A coupling perspective on the management control systems package

(1)

A coupling perspective on the management

control systems package

case “water board Hunze and Aa’s”

Master Thesis

J.J.S. Haanstra

University of Groningen

(2)

1

A coupling perspective on the management

control systems package

case “water board Hunze and Aa’s”

Abstract

This thesis discusses a public organization’s management control systems package and follows the suggestion made in literature, about studying the management control systems package by using coupling theory of Orton and Weick (1990). This thesis applies this theoretical framework, including refinements, to the single case study ‘water board Hunze and Aa’s’. This case study shows that an organization can include multiple different management control systems in its management control systems package and that the level of coupling between those systems can be different. Factors that influence the level of coupling have effectiveness, legal and agency origins. The case study also shows that two management control systems can have two levels of coupling simultaneously and that the level of coupling can change when time passes. Finally, this thesis discusses the advantages and disadvantages of using coupling theory and concludes that the theory is useful, but that improvements can be made.

Key words: management control systems (package), coupling theory, water boards

Contact information

Name : Jolina Haanstra Student number : 1710036

E-mail address : jolinahaanstra@gmail.com Telephone number : + 31 (0)50 3640858 Additional information

Country : The Netherlands University : Groningen

Faculty : Faculty of Economics and Business Master : Business Administration

Spezialization : Organizational and Management Control Supervisior : B. Van der Kolk

(3)

2

Preface

Dear reader,

In front of you is a very special page. It is one of the first pages of my thesis, but is simultaneously also the last page I wrote during my study time at the Rijksuniversiteit Groningen. Between 2007 and 2011 I did a Bachelor in Business Economics and from 2011 to 2013 I did a Master in Business Administration, specialization Organizational & Management Control. This thesis is the final sprint to my graduation of this Master. Many people have made this sprint possible. In the first place I would like to thank water board Hunze and Aa’s and especially the interviewees who gave me a very hearty welcome and who took the time to explain the ‘whats’, ‘hows’ and the ‘whys’ of their organization. In the second place I would like to thank my supervisor Mr. B. Van der Kolk who supported, advised and encouraged me along the ‘thesis way’. And last, but not least I would thank my family and friends who supported me already long before my final sprint was started.

Enjoy reading this thesis.

Yours sincerely,

(5)

4

1. Introduction

In the last decades much literature has been published about management control1. This is not surprising, because often it is hypothesized that organizational effectiveness largely depends upon the (management) control system (Merchant and Van der Stede, 2007; Herath, 2006; Fisher 1998). Management control failures can cause financial losses, reputation damage and possibly lead to organizational failure (Merchant and Van der Stede, 2007). An effective management control system on the contrary, encourages managers to make decisions to reach organization objectives (Fisher, 1998).

Usually organizations have multiple management control systems (Malmi and Brown, 2008; Merchant and van der Stede, 2007). The collection or a set of controls and control systems can be called a management control systems package (Malmi and Brown, 2008). Although the importance of studying the control package is acknowledged for decades (Abernethy and Chua, 1996), it has not been exposed to much academic attention yet (Malmi and Grandlund, 2009; Sandelin, 2008; Malmi and Brown, 2008 and 2005; Abernethy and Chua, 1996; Otley, 1980). In contingency-based research there is much attention paid to specific elements in a control system in isolation from other controls (Malmi and Brown, 2008; Chenhall, 2003, Fisher, 1998,). Studying the management control systems as a package, however, is important.

Firstly, management control systems do not operate in isolation, but they are part of a broader control system (Malmi and Brown, 2008; Otley, 1980). If the linkages between the control systems are not adequate, the control package may not fulfill its intended function (Fisher, 1998). For example, misalignment between control systems can result in ineffective control (Malmi and Brown, 2008). Gaining insight into management control systems package may facilitate the development of the design of management control systems, which can support achieving organizational goals (Malmi and Brown, 2008).

Secondly, according to Chenhall (2003), studying control mechanisms in isolation can lead to serious model under specification and can lead to spurious findings. Because of the interrelationships between control mechanisms, it is difficult to determine the isolated effect

1

(6)

5

of any specific means of control (Otley, 1980). According to Otley (1980) it is a ‘folly’ to construct a theory of a control system outside of the context of an overall control package.

Because of these reasons there have been multiple calls since the 1980’s, to study the management control systems package (e.g. Sandelin, 2008; Malmi and Brown, 2005 and 2008; Verbeeten, 2005; Otley, 1999; Fisher, 1998; and Otley, 1980). This thesis responds to these calls. The general research question in this thesis is: how are the management control systems of a public organization in a management control systems package coupled, and why are they coupled the way they are? Currently there is not much theory which can help researchers to study the relationships between the systems in a control package (Malmi and Brown, 2008). Suggestions have been made to use the framework of Orton and Weick (1990) for assessing the level of coupling between control mechanisms (Järvenpää and Länsiluoto, 2011; Malmi and Brown, 2008). This thesis will follow this research opportunity and will be supported by empirical research, because there is still no clarity about why and how parts of a control package are used and preferred in practice (Malmi and Brown, 2005). It has been advised to examine a single organization in studying the complex topic of control as a package (Otley, 1999). A case study will be conducted in the public context, namely in one of the Dutch water boards. Water boards are regional public authorities which are endowed with the responsibility for the quantity and the quality of (regional) surface water in the Netherlands (Lamers et al., 2010, Van Steen and Pellenbarg, 2004).

(7)

6

The aim of this thesis is therefore two folded; on the one hand I hope I can contribute to the limited body of knowledge about management control systems package, on the other hand I hope that I can help water boards in gaining more insight into their management control systems package, through which they maybe can improve their management control systems package so that they can meet the environmental demands and the social and political wishes.

