The development and effects of control systems in governmental collaborations

MSc BA: Organizational & Management Control

The development and effects of control systems

in governmental collaborations

Rashree Ramdat Tewari Student number: S1744569

Email: rashree.rtewari@gmail.com

Rijksuniversiteit Groningen Faculty of Economics and Business

Msc Business Administration

Specialization Organizational & Management Control

Supervisor: drs. R. A. Minnaar

2

Preface

This thesis is the result of my graduation process of the specialization Organizational & Management Control from the Master Business Administration at the RUG.

It is focused on control in governmental collaborations, which I found an interesting and current topic since the number of governmental collaborations is growing and these collaborations have to deal with many pressures and changes in the environment these days. I would like to thank several people for their support in writing my thesis. First of all, my supervisor Reinald Minnaar for his feedback, and my second supervisor Jeltje van der Meer-Kooistra for co-reading my thesis.

During my research, I held a case study at Regio Groningen-Assen, which is a collaboration between different provinces and municipalities. I would also like to thank all the interviewees that made it possible to collect the data needed for this thesis.

Hopefully you will enjoy reading this thesis,

3

Abstract

This thesis explores how internal and external factors influence the development of control systems in governmental collaborations and how these systems influence the collaboration. The current literature identified four factors that can influence the development of control systems in general; strategy, size, structure, and organizational uncertainty. These factors were also found to influence the control systems in a governmental collaboration, although the effects of these factors sometimes differed from what would be expected from the literature. Other factors that can influence control in governmental collaborations are the recession and budget cuts, politics, societal changes, changes within the collaborating organizations, and whether the collaboration is a legal entity or not. These factors can have several effects on control systems and can, amongst others, create a need to increase the flexibility of these systems and determine the amount and types of control systems that are adopted. Some main effects of the control systems within a governmental collaboration found are an increased transparency for the participants (e.g. regarding their tasks and responsibilities) and a better possibility to adapt to future developments.

4

Table of content

2.3 The effects of control systems ... 9

2.4 Control systems in collaborations ... 10

2.5 The development of control systems ... 11

2.6 Pressures for governmental organizations ... 12

2.7 Changing control systems and environment ... 13

2.8 Concluding ... 14 3 Research methodology ... 15 3.1 Introduction ... 15 3.2 Literature review ... 15 3.3 Interpretative research ... 15 3.4 Case study ... 15

3.5 Data collection method ... 16

3.6 Reliability and validity ... 16

3.7 Data analysis ... 17

3.8 Concluding ... 17

4 Case introduction ... 18

4.1 Introduction ... 18

4.2 Structure of the organization ... 18

5 Case description and analysis ... 20

5.1 Introduction ... 20

5.2 Control systems ... 20

5.3 Development of control systems ... 23

5.4 Changes in control systems ... 25

5.5 Concluding ... 28

6 Case implications ... 28

6.1 Introduction ... 28

6.2 Case implication ... 29

5

7 Conclusion, limitations and suggestions for future research ... 33

7.1 Introduction ... 33

7.2 Answers to the research questions ... 33

7.3 Limitations ... 36

7.4 Suggestions for future research ... 36

8 References ... 37

6

1 Introduction

“An important trend exists towards increased inter-organizational collaborations and information integration among government agencies” (Gill-Garcia, 2012, p. 270). Reason behind this is that governmental organizations are dealing with many problems and challenges these days (e.g. regarding the economic development), that cannot be solved by a single organization (Mcguire, 2006). These collaborations, which refer to inter-organizational strategies, executed by two or more public organizations, regardless of their form (e.g. an agreement, alliance or consortium) are seen as the solution for the problems and challenges public organizations face (Brinkerhoff, 2002; Sorrentino et al., 2010).

Arguments that promote public collaborations are, amongst others, an improvement of efficiency and the ability to comply to future expectations by a reorganization, reuse of assets (e.g. facilities and maintenance contracts) or sharing investments, which can be achieved by cooperating (Weerakkody et al., 2008). Governmental organizations are however concerned with more objectives, such as policy effectiveness, accountability, openness and equity. A collaboration between governmental organizations can contribute to achieving these goals by improving service outcomes through less fragmentation, a better coordination between the collaborating public agencies, and the access gained to important information and expertise (Agranoff and McGuire 2001, Milward and Provan 2003, Gill-Garcia, 2012).

Control systems, which include plans, methods and procedures used by an organization to meet its objectives, are important factors that support organizations in achieving their goals and coping with the changing environments, priorities and demands. The contextual changes make it necessary to regularly assess and evaluate control systems, and verify if these systems are still suitable (Goa, 1994).

