University of Groningen
Faculty of Economics and Business
Business Administration, Organisational & Management Control Master Thesis
Groningen, November 2009
An integrated model for management control optimisation:
The relationship between contingency theory misfit and
management control problems
Student: Peter Wiltjer1 Student number: 1258753
Thesis supervisor university: Mw. Dr. S. Tillema
Second thesis supervisor university: Prof. Dr. H.J. ter Bogt
Abstract
In the management control literature, management control problems (lack of direction, motivational problems and personal limitations) are usually regarded as inherent to human nature. Therefore, it is assumed that these problems have to be addressed by implementing behaviour-influencing devices (management controls). Conversely, this paper suggests that such problems may result from other, underlying, organisational problems. It implies a cause-effect relationship between the alignment of an organisation’s control package vs. its contingencies and these various problem categories; non-optimal alignment is regarded as a catalyst for control problems. Furthermore, it provides empirical support for this approach, and it examines how this approach may be useful for managers in their pursuit of optimal control.
Table of contents
Abstract ...2
Table of contents...3
1. Introduction...4
2. Theoretical background ...6
2.1 Defining Management Control (MC) and Management Control Systems (MCS)...6
2.2 Internal consistency...7
2.3 A framework for analysing control packages ...7
2.4 The Malmi & Brown (2008) control package...9
2.4.1 Planning controls...9
2.4.2 Cybernetic controls ...9
2.4.3 Reward and compensation controls... 10
2.4.4 Administrative controls ... 10
2.4.5 Cultural controls... 11
2.5 Macro control variables ... 11
2.6 Environmental and organisational characteristics... 12
2.7 Management control fit and misfit ... 14
2.8 Control problems... 15
3. Design and research method ... 17
3.1 Research design ... 17
3.2 Methods of data collection and analysis ... 17
4. Empirical results and analysis... 20
4.1 Case introduction ... 20
4.2 Contingency changes... 20
4.3 Organisational control problems resulting from misfit ... 21
4.3.1 Planning controls... 22
4.3.2 Administrative controls ... 23
4.3.3 Cultural controls... 23
4.4 Macro control variables and control problems... 24
4.5 Empirical support for the conceptual framework... 26
4.6 A general approach for dealing with the control problems described above ... 27
5. Conclusion ... 30
5.1 Summary / discussion... 30
5.2 Recommendations for future research... 32
1. Introduction
For business organisations, being “in control” is assumed to be an important prerequisite for long-term survival. According to Merchant & Van der Stede (2007, pp. 11), organisational success depends on “good control”, which has to be future-oriented and objectiveness-driven. This implies the need for flexible and adaptable controls. Merchant & Van der Stede (2007, pp. 8) also state that management control is necessary for dealing with control problems, which they define in behavioural terms (lack of direction, motivational problems and personal limitations). In their view, management control is an instrument used by the management of an organisation for directing the behaviour of its employees, in order to avoid employee misbehaviour and otherwise undesirable employee limitations, which are assumed to undermine organisational performance. Contingency theory provides a more elaborate view on the relationship between controls and control problems (Donaldson 2001). Given the large variety in management controls (MC) and management control systems (MCS), firms have to select a firm-specific set of controls (Abernethy & Chua 1996, pp. 573), which optimally fits with environmental and organisational
characteristics, in order to achieve optimal performance (Fisher 1998, pp. 52). The set of controls
is also known as a control package (Otley 1980), while the characteristics are assumed to be represented by contingency variables, both environmental and firm specific (Gresov 1989; Gerdin & Greve 2004; Cadez & Guilding 2008; Donaldson 2001).
According to Merchant & Otley (2007, pp. 785), the control package has to be able to:
“[…] help an organisation to adapt to the environment in which it is set and to deliver the key results desired by stakeholder groups.”
any misfit between a firm’s control package and its contingencies may allow/encourage employee misbehaviour and limitations (MC-problems), which causes non-optimal organisational performance.
This paper is an attempt to provide evidence for this integrated view on the nature and origin of control problems. In the context of this view, it tries to identify explicitly the problems that could arise as a result from non-optimal fit between a control package and its contingencies, and to investigate how managers can use the above integration of theory for countering those problems in practice. In order to do so, it uses the following questions as an outline:
- What are the elements that make up a control package?
- Which contingencies (i.e. environmental and organisational characteristics) could influence a control package?
- What problems may result from any level of misfit (i.e. inconsistencies between a control package and its contingencies)? How do these problems relate to the MC-problems identified by Merchant & Van der Stede (2007)?
- How do organisations handle the problems identified, and what are the consequences of the above-suggested integrated view on control for managerial problem handling in practice?
2. Theoretical background
In this chapter, some relevant theoretical constructs and concepts will be clarified. First, the definition of management control and management control systems will be specified. Thereafter, the idea of controls operating as a package will be elaborated. Subsequently, an overview of possible contingencies will be described. Based on this overview, possible problems concerning the alignment of a control package with its contingencies will be proposed.
2.1 Defining Management Control (MC) and Management Control Systems (MCS)
The definitions of MC and MCS are highly interwoven with other definitions in the field of Management Accounting and Control. According to Chenhall (2003, pp.129):
“The terms management accounting (MA), management accounting systems (MAS), management control systems (MCS), and organisational controls (OC) are sometimes used interchangeably. MA refers to a collection of practices such as budgeting or product costing, while MAS refers to the systematic use of MA to achieve some goal. MCS is a broader term that encompasses MAS and also includes other controls such as personal or clan controls. OC is sometimes used to refer to controls built into activities and processes such as statistical quality control, just-in-time management.”
It follows from this definition that a MCS is very broad, and according to Merchant and Otley (2007), it encompasses almost every part of the organisation. In order to narrow this view on MC, Merchant & Van der Stede (2007) have made a distinction between two basic functions of control: strategic control and management control. The former is focussed on strategy formulation and strategy updating, while the latter is primarily focussed on influencing the behaviour of employees. Nevertheless, this text incorporates both management control and strategic control in its definition of a control package, because strategy is considered to be part of, instead of contingent to, control. Section 2.6 will elaborate more on this issue.