(8)

7

2. Literature review

Many non-profit and profit organizations are large and complex. The consequence is that all decisions cannot be made by one person and that principals (e.g. managers) must hire agents (e.g. subordinates). This is called ‘decentralization’. The result of decentralization is that lower level managers have freedom to make decisions themselves (Bhimani et al. 2008). A principle, however, can have difficulties in observing agents’ efforts as a result of information asymmetry (Zimmerman, 2006). A large part of the (business) economics literature therefore predicts that people act in their own interest to maximize their own utility, which can be detrimental for the welfare of the principal and/or firm (the principal-agent problem), which leads to agency costs (Zimmerman, 2006). Employees can also perform inadequately, because they do not know what the organization wants from them (problem of lack of direction) or because of other personal limitations (Merchant and Van der Stede, 2007). To reduce these (agency) problems, firms can install monitor systems, like management control systems, which can provide incentives to agents to act in such a way that it is beneficial for the firm (Zimmerman, 2006). Hence, management control systems are important for organizations.

Many studies of management control systems have long been colored by contingency theory (Chenhall, 2003). Contingency theory argues that there is no universally applicable management control system for all organizations (Otley, 1980), but that the appropriateness of management control structures is influenced by circumstantial factors (Speklé, 2001), like the (external) environment, technology, organizational structure, size, strategy and national culture (Chenhall, 2003; Robbins and Barnwell, 2006). Fisher (1998) argues that a better fit between an organization’s control systems, results in better performance for both individuals and organizations. In this sense the one control system can be seen as a contingency factor for an other control system.

(9)

8

For studying the management control system package, the first necessity is to define the concepts of (management) control and management control systems (2.1) and management control systems package (2.2). Thereafter a framework of management control categorization to apply in this thesis will be determined (2.3). Subsequently, coupling theory (2.4) and the linkage between coupling theory and the management control systems package will be introduced. Finally, attention will be paid to the public context (2.5)

2.1 Control

It is widely stated that there is no universally accepted definition of the term ‘control’ (Merchant and Van der Stede, 2007; Speklé 2001; Fisher, 1998, Ouchi, 1979), through which multiple (sometimes overlapping) conceptions of management control are used (Malmi and Brown, 2008). For example, Flamholtz et al. (1985, p 35) define control as: “… attempts by the organization to increase the probability that individuals will behave in ways that will lead to the attainment of organizational objectives.” A more specific definition is given by Bhimani et al. (2008, p. 615), which state that: “ A management control system is a means of gathering and using information to aid and coordinate the process of making planning and control decisions throughout the organization and to guide employee behavior. The goal of the system is to improve the collective decisions within an organization”. According to Merchant and Simons (1986) (referred to in Speklé, 2001) control definitions include often two elements: they focus on the behavior of organization participants and the effects of these behaviors on organizational outcomes. According to Speklé (2001) these elements capture a quite significant part of the management control literature.

The concept can be sophisticated by making a distinction between the two purposes of internal accounting systems: decision making and control (Fama and Jensen, 1983; Zimmerman, 2006; Malmi and Brown 2008). Many empirical studies emphasize the decision making purpose, at the cost of a control perspective (see Zimmerman, 2001 for an elaboration). The main difference between the two is that a decision-support system does not contain a mechanism to monitor subordinates (Malmi and Brown, 2008).

(10)

9

example about environmental uncertainty. From a control perspective a given level of environment uncertainty would lead to a less extensive use of accounting information/controls than from a decision-support perspective. The latter stimulates the extensive use of accounting information. Malmi and Brown (2008) align their definition of management control with the definition of Merchant and Van der Stede (which is used in other studies as well, e.g. Nieminen and Lehtonen, 2008), but extended the definition by excluding decision-support systems explicitly. In this thesis I will use this definition

“Management controls include all the devices and systems managers use to ensure that the behaviours and decisions of their employees are consistent with the organization’s objectives and strategies, but exclude pure decision-support system.” (Malmi and Brown, 2008, pp. 290)

Malmi and Brown (2008) call these controls ‘management control systems’ when these controls are complete systems, contrary to a simple rule, like ‘not to twitter during meetings’.

The term ‘organizational control’ is not used, because this term also refers to non-human control issues like for example inventory control (Malmi and Brown, 2008). Merchant (1982) notes, however, that management control is all about human behavior

2.2 The management control systems package The term ‘package’2

is also not pronounced in one voice (Malmi and Brown, 2008). Otley (1980) is one of first who refers to the concept of package (Sandt, 2009). He states that there are extensive interdependencies between the accounting information systems design, the management information systems design, the organizational design and other control mechanisms (like personnel selection) which form together a package.

An other definition sometimes used, is given by Abernethy and Chua ( 1996, p. 573) who argue that “control systems operate as a ‘package’, when they are internally consistent – that is, they are designed to achieve similar ends”. Sandelin (2008), who also studies a management control systems package gives another interpretation to this ‘internal consistency’. Sandelin demonstrates that internal consistency can be built by prioritizing a

2

(11)

10

certain form of control (a primary mode of control) which shape the design of the control package and the secondary modes of control that complement the primary one.

Some authors do not even give a clear definition of the concept. Dambrin et al. (2007), for example, does not make a clear distinction between the term ‘management control system’ and the term ‘control mix’, which they use as a synonym for ‘package’.

Malmi and Brown, on the contrary, do make a clear distinction between the two terms. Their definitions will be used in this thesis. Malmi and Brown (2008) call a complete system of management controls a management control system3, when it is designed and coordinated intentionally. As this is not always the case and because different systems are often introduced by different people/ business units at different times in most organizations, Malmi and Brown do not regard this as a single management control system, but refer to the concept of ‘package’. The term package raises the questions whether and how the management control systems in the package are related (Sandt, 2009).

2.3 Management control categorization frameworks

Multiple frameworks about management control systems can be used for studying the design of management control systems as a package (Sandelin, 2008). Categorizations have, for example, been made on the characteristics of organization activities (Speklé), the use of the system (Simons), and on the object or focus of control (Ouchi, Merchant and Van der Stede). A rather new framework is made by Malmi and Brown (2008).

Simons (1995) distinguishes four levers of control: diagnostic control-, beliefs-, boundary- and interactive control systems. This framework can especially be used for studying how managers use a management control system, rather than studying the design of the system (Sandelin, 2008). Moreover, the framework strongly focuses on top-management level and does not take more informal controls in consideration, through which it is unlikely that the framework can explain the whole control system adequately (Ferreira and Otley, 2009).

