimentation also led to a tool that better incorporates the different circumstances in the various settings the tool is used in. Especially for the smaller nursing homes the MCS became more useful after experimentation and adjusting the tool, as mentioned in the previous section. 5.4 Professionalism
The views on professionalism differed between managers that were rather new to their position or to the organization and the managers that had years of experience in health care in a middle management position. Because the MCS concerns national norms set by the Dutch health care authority, as well as norms specifically used within the organization, a differing level of professionalism is to be expected. As one manager that joined the organization during the development process said:
“I prefer to use the tool on my tablet, and then I will have a look at how it works. I like to research this myself. This gives insight into how the calculations work. Sometimes I find something I can use for myself.
All managers however acknowledged the tool did give a clear insight into the underlying norms. This did not occur due to the degree to which individual employees show commitment to their profession and the current organization, but when confronted with these norms, they were discussed in project group meetings. This led to a better understanding of the results in the MCS, as well was a better understanding of yearly budgets and personnel expenses.
A clear distinction in professionalism during the development of the MCS can also be made between those managers that struggled to keep expenses within the yearly budgets and managers that had no problems staying within budget. Those managers that struggled to keep their
26 personnel expenses within budgets were more dedicated to learning from participating in the project. The MCS was used most by these managers, which not only used the tool to plan their time schedules, but also to justify their financial results. As manager 2 commented, the tool had played a great role in justifying his locations results to his superior and the finance manager.
Based on my years of experience I am familiar with the numbers. When I saw the results of the calculations it became clear to me this was impossible to achieve.
Besides the managers learning about norms and using these norms to improve their financial results, manager 2 also felt the development process had also improved understanding at the finance department.
During the development process we frequently discussed these issues. The finance department quickly understood the issues that arose with the operation of departments with fewer then 24 clients. We included the finance manager and demonstrated the budgets for these departments were insufficient.
5.5 Double loop learning
The second part of the interviews focused on the changes the organization was going through. Due to the large transition in the Dutch health care system and budget cuts by the Dutch government, the organization was facing several challenges to ensure it would stay profitable. This led to many changes in the organization, from reduction in overhead functions, to completely restructuring facility services in the nursing homes. In 2014 a large reorganization was announced.
Besides these changes initiated by top management, on a lower level changes were also visible. Some of the managers had started to restructure some of the functions in their department. For instance the way of organizing the night and weekend shifts of the nurse practitioner was discussed within two of the three large nursing homes. During 2014 the managers in one of the nursing homes also wrote a pressing letter to top management urging for a reconsideration of the way the budgets were divided, indicating that with the budget their departments had it would be impossible to responsibly organize their services.
These changes, among other changes mentioned by interviewees, were the next point of focus in the interviews.
27 The decreases in budget, accompanied by a higher demand by their clients, were mentioned by all interviewees as an important change they had seen in the organization in the past years. As manager 5 commented on the changes:
“The health care demand is increasing and at the same time our budgets are decreasing. We have to do more with fewer resources. At the same time the demands by our clients, and especially their family, are also increasing… It also increases the workload for employees to be able to deliver the excellent care services the organization strives to deliver.”
When asked what the most important changes have been in the past two years, manager 3 responded:
“Among the most important changes is that demands by our clients are changing. Not just demands by individual clients, but also the demand in health care in general. We see this leads to pressure on employees. We have to highlight this increasing demand, but also keep an eye on the business dimension.”
This business dimension, the focus on the profitability, was mainly due to the new ways of financing health care. In the past the budgets were set for a year and health care providers knew exactly what their budget was. This has shifted to a system linked to actual occupation of the nursing homes and the weight of support needed by the clients. Most recently the budgets were also cut by almost 5 percent and closing of one of the nursing homes was announced due to government policy.
This shift to a business approach to health care is an important shift in focus within the organization. Revenues are increasingly volatile and managers have to focus on efficiency while also retaining the high quality of care the organization demands. This has led to initiatives to increase efficiency. One example of this focus was given by manager 4:
“How can we deploy our resources best? I think flexibility is very important. Not everyone is used to this. For instance everyone starts work at 7:00 am. I asked them to consider if this was really needed. Is it possible to start some shifts at a later time?”
This seems like a small adjustment in the time schedule. However these small changes actually might lead to significant cost reductions, since direct personnel expenses amount to approximately 60% of the budget. It is also a large change in norms used when scheduling personnel. The time schedules used to be roughly divided in three shifts of 8 hours per day, starting at 7:00 am, 15:00 pm and 23:00 pm. Changing these schedules meant a significant change for employees that had to be more flexible.