2.2 Internal consistency
Otley (1980, pp. 422) first mentioned the idea of different controls operating as a package, when he remarked the overlapping and interrelating effects of different controls. Recently, Malmi & Brown (2008, pp. 287) defined a control package as follows:
“A management control systems (MCS) package is a collection or set of controls and control systems. The individual control systems may be more traditional accounting controls such as budgets and financial measures, or administrative controls, for example organisation structure and governance systems, along with more socially based controls such as values and culture. Organisations may have numerous controls present, and they all may be used to some extent to align individual’s activities with organisational goals.”
This set of controls was already described by Flamholtz et al. (1985), who proposed a classification in which some of the controls mentioned by Malmi & Brown (2008), such as culture and structure, were regarded as control-context variables. Nevertheless, they assumed several interrelationships between the various controls. It follows that the various controls within a package are interdependent, and therefore, as Abernethy & Chua (1996, pp. 573) asserted, they should be internally consistent, which means that they are designed to achieve similar ends. Mak (1989) argues, as a moderate representative of configurationalism (Donaldson 2001, pp. 141), that this internal consistency among the various controls within a package may be of more importance for organisational performance than the fit between a control package and its contingencies. So it may be important, while observing some deviation between the configuration of a control package and the requirements due to changing environmental and organisational characteristics, to keep the alignment of the controls intact. It is likely that this will be a complicating factor in dealing with the problems arising from deviations, for it makes the tempting solution of adjusting individual controls less attractive.
In the literature, several variables can be found by which the internal consistency of a control package can be measured. Before elaborating on these in section 2.5, the next sections will provide a more comprehensive description of the various elements that make up a control package.
2.3 A framework for analysing control packages
Over the last thirty years, researchers have developed many alternative typologies, or frameworks, in order to describe and analyse MC-practices within firms. Sandelin (2008, pp. 325) listed some influential frameworks as follows:
techniques, or at least of practices mediating information, thus ignoring more subtle ways of motivating and coordinating organisational behaviours. Also, Simons’ (1995)
framework focuses on formal, information-based control practices and is primarily applicable to investigation of how managers use rather than design MCS in an attempt to control strategy. Moreover, the organisational control framework of Flamholtz (1983, 2005), Flamholtz et al. (1985) seems to lack specificity although it is informative about the need to align organisational control elements.”
Based on these considerations, Sandelin has chosen to use the prominent object-of-control framework proposed by Merchant & Van der Stede (2007), while admitting that this framework has its limitations in explaining relationships between individual controls (Sandelin 2008, pp. 326). In order to circumvent these shortcomings, the remainder of this paper builds on another, recently developed framework, proposed by Malmi & Brown (2008). According to these authors, their framework integrates aspects and elements of the various predecessors that have been proposed over the last four decades, in an attempt to compile a framework that is as comprehensive as possible. Although they did not have the intention to eliminate the before-mentioned shortcomings, their framework appears useful because its classification of controls captures the richness of several significant research streams associated with the various individual controls. Figure 1. provides the framework graphically.
Fig. 1. Control package. Source: Malmi & Brown (2008, pp. 291).
2.4 The Malmi & Brown (2008) control package
As shown in Fig. 1, there are five main types of control in this typology: planning, cybernetic, reward and compensation, administrative and cultural controls. Each of them will be described below, while drawing not exclusively on the Malmi & Brown (2008) paper, but also on descriptions in former typologies, in order to create a more comprehensive picture.
2.4.1 Planning controls
Planning is focussed on target setting and on the alignment of organisational and individual goals ex ante (Flamholtz et al.1985, pp. 39). Malmi & Brown (2008) made a distinction between action planning (usually a twelve-month period) and long-range planning. The latter seems very similar to Merchant & Van der Stede’s strategic planning (2007, pp. 331).
2.4.2 Cybernetic controls
Green & Welsh (1988, pp. 289) defined cybernetic control as follows:
“By cybernetic, we mean a process in which a feedback loop is represented by using standards of performance, measuring system performance, comparing that performance to standards, feeding back information about unwanted variances in the system, and modifying the system's comportment.”
According to Malmi & Brown (2008), especially the linking of behaviour to targets and the associated accountability for variations in performance upgrade a cybernetic system from being a decision-support system to a MCS (see also Zimmerman 2001). They distinguished four basic cybernetic systems that can be part of a control package: budget systems, financial measurement systems, non-financial measurement systems, and hybrid measurement systems.
293). According to Vaivio (1999, pp. 410), they could provide more comprehensive control, without being hindered by the limits of financial measures. Examples of non-financial measures are customer satisfaction and market share. Finally, hybrid systems integrate financial and non-financial measures in order to create a comprehensive picture of the performance of the organisation. The most prominently used hybrid system is the balanced score card (BSC) proposed by Kaplan & Norton (1996), which consists of four interrelated perspectives (top-down): financial, customer, internal business processes, and learning and growth. Core of this system, or model, is an assumed causal relationship between the perspectives. Although strongly criticised by Nørreklit (2000) for this alleged causality, the model has become an integral part of MC practice (Ittner & Larcker 1998, pp. 217).
2.4.3 Reward and compensation controls
Although often regarded as supporting cybernetic controls, reward and compensation systems are a separate control type in the framework (Malmi & Brown 2008, pp. 293), because they can have other functions, including retaining employees and encouraging cultural control. The primary objective of this kind of control system is to increase motivation and performance, by aligning goals and activities of individual employees with those of the organisation (Bonner & Sprinkle 2002). It has three main types of benefits: direction of effort, inducement of effort, and supporting employee selection (Merchant & Van der Stede 2007, pp. 395; Bonner & Sprinkle 2002, pp. 306).
2.4.4 Administrative controls
Malmi & Brown (2008, pp. 293) describe administrative control systems as a collection of controls that direct employee behaviour through the defining of responsibility structures, the organising of groups, the monitoring of performance and the guidance of behaviour. Therefore, administrative control sets the terms for assigning rewards and compensation.
the systems used for matching and coordinating activities of the various functions and hierarchical levels. Policies and procedures provide clarity about what behaviour to exhibit or to avoid. Merchant & Van der Stede (2007, pp. 76-79) call them action controls, and made a further distinction between behavioural constraints, pre-action reviews, action accountability, and redundancy.