(12)

11

Speklé (2001, 2002) distinguishes control archetypes to explain the diversity in management control structures with help of transaction cost theory (TCE). Studies about control archetypes provide an exhaustive taxonomy of ideal controls (Caglio and Ditillo, 2008). Speklé (2002) himself notes that his framework is still in its infancy, that variables have to be defined in conceptual terms and that it has not been studied properly. Moreover, studies about archetypes have mainly focused on simple archetypes of control and do not fully explain more complex situations (Caglio and Ditillo, 2008).

A rather ‘young’ framework is made by Malmi and Brown (2008), who suggest a typology consisting of five groups of controls (planning controls, cybernetic controls, reward and compensation controls, administrative controls and cultural controls). Although this framework sounds rather promising, it does not have proved itself yet.

Ouchi’s framework is based on the focus of control (Kamminga and Van der Meer-Kooistra, 2006). Ouchi (1977, 1979) derives three types of controls based on the two dimensions ‘knowledge of the transformation process’ and the ‘ability to measure outputs’. Control can be focused on activities (behavioral control), on outputs (output control) or on the organizational culture (clan control). Merchant and Van der Stede (2007) elaborate this framework and distinguish four types of objects-of-controls (results-, action-, personnel-, and cultural control). The attractiveness to use this framework will be discussed after the four types are amplified.

(13)

12

In contradiction with results control, action controls directly influence employees’ behavior. Action controls are about employees (not) performing certain actions, through which the organization can obtain its interests and are therefore only useful when a) organizations are able to determine which actions are (un)desirable and b) when organizations are able to observe that the (un)desirable actions are (not) taken. Action controls can manifest themselves in four basic forms: behavioral constraints (physically or administratively), pre-action reviews, action accountability, and redundancy. Except for action accountability, the basic forms are inherently preventing.

Both action and results control are usually not perfect. To overcome this problem, organizations make use of personnel and cultural control mechanisms, which Ouchi combines as ‘clan control’. Personnel controls focus on the types of people and about that employees control and motivate themselves. Personnel controls include 1) selection and placement of employees, 2) training, and 3) job design and the provision of necessary resources to employees in order to do a good job.

Cultural controls is about mutual monitoring, which means that a group exhibit pressure on individual employees who deviate from group’s shared norms and values, traditions, ideologies and ways of behaving. This form of control is most effective when there are emotional ties between employees. Organizational culture is a quite stable control mechanism over time and manifests itself in written and unwritten documents that direct employees’ behavior. Culture can be shaped by codes of conduct (like mission statements or management philosophy), group rewards, physical and social arrangements, intra-organizational transfers. One of the most important advantages of cultural and personnel controls is, that it is applicable to almost every setting in contradiction to action and results control. The degree of effectiveness though, depends on the specific context. The strength of social ties for example, can differ a lot between organizations.

(14)

13

objects of control which makes it relatively easily applicable to different organizational contexts and that it incorporates both formal and informal control mechanisms (Phister, 2009). These features make it attractive to use this framework in this thesis. A disadvantage of this framework is however, that it lacks in specifying the couplings between the control mechanisms (Phister, 2009 and Sandelin, 2008). The object-of-control framework will therefore only be used as a medium for analyzing the couplings between different control systems (Sandelin, 2008, Sandt) and will be complemented by coupling theory.

2.4 Coupling theory and the management control systems package

Almost 40 years ago, Glassman (1973) noted that the coupling concept is a useful theory for understanding the stability of different kinds of living systems, like organizations and society systems. Coupling theory is a tool that focuses on relationships between coupling elements (Beekun and Glick, 2001). Coupling elements can refer to ‘anythings’ (actors, management control systems, subunits etc.) that are related in some way with each other (Weick, 1976; Beekun and Glick, 2001). Weick (1976) notes that an important advantage of coupling theory is, that the theory assumes that an organization consists of multiple building blocks. This decomposability makes it possible to analyze (organization) complexity (Simon, 1996). A difficulty is however, distinguishing a system that is still dividable from a system which can be taken as elementary (Simon, 1996).

(15)

14 2.4.1 Orton and Weick’s framework

There are at least two main perspectives about what determines the type or degree of coupling. Glassman (1973, p. 84) argues that the degree of coupling between systems ‘depends on the activity of the variables which they share’. Orton and Weick (1990) argue that the type of coupling depends on the degree of distinctiveness and responsiveness (see below).

Especially the specific term ‘loosely coupling’ has had a lot of attention in literature. Derived from the above definitions, there are at least two main definitions of loose coupling as well. Glassman (1973) claims that loose coupling means that systems share few variables or that the shared variables are very weak, compared to other variables that influence the system. In this thesis I will apply the second perspective which is of Weick (1976) and Orton and Weick (1990). Weick (1976) states that loose coupling means that ‘coupled events are responsive, but that each event also preserves its own identity and some evidence of its physical or logical separateness’ (Weick, 1976, p. 3). The term ‘coupled’ captures the responsiveness dimension and ‘loosely’ captures the distinctiveness dimension which refers to the preservation of coupling element’s own identity and separateness (Orton and Weick, 1990). The two dimensions give rise to three other forms of coupling, namely non-coupling, tightly coupling and decoupling (Orton and Weick, 1990). A way to present such a dialectical concept is with help of a two-variable matrix (Orton and Weick, 1990; Bobko, 1985) which is presented in figure 1.

(16)

15

When there is both no responsiveness or distinctiveness in a system than it is a non-coupled system, which means that there is not really a system; when there is no responsiveness, but there is distinctiveness the system is decoupled; when there is responsiveness without distinctiveness the system is tightly coupled; and finally when there is both responsiveness and distinctiveness the system is loosely coupled as described above.

The reader should be aware that a tight coupling between coupling elements according to Orton and Weick differs from Merchant and Van der Stede’s (2007) view on a tight management control system. Merchant and Van der Stede, (2007, p. 128) define the degree of tightness as the ‘degree of assurance that employees will behave as the organization wishes’.