28 Another change that was mentioned by almost all interviewees, concerned the budgeting process. For several years the budgets were based on the norms published by the Dutch health care authority and that norms were used for establishing the annual maximum rates per ZZP. The finance department used these rates to divide expected revenues among the care departments, therapists, facilities departments and housing and overhead expenses. During 2014 this budgeting process and the use of these norms was often criticized by middle management. They felt the care departments were not given enough resources to provide reliable care to their clients. To maintain the quality, budgets were overstepped at almost all departments. Eventually after several months some middle managers expressed their concerns about these budgeting norms to top management, arguing they could not deliver the excellent care the organization strived to deliver to its clients. This discussion about the norms led to an alteration in the budgeting process, where the minimum personnel expenses, based on calculation done in the MCS, were also incorporated into the yearly budgets. As manager 2 commented:
“It was obvious at some time that budgets would be exceeded. In the MCS this was perfectly clear. There was no relation between the MCS and the budgets. It led to an allocation of resources that departments could not be held accountable for. I now see a movement toward a new budgeting process, apparently that was a learning process in the organization.”
Manager 3 also mentioned the use of the MCS as input for the budgeting process as an important change in the organization. However manager 1 did not think the MCS had been used in the budgeting process. This might be explained by the difference between several of the nursing homes, where some were still budgeted using only the norms, which led to budgets that were sufficient for delivering the care the clients needed. Other departments however needed extra resources, due to an inefficient location or varying occupancy. The manager of a location that had the highest budget overrun actually felt the MCS had been important in the budgeting process.
Based on these examples, it seems clear the organization had quite a few occurrences of double loop learning when confronted with the changes in the health care system in the Netherlands. As Argyris (Argyris, 1977) argues, double loop learning is most important when organizations are confronted with change. It allows organizations to adjust to the new circumstances they have to operate in. When confronted with budget cuts and a big transition in the national health care system, as well al as higher demand in health care services by their clients, the organization was able to discuss and alter norms regarding the deployment of personnel and the budgeting
29 process, and was able to view their operations with a business approach it was not accustomed to. The next focus of the interviews was the role the MCS and the development process had in these occurrences of double loop learning.
5.6 Developmental approach and double loop learning
After discussing the changes the organization was facing, the interview focused on the role the development of the MCS had played in these changes. Views on the role of the development of the MCS in the ability of double loop learning in the organization differed among the interviewed managers. While some thought the development process had a great role in the changes in the organization, others felt the MCS had no part in these changes.
Manager 3 felt the development and implementation process were very important:
Yes, that process has been very important. Mainly because during that process, the managers discussed the matter with the controller and HRM advisor. That way it became clear they could do it differently. This was very valuable. The development and implementation process was crucial to give managers this insight.
This manager also said that before the implementation project, managers did not know what to do when a budget overrun occurred. As she said:
Before the implementation process, managers sometimes knew they had too much employees at their department. But they just said “I have these people on my payroll, so I have to pay them, I can’t do anything about that”. In the implementation process, these ideas were challenged and solutions to the problems were sought together with the controller and HRM advisor.
These comments clearly point out the importance of the co-development of the MCS for this manager. To get the managers, the controller and the HRM advisor together to review the results of calculations in the MCS has led to new ways of solving these problems. The manager felt these discussions would not have taken place without this development and implementation process. It also led to a better understanding of the results in the MCS by middle management.
It gave managers a lot more insight. If they had only been able to operate the tool, report on the results, and get on with their usual way of working, it would not have been useful. But especially the implementation process has been crucial to change this.
Another manager felt the MCS had not played a role in her daily work, or in the changes she made at her department. She only used the tool when changes occurred at her department, for instance when expanding the department, of when occupancy varied. She would use it then to
30 calculate a new time schedule based on the new budgets, but would not alter the norms that she had always used. A possible reason for this might be explained due to the department she was working at, which was not affected by the changes in the organization as much as other departments. She also mentioned it was a very steady department, without many changes in occupancy and staffing requirements. There might not have been a need for changing, of for double loop learning, at her department.
One of the managers involved in the discussion on budgeting norms felt that the development of the tool had some role in being able to start this discussion, but saw some other circumstances that also contributed.
I think it helped us to clarify what resources we had in 2014. This was based on the allocation of resources in the yearly budget. But another factor was that one the managers already had insight in the financials and asked the finance department to further explain this allocation. It is not entirely because of the MCS, it is also because of the insight in the budget some managers already had.
He did feel the development of the MCS might have contributed to the insight in financial norms.
I think it contributes to cost awareness, realizing that a tariff has influence on the available resources. One location did not pay attention to the ZZP their clients had. Because they had mainly the low ZZP, they also had smaller budgets. The MCS showed the impact of these tariffs for lower and higher ZZP’s.
During the interviews it was obvious that one of the main reasons the organization was able to discuss the norms they were using, was the insight managers had in these norms. The development of the tool has contributed to this insight through the experimentation during the development process. The financial norms were also clearly incorporated in the MCS, allowing managers to assess them and, when it was obvious the allocated resources were not sufficient, to substantiate their claims they needed more resources. Since these discussions led to an alteration of the budgeting norms, this clearly influenced the ability for double loop learning.