2.4.5 Cultural controls
The last type of control to be described is cultural control, subdivided into three elements: clans, values, and symbols. When organisational culture is employed to regulate behaviour, according to Malmi & Brown (2008, pp. 294), it can be regarded as (part of a) control package. Clan control has been defined by Ouchi (1979), who considers a unique organisational (sub)culture as a clan. Within this clan, a socialisation process creates alignment between individual and organisational, or group interests, through the establishment of shared values, such as beliefs about environmental awareness, and symbols, such as dress codes or building/workspace design (Malmi & Brown 2008, pp. 294-295).
2.5 Macro control variables
Besides using a control framework, as elaborated above, for describing the control package of an organisation, several macro variables for characterising that control package are distinguished. Fisher (1995, pp. 28), for example, provides a non-exhaustive overview of such variables, which have the form of a continuum, and are usually described by their extremes. Figure 2. contains a list of the most prominent macro variables that are proposed in the literature.
• Tight vs. Loose control
• Interactive vs. Programmed control • Coercive vs. Enabling control • Objective vs. Subjective control • Behaviour vs. Results control
Fig. 2. Macro variables for classifying a control package.
staff specialists for the preparation and interpretation of control information, while interactive control requires an active attitude from all management levels (Simons 1995). The coercive vs. enabling control variable focuses on the degree of worker empowerment and workers’ freedom of action (Ahrens & Chapman 2004). The meaning of the other variables in the figure may be considered straightforward. The objective vs. subjective control variable is focussed on how employees are assessed by their superiors; based on subjective management judgment, or based on objective performance information (Fisher 1995, pp. 28). The behaviour vs. results control variable (Ouchi 1979) focuses on the choice between control based on “hard” results and control based on “soft” behaviour.
Besides to describe the general orientation of a control package, those macro variables could also be useful for judging the internal consistency of a control package. For being internally consistent, it could be expected that all of the control types within the control package should be positioned equally on a particular continuum. Fisher (1995, pp. 28) states this as follows; if a control package is defined as “tight”, al control mechanisms are used relatively strictly. For the purpose of this paper, the macro variables described can be helpful, because they can be regarded as the muscles in the skeleton of an internally consistent control package.
Together, the control framework and the macro variables represent the constitution and orientation of a control package, which presumably must be in line with its contingencies (i.e. the characteristics of the organisation and its environment). In the next section, the contingency theory will be elaborated in order to identify those possible characteristics.
2.6 Environmental and organisational characteristics
The contingency approach is based on the premise that optimal functioning of a control package depends on the particular elements of a firm’s context. Otley (1980, pp. 413) states this as follows:
“[…] a contingency theory must identify specific aspects of an accounting system which are associated with certain defined circumstances and demonstrate an
appropriate matching.”
According to the contingency theory, this appropriate matching will result in increased organisational performance (Fisher 1998, pp. 48). In order to be able to measure the influence of contingency variables on a control package, this paragraph will give an overview of the most prominent contingency variables that are in use in current contingency research.
recognition (Fisher 1995; Gresov 1989). This approach assumes interrelationships between individual contingencies, and acknowledges that MC (and therefore a control package) must be tailored to multiple contingencies simultaneously. Where different contingencies impose conflicting demands on a control package, a trade-off has to be made.
An elaborate overview of possible contingent control variables is provided by Fisher (1998, pp. 50). Fig. 3. provides this overview graphically.
Fig. 3. Contingent control variables. Source: Fisher (1998, pp. 50)
As mentioned in section 2.1, in this paper, strategy (category four in figure 3) is considered to be part of, instead of being contingent to, the control package. Therefore, the focus will be on the remaining four contingency categories, which will be introduced below.
complexity. Obviously, uncertainty can be regarded as one of the most prominent challenges for being in control, because the latter implies the avoidance of surprises.
In the contingency literature, technology and interdependence (the second category) are treated as a separate category, although, in fact, they could be regarded as part of the industry, firm and unit variables (the third category). Variations in size, structure, technologies used, and similar variables, are expected to put different demands on controls. Finally, according to the fifth category in figure 3, managers often have to choose between measurement of outcomes and measurement of behaviour, because both are not always possible, due to the nature of the technology used and the tasks assigned (Ouchi 1977, pp. 100). This choice
All these contingencies may have an impact on a control package as a whole, or on specific control types within the package. Many of those contingencies are subject to change, which makes it difficult for management to keep their control package up-to-date.
2.7 Management control fit and misfit
A fundamental problem, which makes a perfect fit between a firm’s control package and its contingencies nearly impossible beforehand, is mentioned by Fisher (1998, pp. 56), and is in line with the multiple contingent approach described before (see also Fisher 1995). He asserted that it is impossible to tailor a control package to all contingency variables simultaneously, because of conflicting demands placed on the package by the various contingencies. It follows that, in designing a control package, trade-offs must be considered in prioritising the conflicting contingencies. Here, these trade-off decisions are called “design problems”. Once designed, the control package has to be implemented, what could give rise to what could be called “implementation problems”, for instance misunderstanding and resistance of staff. Finally, reconfiguration of the control package in its entirety, or additions or deletions of individual controls, may be necessary as a result from changing environmental and organisational characteristics. This may result in what could be called “reconfiguration problems”. According to Mak (1989, pp. 274), this may be costly and may result in an internal inconsistent control package, which may lead to non-optimal performance.
2.8 Control problems
With the conceptual toolbox developed above, it is now possible to elaborate on the control problems resulting from a certain level of misfit between a firm’s control package and its contingencies. The misfit can take two appearances. First, individual controls may be non-optimally designed in the light of one of the changing contingencies, i.e. the control package suffers from technical imperfections. Second, the control package as a whole, as described by the macro variables mentioned in paragraph 2.5, may deviate from the requirements imposed by a particular contingency (i.e. the control package is used in a non-optimal fashion).