Given the growing number of collaborating governmental organizations and the importance of control systems in meeting their objectives and dealing with the changing environments that these collaborations face nowadays, it is interesting to investigate how these control systems develop within a governmental collaboration and how they affect the collaboration.

This has not been examined in the current literature, so a literature gap exists. There are some articles regarding the factors that influence the development of control systems in general. These articles, however, are not focused on (governmental) collaborations. This is an important distinction, since the involvement of different parties with possibly conflicting interests might affect the development of control systems. In addition, public organizations have to deal with many institutional pressures (Gupta et al., 1994), which might also influence the development of control systems. Further, despite the importance of control for organizational performance, it is relatively unexplored how these control systems affect organizations (Kinney, 2000).

7

Two research questions will be addressed in this thesis;

1) How do internal and external factors influence the development of control systems in a governmental collaboration?

2) How do these control systems influence the collaboration?

The remaining part of this thesis will be structured as follows; the next chapter explores the current literature regarding control systems and governmental collaborations. In the third chapter, the research methodology will be elaborated. Following this, the case that will be central in this thesis is introduced in chapter four. Chapter five contains a further description of the case and the results from the case study. These results will be analyzed in chapter six. Finally, there will be a conclusion in chapter seven, including the limitations of this thesis and suggestions for future research.

2 Literature review

2.1 Introduction

In this chapter, literature regarding control systems will be reviewed. Since this thesis aims to investigate how control systems are developed and affected by internal and external factors, and how these systems influence the collaboration, these aspects will be discussed in this chapter. In the following section control systems in general will be discussed and a definition of the concept will be given. Section three continues with the effects of control systems. Since this thesis is focused on collaborating governmental organizations, some information regarding control systems in collaborations and their effects will be given in section four. In order to get some information on factors that might affect control in governmental collaborations, despite the literature gap, section five is focused on factors that influence control systems in general, and therefore might also be interesting for the research questions of this thesis. Section six continues with some additional pressures governmental organizations are subjected to, that could possibly also influence the control in governmental collaborations. Section seven looks at the changes in control systems, and finally, a conclusion is given in section eight.

2.2 Control systems

Control systems can be defined as: “A set of all actions and measures which apply to activities performed in the organization being carried out in accordance with the interests of the organization” (Romaniuk, 2011, p.110). “It comprises the plans, methods, and procedures used to meet missions, goals, and objectives and, in doing so, supports performance-based management” (Gao, 1999, p.4).

8

The first step, establishing standards, will also be considered as part of the control process during this research.

Control systems can further be classified in three categories, based on the object they control. These categories are action control, result control and personnel control Merchant (1982).

Action control aims to encourage employees to take desirable actions and prevent them from taking actions that are not in the best interest of the organization. This can be achieved through behavioral constraints, such as physical devices (e.g. locks) and administrative constraints (e.g. separation of duties). Another sort of action control is administrative accountability that holds employees responsible for their own actions. Three conditions are important for this type of control. For one, employees must know what type of behavior is acceptable. This has to be defined and communicated through rules, policies, and procedure manuals. Second, employees’ behavior has to be measured, and third sanctions or rewards have to be determined for behavior that does or does not meet up to the standards. A third type of action control is pre-action review, which aims to prevent potentially harming actions before they are completed through direct supervision and reviewing.

Result control is focused on the achieved results, and result accountability is used as a control method. Other than its focus on results, the steps of results accountability are comparable to those of action accountability. Desired results (e.g. standards and budgets) have to be defined and communicated throughout the organization, they have to be measured, and sanctions or rewards must be given for certain results.

The third category of control systems noted by Merchant is personnel control, which depends on the personnel to do what is best for the organization, and it assists them in doing so. This type of control is appropriate when there is goal congruence. Methods that are used to further improve performance are upgrading personnel capabilities (e.g. through training and strict selection criteria), communication (e.g. of expectations and information regarding coordination) and peer control (e.g. in the form of work groups) (Merchant,1982).

Emch (1954) has noted that control should not merely be focused on providing measures for planning adequately and evaluating the achieved performance. An important aspect of control should also be to stimulate initiatives of employees and executives. He states that a mistake often made is using control as a dictatorial tool for centralized decision making that leaves no room for input from the lower levels.