2.4.2 Beekun and Glick’s subdimensions

(17)

16 2.4.3 Brown, Malmi and Booth’s subdimensions

As opposed to Beekun and Glick, Brown et al (2007, referred to in Sandt4, 2009) did take the dialectical interpretation into account and developed the subdimensions of Beekun and Glick further. They specify three subdimension for both the responsiveness and the distinctiveness dimension. For ‘responsiveness’ these are: ‘dependence’, ‘directness’ and ‘strength’. Dependence refers to the extent that there is exchange between coupling elements. This depends on the relative magnitude of the exchange and the availability of possible substitutes for the exchange. Directness refers to the extent how directly the coupling elements can ‘communicate’. Strength refers to the degree frequency, intensity and the probability that a change in one element cause change in other elements. When coupling elements score high on these subdimensions, the score on the responsiveness dimension is high.

The three subdimensions of ‘distinctiveness’ are: ‘focus’, ‘use’ and ‘components’. Focus refers to the different foci coupling elements can have. In the language of management control systems, personnel controls for example focus on the type of people and about that employees control and motivate themselves, whereas result controls focus on the results produced. Use refers to whether elements are used ex-ante or ex-post and to the differences in nature of activities which can make them distinct. The difference in components is the last subdimension of distinctiveness and refers to the different the different tools and techniques used by coupling elements. When coupling elements do not much differ on these distinctiveness subdimensions, the score on the distinctiveness scale is low (Sandt, 2009).

2.5 The public context

Public organizations differ from private organizations (Rainey et al., 1976). The main raison d’être of public organizations is to provide public services (Merchant and Van der Stede, 2007). Public organizations are often large and have multiple stakeholders who have an interest in the organization, its goals and its performance. These stakeholders, however, often have opposite interests which makes it difficult to set clear goals and to assess how well the management control package is performing. And even if there are clear goals, it can still be difficult to summarize them into quantifiable performance indicators as opposed to for-profit

4

(18)

17

organizations which have for example profit measures (Merchant and Van der Stede, 2007)5. One of the consequences is that the individual performance of managers and employees is also difficult to measure. Moreover, it is difficult to motivate certain behavior by financial incentives as a consequence of law, contract or the public opinion in non-profit organizations (Merchant and Van der Stede, 2007).

Although non-profit organizations differ from for-profit organizations, they both need a good control system to reduce agency problems and they have both the same basic control systems (results, action, personnel and cultural), which must be aligned, at their disposal (Merchant and Van der Stede, 2007). According to Järvenpää and Länsiluoto (2011) especially public organization have difficulty in managing the management control systems package, because of the divided ownership structure as described above. This complexity makes it interesting to study the management control systems package in the public sector.

5

(19)

18

3. Research method

Empirical research in studying the control package in a (non-)profit organization is currently scarce, through which there is still no clarity about why and how parts of a control package are used and emphasized in practice (Malmi and Brown, 2005 ). In the following, the research method for the empirical part of this thesis will be described. As already mentioned in the introduction of this thesis, the empirical research will be about water boards which will be introduced in section 4.

The theoretical framework described in section 2 will be applied to a practical context by using a case study. According to Yin (2009, p. 18) “a case study is an empirical inquiry that investigates a contemporary phenomenon in depth and within its real-life context, especially when the boundaries between phenomenon and context are not clearly evident.” This research method for studying the management control systems package seems suitable. A first argument is put forward by Ahrens and Dent (1998, p. 33), who state: “To obtain better understandings of how management accounting functions in practice, field studies that bring the messy world of organizations closer to the reader are needed.” J.G. March (1973, referred to in Weick 1976) notes secondly that studying coupling theory, requires a methodology that highlights the rich detail of context. Finally, because the research concerning management control systems package is still in an explorative phase, the case study method is preferred (Sandelin, 2008). Case studies are a useful research method in developing theory, because they are especially appropriate for the ‘why?’ and ‘how?’ questions (Blumberg, 2008).

Within case research, a distinction can be made between single and multiple case studies (Blumberg et al., 2008). According to Blumberg et al. (2008) investigating a case in multiple contexts is usually better, than basing results on just one case. Nevertheless, this thesis incorporates a single case study, because it has been advised to examine a single organization in studying the complex topic of the (management) control (systems) package (Otley, 1999).

(20)

19

those management control systems (Tessier and Otley, 2012). This difference is acknowledged in this case study by not only interviewing the (in)direct manager(s), but also by interviewing two subordinates. Information about the interviews and the interviewees is listed in table 2. The total interview time was about nine hours, which resulted in sixty-six transcription pages6 (single spacing, font size 12). These reports were in Dutch and processed in spoken language. The individual reports were sent to the interviewees. After processing of their comments and their consent, the transcriptions were used for this thesis.

Two steps had to be taken in the research process. First, there must be a good understanding of which management control systems are in place. The second step is studying how they are coupled and why they are coupled the way they are. To get an idea of the management control system package, a research tool of Ferreira and Otley (2009), the framework of Merchant and van der Stede (2007) and the researcher’s own insights were used. The disadvantage of Merchant and van der Stede’s framework is that they do not always provide clear guidelines for case study research, while Ferreira and Otley do provide a list of possible interview questions. The (sub)dimensions of the Orton and Weick’s framework were more difficult to study, because there are no real operational definitions of the (sub)dimensions for case study research as far as this author know. Often this appeared from the context. It is also very difficult to formulate questions about the coupling between management control systems upfront, because beforehand it is not clear which management control systems exactly are in place and which are logically coupled in the specific context. This appeared during the interviews, through which the questions which are related to the coupling between management control systems emerged particularly during the interviews or the information was already told by the interviewees themselves. The interviewees were all inclined to tell much information on their own accord. As the interviewees have different positions and the objective of each interview was different, the topics discussed were also different. Moreover, the information provided in the interviews was used for the next interview.

In consultation with the secretary-director of the water board, one department of the organization was chosen as the focus of analysis. He also gave permission to interview certain employees from that department. The chosen department ‘Safety and Sufficient Water’ does

6

(21)

20

not only employ the most people, but also executes one of the main objectives of water boards, namely providing safety and the right quantity of water to the citizens of the area.

In the next section, water boards in general will be first introduced, because they are special public organizations and differ from other public organizations. Next the specific case study water board ‘Hunze and Aa’s’, will be introduced.