As mentioned in the introduction, Merchant & Van der Stede (2007, pp. 10-11) provide a list of what they call “management control problems”. Those problems, which are personal in nature and strongly focused on behaviour, are: lack of direction, motivational problems and personal limitations. Lack of direction means that employees do not perform properly because they do not know what the organisation wants to achieve and/or what their individual role is. Motivational problems commonly arise due to effort aversion or self-interested behaviour. Examples include wasting, mismanaging, falsifying, stealing and abusing organisational resources. Problems of personal limitations become visible when employees, who know what is expected from them, and who are highly motivated, are unable to perform appropriately, because of, for instance, a lack of required intelligence, training, experience, stamina, or a lack of knowledge or information. These problems are called MC-problems because, according to Merchant & Van der Stede (2007), they can be regarded as the primary inducement for the existence of MC. However, as stated in the introduction, they cannot be linked directly to a certain level of misfit between a control package and a firm’s contingencies, which is an organisational matter, leading to organisational problems. Therefore, a second set of control problems can be suspected as a result from control misfit: organisational control problems. Those could be, for instance, communication problems, information quality problems or reputation problems. This idea is visualised in figure 2.
(mis) fit
Fig. 4. Conceptual framework
Unfortunately, such organisational problems, with a special focus on the relationship between a firm’s control package and its contingencies, are not described in the MC literature. Therefore, empirical research is required to identify and to analyse them. The results of this analysis will provide, additional to the theoretical concepts and constructs presented above, the foundation for the empirical investigation of the final research question: how do managers handle the problems resulting from non-optimal control. In the next chapter the design of this twofold empirical research will be described.
For a good understanding of the analysis of the empirical results it is important to note that the term “misfit” itself does not represent any problem or other concrete phenomena. It can only be used to indicate that the control package and any contingency do not perfectly fit. This misfit can not be identified itself, it can only be determined through the identification of the sympoms it causes, namely the control problems. Therefore, the empirical reseach will start with an investigation of the organisational control problems. When these are mapped clearly, it is time to investigate their relationship with the contingency misfit and the MC-problems. So, in the upcoming chapters, the primairy objective is to show that organisational control problems can be, at least in several occasions, the key for the analysis of how to eliminate the misfit, which, according to the conceptual framework, will automatically lead to a reduction of related MC-problems.
Chapter three will describe the empircal research design. Chapter four contains the results and the analysis of the empirical results. Finally, chapter five contains a conslusion and some theoretical considerations. Contingency variables Control package Organisational control problems MC-problems: (Merchant & Van der Stede 2007):
- lack of direction - motivational
problems - personal
3. Design and research method
The empirical research draws on an exploratory field study of a relatively small Dutch organisation 2 . Facing several important institutional and environmental changes and developments, which have a strong impact on its MC, this organisation is ideally suited for the analysis of MC-problems. First, this chapter provides an argumentation for the choice of the research method used. Thereafter, it provides an overview of the methods of data collection and data analysis. Finally it includes a short introduction of the case organisation.
3.1 Research design
The research was conducted in the form of an exploratory field study (Yin 2003, p. 6), because of this method’s theory building potential, verifiability, and empirical validity (Eisenhardt 1989, p. 547). This research method is expected to provide maximum insight in the practical relevance of the proposed framework, and can probably reveal specific organisational control problems. Furthermore, it provides an opportunity to evaluate the way managers handle those problems. So, the two empirical research questions are both covered by this research design. Although, for generalisability purposes, a multiple-case study probably would have been a more optimal research method, this study is confined to a single case, because of limitations with respect to time and resources. Nevertheless, the case used can be regarded as illustrative because it includes all elements of the conceptual framework developed above, justifying a single-case study (Yin, 2003, p. 41), and allowing for theoretical generalisation (Lukka & Kasanen 1995). Moreover, according to Flyvbjerg (2006), a properly conducted single-case study is a strong and effective instrument for in-depth research focused on such a dense, versatile subject.
3.2 Methods of data collection and analysis
Data were collected through direct observation, document study, and interviews, over a
period of four months (March till June 2009). The company gave the author access to the day-by-day working environment, for just “being around”, which provided the opportunity for informal conversations about organisational matters, working conditions and worker attitudes. In addition, the author was provided with several documents, ranging from a general mission statement and a strategy roadmap to very detailed consultancy reports and improvement plans. Especially the latter ones provided information directly relevant to the reasoning in this paper. In the final stage of the research period, five semi-structured interviews were carried out, lasting, on average, about 50 minutes. The interviewees were carefully selected in order to cover the entire control package
2
used and the most important organisational and environmental developments. They included the chief executive officer (CEO), the business controller, the operations manager and the human resource manager. Because of his recent departure, also the former business controller was interviewed. All interviews were tape-recorded and transcribed verbatim, in order to be able to make a comparison of allegations and findings. Together with the data from the documents provided by the organisation, and the day-by-day experience and conversations (listed in draft), these interview reports constitute the complete data set.
Data analysis started with an investigation of the most important contingency changes which the case organisation experienced, for which the daily conversations proved most useful. Additionally, the most important organisational and management control problems were identified, by combining the data from the consultancy reports, which focused primarily on those problems, with the day-by-day information. During the interviews, this knowledge was again verified by questioning the interviewees about these topics. The interviews also provided the opportunity to map out how the interviewees thought about how to address the problems. To prepare for the interviews, a general questionnaire was composed, which covered all the relevant aspects of the analysis. However, because each interviewee has its own educational background, managerial responsibilities and organisational focus, this questionnaire was only used in general terms. This provided the opportunity to focus on the particular parts of the questionaire that are of special interest for the interviewee in question, as the human resource manager is particularly involved in reward and compensation controls, the controller in planning and cybernetic controls, and the operations manager in culture and planning controls. This approach allows for in-depth understanding of the various focus areas, and it delays the moment of theoretical saturation (Glaser & Strauss 1967); i.e. the point where the interviewer is able to predict interviewee responses based on knowledge obtained from previous interviews and other data sources. This implies a broader scope and provides access to a more extensive set of information.
4. Empirical results and analysis
This chapter discusses the results of the empirical research. First, it describes the case organisation and the most important changes in the contingencies of the organisation. Thereafter, it provides a description of the organisational problems that the management of the organisation identified, and, then, how these organisational control problems led to MC-problems. Based on this description, an analysis follows of how the managers handle the MC-problems they face.