31
6 Discussion and conclusions
The results of this case study indicate the ability for double loop learning and using the developmental approach when developing a MCS are connected. Although the views on the way the MCS and the development and implementation process have contributed to the double loop learning processes in the organization vary among the interviewees, most interviewees feel it has in some way affected the changes in the organization.
There are some explanations for these varying views on the way the development of the MCS had affected the changes. First of all not all managers had been involved in the entire project, because some had joined the organization during the project. If the design phase of the project, which took place in the first months of the project, had a large influence on the ability for double loop learning, managers who joined the organization after this process might not feel the development of the tool had much impact at all. The manager that was most convinced the development of the tool had contributed to double loop learning processes had been involved in almost the entire project. However another manager that had been part of the project group from the start did not feel there was a strong relation between the development process and the changes in the organization, while one of the managers that had joined during the project did feel it had affected the way the organization responded to changes.
Another explanation for these differences might be that not all managers had a lot of experience with the financial norms the organization used to set the budgets. Since the insight in these norms was seen as crucial to be able to discuss the norms, managers that had little insight in them before the project started would feel the impact of the development process was much more important. Managers that already had this insight and understanding of the norms would not have benefited as much as the less experienced managers. This would also explain why the participation in the project, as discussed above, does not fully determine the views of the manager. The managers that were relatively new in the organization did feel there was a strong relation between the MCS and the development and the way the organization had changed, while some the managers that had been working at the organization for several years before the project felt this link was much weaker and had alternative explanations.
It can be concluded that the insight in norms the managers gained from the MCS and the project in which it was developed was very important in critically assessing the norms. Another
32 important aspect was the close cooperation of the managers and the controller and HRM advisor in the interpretation of the results of the calculations in the MCS. It therefore seems the experimentation within the project has been the most important aspect of the development of the MCS. During this phase of the project, the managers and controller worked closely together to interpret the results in the MCS. As a result of this experimentation the managers became familiar with the norms the MCS used to determine budgets and cost estimates. They gained insight in the financial consequences of changes in their time schedules and occupancy of their departments. They were also able to use this insight to discuss their financial results and their budgets with their superiors. In accordance with the research of Wouters and Wilderom this case study also showed the importance of experimentation in the development of a MCS that was regarded as enabling middle management in their daily work (Wouters & Wilderom, 2008). Based on the results of this research the proposition, as formulated in section 2, “adopting a developmental approach to developing a management control system increases the organization’s ability to apply double loop learning”, seems to be supported by this case study. Alternative explanations however do exist. During the development of the MCS the Dutch health care systems was rapidly changing. Because of these changes the organization was forced to change their norms to assure it would stay profitable. These circumstances the organization had to deal with, might also have had a great impact on the ability for double loop learning. And although middle management feels the development of the MCS had impacted and enabled double loop learning in the organization, the changes that were the result might have also have occurred without the development of the MCS. Therefore to gain further insight in the role of the development of the MCS, it would be interesting to research this from the perspective of top management. Top management has of course a decisive role in setting norms and values. However the comments of the member of the steering committee, who was also member of the management team, do indicate that the development process also contributed to double loop learning at the level of top management.
Another direction for future research might be if the ability for double loop learning in an organization affects the choice for the developmental approach. As explained in section 2, the theories might complement each other. This could indicate that organizations capable of double loop learning might be more inclined to adopt the developmental approach when developing a new MCS. As we start to gain more knowledge on the advantages of the developmental
33 approach, it could be beneficial to also understand under which circumstances organizations are most likely to adopt this approach successfully.
This research has some limitations. It describes the link between the developmental approach and double loop learning based on a single project in one organization. Due to time limitations only participants of the project were interviewed. Setting organizational norms and values is the responsibility of the board, therefore as suggested earlier, insight in the views of top management might further our understanding of this link. Future research might also distinguish between participants of a project and managers that did not participate to gain further insight in the role the project affected double loop learning, and the influence of the MCS itself.
It is also impossible to exclude all other influences that might have contributed to the double loop learning processes, such as the changes occurring in the Dutch health care system. Future research in other settings will have to be done to rule out any such factors. Another factor that might have influenced the results of this research concerns the time between the development of the MCS and the interviews. This research was conducted nearly 2 years after the initial development of the MCS. Recollections of the role the development had in discussing the norms might have been influenced by events that occurred in the period between the interviews and the project in 2013. A longitudinal case study, in which a researcher works within the project, would overcome this limitation. Although I have been involved in the project, at the time this participation was not aimed at conducting this research. A longitudinal case study might result in a deeper understanding of the relation.