Table 1: List of secondary data sources (excluding articles in academic journals)

Document Written by Subject

Begroting 2012 (budget) Water board Hunze and Aa’s, 2012

Budget of 2012

Beheerplan 2010-2015 (management

plan)

Water board Hunze and Aa’s, 2010

Vision, mission,

developments, actions and objectives that water board Hunze and Aa’s want to achieve for coming period EU Water Framework Directive European Parliament and

Council, 2000 Informatieblad Beoordelen en Belonen Stichting Arbeidsmarkt- en Ontwikkelingsfonds Waterschappen, 2008

Providing information about the performance-contingent payment

Multiple progress and planning overviews

Water board Hunze and Aa’s Progress and planning overviews of the different tasks performed by the operation managers Waterschapswet, 1991 Tweede Kamer der Staten

Generaal, 1991

Legislation relating to water boards

Waterwet, 2009 Tweede Kamer der Staten Generaal, 2009

Legislation about the

management and use of water systems in the Netherlands www.drentsarchiefnet.nl Drents Archief Provides information about

the inventory of preserved archives which are related to Drenthe, about contact addresses etc.

www.hunzeenaas.nl Water board Hunze and Aa’s The website contains information about the organization, rules citizens have to comply with etc. The site changed dramatically on 16 October. The site is used before and after this date. www.uvw.nl Unie vanWaterschappen Provides a platform about all

(22)

21 Table 2: List of interviews done (primary data sources)

Date Length

(m)

Location Position Code Educational Background Objective

05/10/2012 110 Veendam (head office)

Secretary-director since 2007 SD Cultural engineering (bc) and Business (MSc)

Gain insight in the Hunze and Aa’s organization

17/10/2012 110 Veendam (head office)

Department manager of SSW since 2011

DM Civil engineering and Applied Business (bc) and is currently studying Business Development and Innovation and Public Strategy (MSc).

Gain insight into the department of SSW, which management control systems they use and how these are related

13/11/2012 100 Siddenburen (rayon North office)

Operation manager of rayon North since 2011

OMRN Mechanical engineering (bc) Gain insight into what the operation manager does, how he is controlled by the department manager and how the relations are between those management control systems. There was particular attention paid to the planning and progress overviews. 16/11/2012 90 Veendam

(head office)

Member of the executive board (responsible for finance and general affairs) and the general board, and loco dikereef at Hunze and Aa’s since the founding of Hunze and Aa’s in 2000, member of diverse governing bodies at LTO and Cosun, and is business owner of a farm

MEB Agricultural engineering (MSc) and extra courses in subjects as

accounting and informatics

Gain insight into Hunze and Aa’s from a board perspective, how the organization has developed the last twelve years and insight how the

organization performs nowadays.

21/11/2012 120 Veele (rayon East office)

Operation manager of rayon East since the founding of Hunze and Aa’s in 2000

(23)

22

4. Case study: Water board Hunze en Aa’s

Section 4.1 introduces water boards in general. Subsequently Hunze and Aa’s will be introduced (4.2) and the changes that the organization went through since the establishment (4.3). Thereafter a description and analysis of the management control systems package will be given (4.4). Finally the findings of the analysis will be summarized and the effects of the management control systems package will be discussed (4.5).

4.1 Introduction of water boards in general

For centuries the inhabitants of the Netherlands are in ‘war’ with water. This is because a large part of the Netherlands is below sea level, and because the Netherlands is locked in between the North Sea and the major rivers that flow into the country from other European countries. The control and management of water is organized locally for ages. In the 12th century, local and loosely organized organizations were developed to manage the water level. In the 13th century the present-day ‘water boards’, were founded as the first democratic authorities in the Netherlands on an ad-hoc basis. In the 19th century thousands of local water boards were coordinated and reorganized. Many mergers between water boards took place and eventually the number of water boards was reduced from 3500 in 1850 to 2500 in 1950 (Van Steen and Pellenbarg, 2004). Nowadays there are 25 water boards employing about 11.000 employees, in the Netherlands (www.uvw.nl).

The present-day water boards are regional public agencies which are endowed with the responsibility for flood defense and the quantity and the quality of (regional) surface water in a certain area in the Netherlands (www.waterschappen.nl ; Waterschapswet, 1991). The area of a water board is determined by a “catchment7” and not by the administrative boundaries of municipals or provinces (De Bosatlas van Nederland Waterland, 2010). The resulting regions of such a division are called ‘functional regions’ (Woltjer and Al, 2007) and are perceived as organizationally efficient and effective ( Raad voor het Openbaar Bestuur, 2003).

There are however physic geographical and social developments that threatens the water management. Physic geographical (future) threats include rising levels, a drop in the level of

7

(24)

23

land and changing seasonal precipitation levels (Van Steen and Pellenbarg, 2004). Social threats are an increasing level of urbanization, an increasing shortage of space (www.waterschappen.nl) and high aspiration levels with respect to the quality of water imposed by society itself, must be satisfied (Van Steen and Pellenbarg, 2004). According to Van Steen and Pellenbarg (2004) the major challenge in the next decades will not be a technical one, but will be a financial one. Large sums of money are needed to deal with the issues described above (Van Steen and Pellenbarg, 2004). However, in politics there is pressure to reduce costs of water management in, among others, the Dutch water boards (see for example the COELO report, 2010). Hence, more than ever it is of great importance that water boards have an effective management control systems package.

As water boards are public organizations, much of their structure, authority, tasks etc. is (are) defined in legislation. Water boards are not only affected by national legislation, but also by international policy and legislation (like the ‘Water Framework Directive8_{’), because the} management of water is not bound to national borders (Woltjer and Al, 2007).

Water boards have largely the same characteristics (see section 2) as other public organizations, but water boards also differ. Water boards have a more concrete task than for example municipals and provinces. Also their legal structure differs. Municipals and provinces, for example, have a dualistic structure, which means that members of the executive board are not part of the general board. Water boards have still a monistic structure, through which members of the executive board are also member of the general board. Water boards levy taxes and therefore have a (more or less) self-supporting financial system (Van Steen and Pellenbarg, 2004).