4.1 Case introduction
The case organisation is a former subsidiary of a large monopolist. Recently, regulations have changed which have led to a separation. Now, it has to survive as a stand-alone company in a competitive market. The organisation has approximately one hundred and twenty employees, divided over five departments. About forty percent of the employees is working in the field as technicians, while the remaining sixty percent conducts supporting activities, which range from data management, customer acquisition and marketing to billing and after-sales activities. The organisation is, as mentioned before, subject to important contingency changes, which have a strong impact on the MC-practice within the organisation. However, it seems difficult for the organisation to adjust its control package instantly. In terms of the theoretical discussion in chapter two, the organisation faces a certain level of misfit between its control package and some of its contingencies, which has led to certain organisational and MC-problems. In the next section, several of the problems experienced by our organisation will be described in an attempt to relate them to each other, to the control package and to the changing contingencies. In other words, the case provides an opportunity to investigate the relationships within the conceptual framework constructed in chapter two. This is crucial for the subsequent discussion about how to solve the problems faced, which will be the focus of the second part of this chapter.
4.2 Contingency changes
At the time of this study, the case organisation experienced several contingency changes. One of the most prominent ones was a major shift in legislation. For decades, the organisation had been part of a larger company. Despite the fact that its activities were highly distinct from the other activities of the company (traditionally, it regarded itself as a maverick). The influence of the parent on the case organisation’s MC-practice was substantial. Recent changes in legislation had pushed the organisation further away from the parent; a development that has had several implications for how its MC was organised and used.
organisation distinguished itself especially through the organisation and execution of its core activities and through the use of innovative technologies. Unfortunately, this competitive advantage devaluated, as competitors increasingly mastered those core competences and technologies.
A less important, but noteworthy, contingency change was aging of the workforce. This would, in the eyes of the interviewees, not have a major impact on the control package. However, they admitted that it could be seen as an indication of how the organisational culture is going to develop, and this development could give rise to changes in cultural controls. For instance, as they asserted, flexibility and entrepreneurship may become less important for employees when they get older.
One of the effects resulting from the separation of the case-organisation from its parent company is quite straightforward, but for this analysis of utmost importance. In the past, when the ties with the parent company were much stronger, the control practice within the case organisation was highly influenced by the parent company. However, the parent company did not have a commercial focus, and neither did its controls. The increased separation from its parent damaged the control package of the case-organisation because it made several controls to disappear, while the organisational focussed increasingly on commercial aspects, due to the increased market competition. As a result, at the time of this study, the misfit resulted from an incomplete control package, which also had to be adjusted to the increased focus on commercial issues. Because a degree of misfit is not directly measurable and quantifiable through an analysis of the technical design of the control package, it mainly has to be deduced from the problems that emerge when using the package. Therefore, in the next section, a description follows of the organisational and MC-problems as evidenced by the empirical research.
4.3 Organisational control problems resulting from misfit
4.3.1 Planning controls
The shortcomings with respect to the planning controls were mainly long-term; the short-term action planning was considered to be fairly sufficient. First, the long-short-term strategic planning controls, such as the mission, vision and strategy, were not documented. Although the management certainly had plans and strategies for the future, the documentation of those plans and strategies was a hot issue during the four months of empirical research. Obviously, this was a result of the increased separation from the parent company, which made the former mission, vision and strategy less applicable or even obsolete. The absence of clearly formulated strategy, in combination with the shift towards commercial thinking, contributed to a feeling of lack of direction. Several interviewees asserted that, in their view, the overall situation of the organisation, including the positioning in the market, was relatively unclear so far. This was also evident from the consultancy reports; due to the former monopoly situation of the company to which the organisation belonged, knowledge of the market was limited. Another interviewee emphasised the importance of strategic planning for balancing conflicting interests of the different departments. In his view, the planning was not always sufficient for this purpose, resulting in arbitrary and sometimes subjective strategic decisions. These facts led to a degree of strategic uncertainty that would probably be too high, because the organisation might not be sufficiently prepared to take advantage of upcoming opportunities in the environment to deal with. Although somewhat dependent on individual personality characteristics, this uncertainty increases the interviewees’ feeling of lack of direction.
4.3.2 Administrative controls
The organisational and governance structures, including policies and procedures, could be seen as the fundamental controls in the control package. When those structures are not optimally shaped in the light of the contingencies, the entire business process will suffer from ineffectiveness. As one interviewee remarked: “Those structures are the framework within which the MC takes place”. Perhaps, the most important problems that emerged as a result of shortcomings in those structures are communication problems. Obviously, the essence of organisational and governance structures is communication. In the past, another shift in regulation forced our case organisation to reorganise its structures because, in the light of this contingency change, the alignment of the various departments became sub optimal. The current contingency shifts, as described in section 4.1, are creating a corresponding problem. The interfaces between the various departments were not optimally designed to meet the new requirements of the changed market. For acting proactively towards customers, in order to be competitive, customer orders should flow unhinderedly and smoothly through the business process, which covers all departments within the organisation. However, according to the interviewees, this was certainly not the case. The (imaginary) interfaces between the departments, where the work has to be transferred, were not transparent, and employees, including the management, did not sufficiently communicate their information needs and requirements to the other departments. This led to problems, because often employees did not receive the information they required (i.e. lack of information or knowledge, a personal limitation). Consequently, throughput time took longer than necessary, and frustration arose. Such a self-directed attitude of the departments and their managers (as also shown by the issue about local and central planning) may obstruct the decision-making ability within the organisation. According to the interviewees, this may increase employees’ feeling of lack of direction, at least with respect to beyond-departmental issues. The frustration may result in motivational problems such as effort-aversion. In other words, the misfit between organisational and governance structure and contingencies ultimately may lead, through departmental alignment problems and decision-making problems, to MC-problems.
4.3.3 Cultural controls
three generic competencies that were considered indispensable for each employee within the organisation: learning ability, innovativeness, and pro-activity. But, due to their general scope, and because of the uniqueness of culture, it is almost impossible to assess these competencies/values with respect to their contribution to optimal organisational performance. However, culture will be important for this analysis, but in a slightly different meaning.