The empirical part of this thesis considers a water board and focusses on the following empirical research questions:

 How are the management control systems of the case water board in a management control systems package coupled?

 And why are they coupled the way they are?

8

(25)

24 4.2 Introduction of Hunze and Aa’s

A single case study is conducted in one of the 25 water boards, namely “Hunze en Aa’s”. Water board Hunze and Aa’s is founded in 2000 and emerged from a merger between three water boards and two provincial purification organizations (www.drentsarchiefnet.nl). Hunze and Aa’s is responsible for the regional water management in the east of Groningen and the northeast of Drenthe, two provinces in the Netherlands, and covers an area of 213 000 hectare (www.hunzeenaas.nl). Hunze and Aa’s has a yearly budget of about €80 million and employs about 300 fte in 2012 (Begroting 2012).

The structure of water boards is bipartite and includes the water board management and the civil organization. The water board management includes a general board, an executive board and a chairman which is called a dike reef (Waterschapswet, 1991). The civil organization is managed by a so called secretary-director (SD). This ‘double-function-title’ refers to the fact that this person is secretary for the general- and executive board and is director of the civil organization. The civil organization will be central in this case study.

Figure 2 depicts the current structure of the civil organization. On top is the secretary-director which forms together with the department managers (DM), the management team. There are six departments: Safety and Sufficient water (SSW), Clean Water (CW), Policy, Projects & GEO-information (PPG), Laboratory (Lab), Finance and General Affairs (FGA) and HRM, Facilitating, Communication & ICT (HFCI). The first three departments are the ‘primary’ departments. SSW and CW are more occupied with the execution and PPG is more occupied with policy; the other three departments have a more supportive nature. This thesis focusses especially on the department ‘Safety and Sufficient Water’. The objective of the SSW department is to maintain the waterways, the see dyke, the boezems9 and the artifacts like pumping-engines, barrages and sluices, through which the right quantity of water can be provided and safety can be guaranteed. Mechanic engineers of CW perform the technical maintenance of the artifacts. The department employs about 120 employees. Each department manager controls individual employees. Because this amount of people in SSW is too difficult for one person to manage, the department is subdivided in teams. These teams are: rayon North, rayon East rayon West, Support & Sea dyke and Mura (which is an abbreviation of ‘Musk rats’ control). The first three are

9

(26)

25

Secretary-Director (SD)

Department manager (DM) Safety and Sufficient

Water (SSF)

Operation manager rayon North (OMRN)

Operation manager rayon East (OMRE)

Operation manager rayon West

Operation manager Support &Sea Dyke

Operation manager Mura Department manager Clean Water (CW) Operation manager rayon North Operation manager rayon South Department manager Policy, Projects & Geo-Information ( PPG) Operation manager Projects Opeation manager GEO-Information Department manager Laboratory (Lab) Department manager Finance and General

Affairs (FGA)

Department manager HRM, Facilitating, Commynication & ICT

(HFCI)

Operation manager ICT Staff

(27)

26

geographical determined and are responsible for the maintenance in their area, and the last two are determined by their main task(s). Each team is managed by an operation manager. Each operation manager manages a ‘club’ of about 15 to 20 employees, which includes for example tractor drivers. The rayon’s operation managers operate mainly from their rayon offices and are only a small part of their time present at the head office in Veendam. The subject of interest is especially how the department manager of SSW controls his employees. These include the different operation managers, but also other employees who are not depicted in the organization chart. This last category of employees includes hydrologists, area planners, employees who make decisions about the water level and a safety expert. The department manager controls about 15 employees directly.

A distinction can be made in the type of work that the employees, directed by the department manager of SWW, execute. The rayon operation managers carry out mainly the regular work and the other employees like hydrologists carry out the projects. Because two rayon operation managers have been interviewed, the focus will be mainly on the control of the rayon operation managers who perform the regular work. The biggest task of this regular work of which the operation manager is responsible for, is the mowing10 of waterways’ overgrowth. Other (‘smaller’) tasks that are performed by the rayon teams are for example dredging, pruning and doing talus repairs. Mowing takes place in spring, summer and autumn. The other tasks are mainly performed during winter and spring time. To illustrate certain issues examples of other employees or the rest of the organization are also used.

Before the management control systems package of water board Hunze and Aa’s will be discussed, a short review is given about changes that the organization went through since 2000 and which has influenced the management control systems package.

4.3 Organizational changes in the period 2000-2012

Since the founding of Hunze and Aa’s in 2000, the organization has passed some major changes. In the early days the organization mainly focused on providing quality and there was less focus on costs. The secretary-director tells that about 7 á 8 years ago, it happened that tractor drivers mowed certain water ways in a too high frequency. They started in May and

10

(28)

27

they stopped in November. No information was written down though, through which the organization had not a very good idea of what the tractor drivers where doing. He tells that: “as long there were no complaints about the water level, it was good. No complaints will

arise when the water can go through.”

As a consequence of internal insights and external pressure from society about among others costs, another perspective came into place since 2005. The organization became more flat and much more cost-conscious. The price/quality ratio became the centre of gravity. According to the SD it was effectiveness that had to be changed and not efficiency. He states: “A more

clever way of doings things for things that are unnecessary, is not clever.” The new spearhead

became that the right things should happen at the right moment, which was expressed to the operation managers and tractor drivers by the SD in statements like: “I prefer that you go to

the canteen, when it is not really necessary to mow; Coffee is much cheaper than diesel fuel.”

The cost-consciousness appeared in various management control systems like the organization’s mission, meetings and as precondition in operational plans. To know how low the cost of mowing should be, Hunze and Aa’s use(d) internal benchmarking. A part of the mowing work is put out to a contractor and that cost level is used as a benchmark for Hunze and Aa’s. A few years ago, it appeared that the contractor had cost of €1,40 per linear meter. The mowing by Hunze and Aa’s itself had costs of €2.10 per linear meter. As result of this information, the SD told the operation managers and their employees, that mowing could only be done by the water board itself when it could compete with the contractor. Subsequently, the organization drew up a maintenance management plan11 in 2005 in which it is stated how the maintenance of each waterway should be and how important the waterway is for guaranteeing a good water level. These considerations result in operational plans. In the forthcoming paragraph the coupling between the maintenance management plan and the operation plans, will, among others, be discussed.

4.4 Description and analysis of the management control systems package

From the secondary data and the interviews it followed that Hunze and Aa’s has multiple management control systems, among which a management plan, different operational and

11

(29)

28

investment plans, progress overviews, production cards, budgets and an incidental and structural performance-contingent rewarding system (results control systems), a maintenance management plan (action control system) a mission, vision, core principles, improvement boards, a building which incorporates all kinds of durable features and meetings (cultural control systems), personnel selection, assessment interviews and training (personnel control system).

It seems that the department SSW, and especially the rayon operation managers, relies currently relatively most on result control systems compared to the other departments. According to the SD the department (at least the rayon units), is especially controlled by results/norms because most tasks are not too complicated, through which the mowing task is easy to quantify. He tells that a more difficult issue to quantify is for example the communication objective: citizens must become more aware of the water board Hunze and Aa’s. Both SD and the DM tell that employees have rather much freedom in performing their job.

During the interviews it seemed that most of the management control systems were (to some extent) related. This case study will discuss the management control systems and the couplings between them that were most salient and perceptible for an outsider during the interviews. This means that some of these interconnections will be extensively discussed with use of the Orton and Weick framework, some will only be used to outline the context and some will not be mentioned at all. Space restrictions regarding a master’s thesis do not allow discussing the whole management control systems package in detail.

(30)

29

4.4.1 Coupling between maintenance management plan and operational plans12

The maintenance standards are stated in the maintenance management plan. These standards are subsequently translated in operational plans which state how much kilometers the different rayon teams must mow each week. The rayon’s operations managers tell that they are responsible for the operational plans, but that these plans are made with help of their team members and a supporting employee at the head office. In these operational plans also the progress is depicted through which they are simultaneously also progress overviews. As contexts and insights change, each year in winter the operational plans for the coming year are made or adjusted.

The maintenance management plan and the operational plans have different foci in terms of the coupling perspective described in section 2.4.3. The maintenance management plan focusses on actions (how do we have to maintain the various water ways?)13. The operational plans focus on output/results to make sure that the water ways are clean. They do have overlapping components. The maintenance plan is about the categorizations of the importance level of the water ways, while the mow operation plans of the rayon managers translate these categorizations into weekly plans. The maintenance plan, though, incorporate portrayed images and the operational plans incorporate graphics to display the planning per week or period. The use of the maintenance plan is ex-ante. The operational plans are in principle also ex-ante. The document on which it is depicted is however also used ex-post, because the planning is compared with realizations. When the operational plans are used as comparing or progress device, the plans will be called ‘progress overviews’. It seems that the maintenance plan and the operational plans have a low to medium score on the distinctiveness dimension of Orton and Weick.

The responsiveness dimension on the contrary seems rather high. The operational plans depend on the maintenance plan. The DM enforces this by telling that when adjustments are made in the basis, there are consequences for the planning. Both maintenance plan and the operational plans do not in principle change during the year. The rayon North’s operation manager (OMRN) tells, however, that last year he adjusted his planning, because the planning was not representative on the long run anymore. He adjusted the planning to reflect the actual situation. So the operational plans depend also a little on the practical realization.

12

In this subparagraph ‘operation plans’ refer to the plans of rayon operation managers

13

(31)

30

The directness and strength dimension seems also high. The rayon East’s manager (OMRE) tells that the rayon team gains experience when it performs its work. The rayon teams together with the employees who decide on the water level determine if the maintenance plan should be adjusted. Herein, they take into account whether it is a seldom phenomenon or that it should be a structural adjustment. According to the OMRE a part of the work is uncontrollable, because of the weather conditions. He tells that when structural changes are necessary, the next year in January/February the maintenance plan and the operational plans will change. The rayon’s operation managers are subsequently responsible for the operational plans.

A quite high responsiveness dimension combined with a quite low to medium distinctiveness dimension, assumes a tightly coupling between the maintenance plan and operation plans according to Orton and Weick’s framework. This tight coupling seems logical. The DM explains that the right things should happen at the right moment and that effective plans are needed. The maintenance plan is made by the employees and is approved by the board. The DM tells that possible problems in the planning should be solved by the rayon operations managers and as long as there are no real big problems, this will be done next winter to avoid costs. This seems not a problem. According to the OMRE those changes are marginal and that the outline is right.

4.4.2 Production cards and progress overviews

The rayon operation managers receive every week a progress overview to which extent the planning is realized. These overviews are made at the head office and are based on input of the tractor drivers. The tractor drivers write on a production card, how much kilometers and which waterways they mowed each week. The progress overviews are a management control system for the rayon operation managers. The production cards are rather a management control system for the tractor drivers and a decision-support system for the rayon operation managers. The production cards are, however, important for the progress overviews and therefore interesting to study with respect to the management control systems package.

(32)

31

both management control systems show how much there is mowed in week X and in which profile14, but the progress overviews also depicts this into cumulative graphics with a distinction in the profile and graphics which are subdivided in the different labels. Moreover, depending on which rayon is the focus of attention, production cards also reflect which water ways exactly are mowed and which activity has taken place. The progress overviews are more like a summary. The use is, on the contrary, different. The production cards are used after actions have been performed, while the progress overviews are used ex-ante to direct the team the next week. When the progress overviews are used to compare planning and realization, it is used ex-post though. All these parameters together give the impression that the distinctiveness dimension is rather low.

This is different for the responsiveness dimension. The progress overviews of mowing depend totally on the production cards. A change in the production cards leads automatically to a change in the progress overview. Next, the production cards ‘communicate’ rather directly with the progress overviews, although it is processed by a supporting employee on the head office. This intermediate ‘station’ results in a delay of a few days between the handing in of the production cards on Fridays and the completion of the progress overviews and the sending to the rayon operation managers and the department manager on Wednesdays. Finally, the frequency of adjusting the progress overview is weekly, which indicates a medium to high score on the ‘strength’ subdimension. It seems that the total ‘grade’ on the responsiveness dimension is rather high. According to Orton and Weick’s framework, the progress overviews are tightly coupled to the production cards.