When culture is not confined to “hard” elements able to influence behaviour, but instead described in more “soft” terms, some interesting points can be made. In the words of the CEO of our case organisation, the working environment was traditionally focused on “extinguishing fires wherever they emerge”. In contrast, structural monitoring and managing received less attention. Troubleshooting was an important aspect of daily management. This fire-fighter culture, which the interviewees described as reactive, became less applicable since the contingency changes required the organisation to operate increasingly proactively. Its applicability also deteriorates with a view to the increased importance of business process efficiency and effectiveness, which requires structural monitoring and tempering of ad-hoc decision-making. This misfit between culture and contingencies may make it hard for the organisation to be as competitive as desired (i.e. it results in sub optimal performance). However, the empirical results did not provide any indication that it resulted in MC-problems of lack of direction, motivational problems or personal limitations.
4.4 Macro control variables and control problems
As noticed before, the management control literature provides several macro control variables for characterising the way a control package is used. In order to fit the control package to its contingencies, an organisation not only has to develop a technically suitable control package, but it also has to use the control package appropriately (i.e. it has to position itself on the macro control variable continua introduced in section 2.5), in order to meet the demands imposed by the contingencies. This section will elaborate on the relationships between the control package orientation, as expressed by the macro control variables, and any perceived non-optimal performance as mentioned by the interviewees.
which emphasised an organisation-wide feeling that a continuous stream of plans and initiatives was developed, but many of them remained unfinished, because, during their implementation, the focus of the management had already moved on to newer ones. These problems can be characterised as a combination of misplaced focus as well as effectiveness and efficiency problems. Obviously, this combination may result in feelings of lack of direction among employees who cannot ever be sure about the importance of their work; perhaps, tomorrow, the attention for it by management declines to almost nothing. Clearly, this could lead to motivational problems.
Second, the interviewees qualified the control package as slightly programmed (and thus less interactive). In their view, for example, departmental managers should be more actively involved in the process of preparing and interpreting control information, because they expected this will improve the communication about individual and departmental performance, and will reduce the quantity of this information, which reasonably improves its quality. Till now, some interviewees had the impression that a certain part of the control information was produced automatically, without any input from the manager in question, who leaves this information unused. This can be seen as related to the information problems that are described in section 4.2.2 about cybernetic controls.
Fourth, the interviewees indicated that subjective and objective aspects play an almost equivalent role in the control practice. Because objective control obviously focuses mainly on output, and subjective control has more to do with behaviour, both variables will be described jointly. The control package had, according to most of the interviewees, a strong emphasis on output, although some interviewees did not refer to any inequality. The interviewees were not able to suggest any practical improvements with respect to the positioning of the control package on these variables. However, they preferred more emphasis on behaviour, because, as one interviewee said, short-term improvements in behaviour could result in long lasting improvements in output. Therefore, they agreed that a certain level of subjective control would be necessary. At the same time, they agreed upon the fact that subjective control can lead to discontent among workers, because it could be perceived as inconsistent. This discontent could lead to self-directed behaviour of employees, which is another word for motivational problems. Also in this case, the relationship between those problems and the misfit is a direct one, without any intervention of organisational control problems.
4.5 Empirical support for the conceptual framework
Based on the empirical results, an answer to the third research question can be formulated. The above problem analysis provides empirical support for the idea expressed in the conceptual framework that a great deal of the MC-problems originate, at least partly, from organisational control problems which, in turn, may result from a certain degree of misfit between the control package and its contingencies. As evidenced above, misfit can take two appearances; it may have its cause in the technical composition of the control package, or it can result from the way in which the control package is used by management. However, this dichotomy does not result in extra diversity in the nature of the problems that represent any misfit; both appearances create comparable problems. Several organisational control problem categories have been identified. The first, and most straightforward, is technical incompleteness of the control package, as described, for instance, in the context of the planning controls. The absence of any company specific mission/vision documentation resulted in strategic uncertainty, which created feelings of lack of direction. A second organisational control problem was operational effectiveness and
efficiency problems. In the above description, this was the result of the use of a planning system
pursuit of sub optimal strategies and objectives. The final types of important organisational control problems identified were communication problems and decision-making problems, resulting from the imperfect design of the administrative controls. Besides through those organisational control problems, MC-problems can also result directly from a certain degree of misfit, as evidenced in section 4.3.
In summary, it can be concluded that the ideas expressed in the conceptual framework are supported by empirical evidence. This conclusion provides managers with an integrated understanding of the relationships between contingential developments, management controls (the control package), organisational control problems and, ultimately, MC-problems of lack of direction, motivational problems and personal limitations. This understanding may be helpful for improving organisational performance, as will become clear in the next section.
4.6 A general approach for dealing with the control problems described above
In this section, an answer will be sought to the last research question: how can organisations handle the problems identified? Here, we do not consider it fruitful to address each problem individually, in an attempt to produce tailor-made solutions. Instead, this section attempts to describe a general approach for control optimisation by using the reasoning developed in the conceptual framework.
Before trying to find solutions to any problem, it is important to distinguish between causes and effects, between actual problems and symptoms of problems. Sometimes, it is only possible to fight symptoms without addressing the underlying problem. However, it seems more efficient, and better from a survival perspective, to eliminate the cause, which causes the symptoms to disappear. Therefore, in the context of this paper, before being able to make claims about how to solve any management control problem, it is important to clarify whether a specific problem (category) is a cause or a symptom.
for instance, activity elimination or automation (Merchant & Van der Stede 2007, p. 12), is fighting the symptoms, i.e. the resulting problems. From this perspective, behaviour-influencing devices need to be implemented in order to overcome the MC-problems. Their definition of controls (behaviour-influencing devices) is therefore much more confined than the definition that is used in this paper. In contrast to this negative view on human nature, our interviewees preferred to take a more positive view; they initially expect an external cause for the personality-related MC-problems of their subordinates. In other words, they regard MC-problems as symptoms resulting from more fundamental organisational problems. This principled position has significant implications for how MC-problems can be resisted and contested.
As evidenced in section 4.2, MC-problems can stem from organisational control problems. However, also these organisational control problems can be traced back to root causes, i.e. misfit between the control package and its contingencies, both in the form of technical imperfections of the control package or in the form of wrongly positioned macro variables. Because contingency changes, by definition, cannot be influenced by the organisation, this is the level on which managerial adjustments to the control package can provide fundamental solutions that may eliminate organisational and MC-problems.
Obviously, when the technical composition of the control package is non-optimal, the package has to be adjusted, and when managers conclude that the control package is wrongly positioned on one or more macro variables, the logical solution is repositioning. Unfortunately, the practical implementation of these solutions is less straightforward. Most importantly, managers require “tools” for these adjustments and this repositioning. Our empirical research reveals three such tools which managers consider important in this context. In the eyes of the interviewees, each managerial initiative requires decisions about what leadership style is required in order to reach the objectives, how much worker empowerment is justified given the circumstances, how to communicate the decisions through the organisation in order to obtain consent of all employees involved, and how to design the communicational aspects of the operational process after the policy change.
Technical adjustments to the control package, which have to be made because of contingency changes, such as the introduction of a changed authority structure (administrative control), are usually very specific, which makes it difficult to make detailed claims about how to use the managerial tools in order to support the adjustments.
more aloof from everyday business. This modified leadership style implies an increase in the degree of trust that management provides to its subordinates, and it requires a communication style that overcomes the increased worker independence. Or, as another example, when management wants to make its use of the control package more programmed, it may require a more authority-based leadership style and a more formal communication style, while trust becomes less important because employee initiative is less required.
Unfortunately, because of the structural tendency towards misfit due to the changing nature of the contingencies, and the reactive nature of the above-mentioned adjustments, there will always be MC-problems left which have to be countered by using behaviour-influencing devices. Therefore, the management controls as defined by Merchant & Van der Stede (2007) retain their functionality. However, in the reasoning developed above, they are regarded as a complement to the control package adjustments, or, in other words, as a second safety net.
5. Conclusion
This final chapter starts with a summary of the reasoning developed in this paper based on both literature research and empirical research. Next, it contains a discussion of several theoretical issues related to the subject, which ultimately results in a conclusion about the content and the usefulness of the theory. Finally, some recommendations for future research are suggested.
5.1 Summary / discussion
Obviously, every manager strives for optimal control. This truism immediately leads to the question of how to achieve this aim. In the management control literature, two theoretical schools attempt to shed light on this issue, both from their own perspective. On the one hand, a behaviour-oriented school, represented by, among others, Merchant & Van der Stede (2007), states that optimal control can be realised by directing the behaviour of employees through the use of behaviour-influencing devices, called management controls, in the expectation that organisational performance will be optimal only then, when all employees act entirely in the interest of the organisation. On the other hand, contingency theorists, who take a broader view, assert that control can only be optimal when the package of controls used by the organisation best matches the requirements imposed by all its contingencies. If not, the logical consequence is misfit between de control package and its contingencies, which is associated with organisational problems. Now the question arises: can these problems, together with the behaviourally based problems implied above, be regarded as two sides of the same coin? The empirical research in this paper shows that they are not. Instead, they are part of a sequential relationship: the non-optimal control (i.e. organisational control problems) resulting from any misfit between a firm’s control package and its contingencies may lead to employee misbehaviour and limitations (management control problems), which causes non-optimal organisational performance.
management control problems. In short, the model provides a more structured reasoning for countering problems of various kinds.
The theoretical and empirical research provides a picture of how managers deal with the various problems. First, it suggests two types of misfit: it may appear in the form of technical deficiencies in the control package, or it may be reflected in non-optimal application of the control package. Both types of misfit can lead to management control problems (i.e. lack of direction, motivational problems and personal limitations), with or without intervention of organisational problems such as communication problems, efficiency problems, information problems, and decision-making problems. The research reveals that, although the use of behaviour-influencing devices is considered a necessity, managers expect to be able to counter the above-mentioned problems by making adjustments in the technical composition of the control package, or in the way the package is used. To do this, they claim to make use of the following tools: leadership, trust and communication.
Logically, the integration of both theories reveals some theoretical inconsistencies. For instance, as described in this paper, the term “controls” has as many definitions as there are writers on the subject. Merchant & Van der Stede (2007) limit themselves to controls focussed on behaviour, while contingency theorists use controls mainly in an organisational context. This difference implies, among other things, that the two theories diverge, while both focus on performance optimisation. In order to overcome these differences, this paper makes use of a recently published description of control that incorporates both organisational and behaviour-focussed controls.
Another point of contention, which has major implications for how to handle MC-problems, is the nature of human behaviour. Merchant & Van der Stede (2007) choose a negative approach; they expect that employees themselves tend to act self-interested and non-optimally from the perspective of the organisation. Therefore, in their approach, behaviour-influencing devices are indispensable. In practice, as evidenced in the empirical research, managers may take a more positive approach on human nature. This provides the option to view problems concerning employee behaviour and employee performance as symptoms of underlying problems.
which makes alignment difficult anyway. Therefore, behaviour-influencing controls may be regarded indispensable. However, they are not one and all. Realising that making adjustments to the control package can prevent such problems ensures a more balanced composition of that particular control package.
This realisation is the most distinctive feature of the integrated model, in comparison to the two separate theories. The contingency theory, on the one hand, remains vague on what problems result from any observed misfit, so the relationships between problems and possible solutions remain at an abstract level. This makes it difficult to make claims about any adjustments to the control package. The Merchant & Van der Stede (2007) approach, on the other hand, suffers from the lack of a reference point from which the composition of the control package can be assessed on the criteria of internally consistency, efficiency and effectiveness. It has no mechanism to counter proliferation of management controls, because any undesirable employee behaviour has to be countered with a specific, individual control device, and there is no other mechanism for harmonising this “collection of controls” into a coherent control package than just common sense. The integrated model developed in this paper provides such a reference point, in the form of the alignment of the control package to its contingencies. Over and over, the composition of the control package has to be assessed on its abilities to meet the requirements imposed by its contingencies, which ensures internal consistency.