This tightly coupling seems logic. As the DM tells, the aim of the progress overviews is to unify planning and realizations. According to him the benefit of this tight coupling is that the rayon operation managers have a good idea of how much of the planning is realized, through which they are able to make corrections immediately. The OMRE agrees with this. The OMRE mentions that he receives the progress overviews on Wednesdays, but has a team meeting to discuss the state of affairs on Mondays. He explains that the processing of the production card on meters level by the supporting employee takes three days and even if the information is processed during the weekend, it still cannot be delivered on time. The lack of

14

(33)

32

information is not a problem for the OMRE. For him it is not necessary to have an at the meter exact progress overview on Monday, he only needs a rough estimate of the progress. With help of the production cards he makes a fast calculation himself and sketches it into last week’s progress overview.

4.4.3 The management plan15 and the operational plans16

A water board is obliged to have a management plan regarding the water management in the water board’s area, which is revised each six years (Waterwet, 2009). The current management plan is related to the period 2010-2015. In this document a water board states the main lines of what it is intended to do in relation to policy, management and maintenance in the water board’s area in the forthcoming years. The objective and actions in the management plan are partly determined by external forces, but are also partly determined by the civil organization itself. The document also presents the water board’s general mission, core principles under which it operates, ambitions and vision, trends and developments and the legal requirements which it has to fulfill. This document is used to direct the organization and has from that perspective, next to the external function, also an internal control function, (Beheerplan 2010-2015).

The operational plans, which include several projects, are derived from the management plan. The management plan and the operational plans have overlapping foci. Both reflect objectives and the actions that would lead to these objectives. They also have overlapping components. The objectives and the needed actions are, however, much more concrete in the operational plans and represented in different forms. The DM tells for example that on behalf of the objectives mentioned in the management plan, a process schedule (an operational plan) that includes the necessary activities is drafted. Both operational plans and the management plan are used ex-ante. Therefore, the distinctiveness dimension with respect to the operational plans and the management plan appeared to be low to medium.

The operational plans are based on the management plan. According to the DM there is a one-to-one relationship which indicates a great strength and dependence. It is, however, questionable if this one-to-one relationship holds when the management plan period matures

15

‘Beheerplan’ in Dutch

16

(34)

33

(the reason will be explained below). The management plan is translated in the department plans and subsequently in personal plans, which makes the relationship rather direct. It seems that the score on the responsiveness dimension is high. When this is combined with a low to medium distinctiveness dimension, Orton and Weick’s framework predict a tight coupling between the management control systems. The DM explains that there is a one-to-one-relationship between the management plan and the operation plans to avoid ‘luchtfietserij’17

The coupling between the realization of the operational plans in the organization and the management plan is at the same time also a good example of a non-coupling. During the six-year period, it appears that some objectives or actions stated in the management plan will not be achieved. The SD tells that the water board presents these adjustments in the budget and other budget related reports which are public, and in a document that is given to Brussels each year. The budgets and the other budget related documents show both the plans and the relating cost for the coming year. Changes in the budgets will eventually appear in the operational plans for the next year again. These insights are not adjusted in the management plan though. The distinctiveness between the coupling elements is still the same and therefore low. The management plan on the contrary does not respond to changes in the realizations of operational plans at all. According to Orton and Weick’s framework the operational plans and the management plan are non-coupled from this perspective.

The opinions about this non-coupling differ. The SD tells that on the one hand, the benefit of the management plan is that it forces an organization to look ahead, but on the other hand much can change in a six-year period. The SD therefore prefers a dynamic management plan: “The most ideal situation for me, would be a plan in which you add one year to the plan at the

end of the year, through which the modified insights of the last year can be processed. In this way, the plan is a more actual control instrument. At this moment, this is however only a wish, because this is not (yet) possible under the current legislation…. and would result in an enormous bureaucracy...”. The MEB thinks on the contrary that a dynamic management plan

is not desirable and states: “The management plan is just a direction; a point on the horizon

towards the organization has to work to… It would also be a waste of money to adjust the plan each year. The adjustments become visible in the budgets, and that is the place it should be”. The SD tells that because the adjustments are made in the budgets, there is at least a

17

(35)

34

systematic control’s point of departure. He complements this, however, by stating that at the end of period the management plan is outmoded at a number of points.

Because the operational plans have to respond to obtained insights, which are processed in the budget, it can be concluded that the earlier mentioned tight coupling between the operational plans and the management plan will slightly fade away and become more non-coupled when time passes. This example also shows that, dependent on which coupling element is taken as starting point, the same two coupling elements can be coupled in different ways.

4.4.4 The rewarding system and result control systems

Hunze and Aa’s has in principal two financial performance-contingent rewarding systems18 : an incidental- and a structural rewarding system. The DM states that he tries to split the two rewarding systems as much as possible, but that the boundary is sometimes a little vague. He explains that the one system influence the other system: the people who get an incidental rewarding are always employees who are also qualified for a growth in their structural reward. It is in principle not possible for an employee to qualify for an incidental reward, but not for the structural reward.

One could expect from the Merchant and Van der Stede’s framework that performance-contingent rewarding system is related to the measures used in the result control systems (see section 2), because the department is controlled by results according to the SD. In the case for rewarding the rayon operation managers, these measures could be exploitation budgets or mowed kilometers. The two management control systems could expected to be rather responsive, through which the coupling would be tightly or loosely coupled. This is however not the case, although there is a slight difference between the incidental reward system and the structural reward system.

The incidental rewarding system

A while ago, an incidental performance-contingent reward has been introduced. In agreement with the works council (‘Ondernemingsraad’ in Dutch), it is decided that 20% of the employees should earn an incidental rewarding of 3.5% or 5% each year. The SD tells that

18

A coupling perspective on the management control systems package

A coupling perspective on the management

control systems package

case “water board Hunze and Aa’s”

Master Thesis

J.J.S. Haanstra

University of Groningen

A coupling perspective on the management

control systems package

case “water board Hunze and Aa’s”

Table of contents

Preface

1.

Introduction

2.

Literature review

3.

Research method

4.

Case study: Water board Hunze en Aa’s