Based on the above, it can be concluded that the integrated view on the nature and origin of control problems, which is elaborated in the conceptual framework, is supported by the empirical results, and may be a foundation for managerial decision-making on management control. Nevertheless, there are several aspects that require further investigation. These will be described in the next, final section of this paper.
5.2 Recommendations for future research
focus not only on changing contingencies, but, in a broader context, to the structural influence of contingencies on MC-problems.
Second, it may be fruitful to examine in more detail the relationship between human nature and the various MC-problems, which is primarily a psychological or sociological issue. To what extent are MC-problems inherent to human nature? Or, in other words, do people behave optimally from the perspective of the organisation when the control package perfectly fits its contingencies (i.e. when the organisation is perfectly designed)? As described in this paper, the choice of how to approach the MC-problem issue depends strongly on one’s view of human nature; a choice which is optional so far. Perhaps, this question may not be answered at all, a possibility which may prove to be the most significant threat to the theory suggested in this paper. Finally, the paper contains only a very limited description of how managers cope with the different control problems, for the sole purpose of showing that the theory holds water. However, for the theory to be useful, it is necessary to describe in more detail how managers can fit their control packages to their contingencies in the first place, and also how they can position their control packages on the various macro control variables and how they can maintain this positioning. In addition, research is needed to investigate the value of the defined organisational control variables as signals for the existence of any misfit and as part of the solution for the MC-problems.
Bibliography
Abernethy M.A. & Chua W.F., (1996). A Field Study of Control System “Redesign”: The Impact of Institutional Processes on Strategic Choice, Contemporary Accounting Research, Vol. 13, No. 2, pp. 569-606.
Ahrens T., & Chapman C.S., (2004). Accounting for Flexibility and Efficiency: A Field Study of Management Control Systems in a Restaurant Chain, Contemporary Accounting
Research, Vol. 21, No. 2, pp. 271-301.
Bonner S.E., & Sprinkle G.B., (2002). The Effects of Monetary Incentives on Effort and Task Performance: Theories, Evidence, and a Framework for Research, Accounting
Organizations and Society, Vol. 27, No. 4/5, pp. 303-345.
Cadez S., & Guilding C., (2008). An Exploratory Investigation of an Integrated Contingency Model of Strategic Management Accounting, Accounting, Organizations and Society, Vol. 33, No. 7-8, pp. 836-863.
Chenhall R., (2003). Management Control Systems Design within its Organizational Context: Findings from Contingency-based Research and Directions for the Future, Accounting,
Organizations and Society, Vol. 28, No. 2-3, pp. 127-168.
Donaldson L., (2001). The Contingency Theory of Organizations. Sage Publications, Thousand Oaks, California.
Eisenhardt K., (1989). Building Theories from Case Study Research, Academy of Management
Review, Vol. 14, No. 4, pp. 532-550.
Ekholm B.G., & Wallin J., (2000). Is the Annual Budget Really Dead, The European Accounting
Review, Vol. 9, No. 4, pp. 519-539.
Ferreira A., & Otley D.T., (2005). The Design and use of Management Control Systems: an extended Framework for Analysis. Working paper.
Fisher J.G., (1995). Contingency-based Research on Management Control Systems:
Categorization by Level of Complexity, Journal of Accounting Literature, Vol. 14, pp. 24-53.
Fisher J.G., (1998). Contingency Theory, Management Control Systems and Firm Outcomes: Past Results and Future Directions, Behavioral Research in Accounting, Vol. 10 (Supplement), pp. 47–57.
Flamholtz E.G., (1983). Accounting, Budgeting and Control Systems in their Organizational Context: Theoretical and Empirical Perspectives, Accounting, Organizations and Society, Vol. 8, No. 2-3, pp. 153–169.
Implications for Accounting, Information, and Control, Advances in Management
Accounting, Vol. 14, pp. 139–165.
Flamholtz E.G., Das T.K., & Tsui A.S., (1985). Toward an Integrative Framework of
Organizational Control, Accounting, Organizations and Society, Vol. 10, No. 1, pp. 35– 50.
Flyvbjerg B., (2006). Five Misunderstandings about Case Study Research, Qualitative Inquiry, Vol. 12, No. 2, pp. 219-245.
Gerdin J., & Greve J., (2004). Forms of Contingency Fit in Management Accounting Research – A Critical Review, Accounting, Organizations and Society, Vol. 29, No. 3-4, pp. 303-326.
Glaser B., & Strauss A., (1967). The Discovery of Grounded Theory: Strategies of Qualitative
Research. Wiedenfeld and Nicholson, London.
Green S.G., & Welsh M.A., (1988). Cybernetics and Dependence: Reframing the Control Concept, Academy of Management Review, Vol. 13, No. 2, pp. 287-301.
Greiner L.E., (1997). Evolution and Revolution as Organizations Grow: A Company's Past has Clues for Management that are Critical to Future Success, Family Business Review, Vol.
10, No. 4, pp. 397-409.
Gresov C., (1989). Exploring Fit and Misfit with Multiple Contingencies, Administrative Science
Quarterly, Vol. 34, No. 3, pp. 431-453.
Hambrick D., & Lei D., (1985). Toward an Empirical Prioritization of Contingency Variables for Business Strategy, Academy of Management Journal, Vol. 28, No. 4, pp. 763-788. Hope J., & Fraser R., (2003). Beyond Budgeting: How Managers can Break Free from the Annual
Performance Trap. Harvard Business School Press, Boston.
Ittner C.D., & Larcker D.F., (1998). Innovations in Performance Measurement: Trends and Research Implications, Journal of Management Accounting Research, Vol. 10, pp. 205-238.
Kaplan R.S., & Norton D.P., (1996). Linking the Balanced Scorecard to Strategy, California
Management Review, Vol. 39, No. 1, pp. 53-79.
Lazear E.P., (1998). Personnel Economics for Managers. John Wiley & Sons, New York. Lukka K., & Kasanen E., (1995). The Problem of Generalizability: Anecdotes and Evidence in
Accounting Research, Accounting, Auditing & Accountability Journal, Vol. 8, No. 5, pp. 71-90.