9

He therefore proposed that control should be related to the responsibilities and authorities of the different functions, which he calls "functional control". Control should not merely be limited to the tasks of the controller and top management, and be highly centralized. Instead, it should be an integral aspect of every function in the organization to assure that all the different functions will be truly under control. It should further be focused on supporting the operating personnel and contribute to achieving desired outcomes, not just provide historical reports that are of little use in practice (Emch, 1954).

2.3 The effects of control systems

Control systems can improve organizational performance in several ways. They can improve transparency, give directions for improvements to reach specific goals (Haponava & Al-Jibouri, 2012), and support in managing changes to cope with the changing environment, priorities and demands of an organization (Gao,1999).

Control systems can further improve organizational performance by protecting against personal limitations, such as a lack of clarity for employees regarding their tasks and what is expected from them, caused by a shortage of information, training or ability (Merchant, 1982). To perform their job well, it is crucial for employees to clearly understand their function, authority and responsibilities, and possess other information they need to perform their job effectively (Emch, 1954). Appropriate control systems can contribute to this and thereby improve employees’ performance.

Furthermore, controls can improve performance by protecting against the effects of goal incongruence between employees within an organization or between collaborating organizations, and prevent employees from taking undesirable actions (Merchant, 1982).

Another possible effect of control systems is an increase in motivation. Formal control systems can have an incentive effect (e.g. in organizations that use pay for performance systems) that motivates employees to reach specific targets. Moreover, control systems can increase psychological motivation by influencing employees’ ideas regarding appropriate behavior in a certain situation, which leads to forming personal norms (Tayler & Bloomfield, 2010). Confirming to these norms can give employees a good feeling about themselves and their behavior (Schwartz, 1977), and it can strongly influence their behavior, despite the fact that they have no direct economic impact (Cialdini et al., 1991). Control systems can further indirectly influence employees’ motivation through their preference to act in a similar manner as those around them, which can be seen as descriptive norms (Tayler & Bloomfield, 2010). These norms can still exist when the control systems that led to their development are changed (Tayler & Bloomfield, 2010).

10

leave no room for initiatives to correct inappropriate conditions and possibly poor planning, and make responding to arising situations and opportunities more difficult. Applying unsuited control systems can further negatively affect employees’ interests and enthusiasm. It can even lead to a greater focus on complying with procedures, than on trying to reach the objectives of the organization. Too much control can further create a great amount of paperwork, directives, manuals and reports, which are not always read, and are not appropriate when striving for efficiency (Emch, 1954).

To conclude, control systems can have many positive effects on an organization such as increasing motivation, giving direction, and providing clarity regarding roles, objectives, policies and other important information employees need to perform their job well. On the other hand, too much control imposed on employees by the top can have a negative effect on performance, by leaving no room for employees’ input and initiatives, decreasing motivation and lowering their ability to adapt to changing situations and to improve incorrect plans. A right balance is therefore needed to be able to gain the advantages that control systems can provide, without over-controlling and letting the related negative effects of control lower performance.

2.4 Control systems in collaborations

The previous two sections have discussed control systems and their possible effects. Since this thesis is focused on control systems in governmental collaborations, this section will elaborate on control in collaborations. It will discuss some control systems proposed for collaborations and their effects on the collaboration.

To achieve the desired results in collaborations binding agreements can be necessary. These agreements could take the form of service level agreements (SLAs) (Arenas & Wilson, 2008). “A SLA is a formal contract used to guarantee that consumers’ service quality expectation can be achieved” (Wu & Buyya, 2010, p.1). It explains the service that will be provided, its quality, date of delivery, and security constraints (Arenas & Wilson, 2008).

Miller (1999) has emphasized the importance of clear identification of process ownership. He states that it is important to explicitly describe the responsibilities of the different parties using detailed documentation, which usually includes service level agreements (SLAs). These SLAs should describe the services and activities of the different parties, identify financial responsibilities and performance metrics. Examples of these performance measures are: targets for accounts payable (e.g. early-pay discounts obtained), results from user satisfaction surveys that point out good performances and possible points for improvement, quality measures and ratings for compliance with the SLAs. Using clear SLAs could prevent activities from being overlooked by the different parties and prevent “service creeps” which are additional requests that were not agreed upon (Miller, 1999).

11

Howcroft and Richardson (2012) also mention the importance of strict standards as control methods. They distinguish between two methods: Service Level Agreements and Key Performance Indicators (KPIs). Where SLAs focus on the service provision relation and the agreements between the collaborating organizations, KPIs are formed to measure to what extent actual performance meets up to what is expected (Howcroft & Richardson, 2012).

Authors such as Haponava and Al-Jibouri (2012) have stated that KPIs are usually used to review projects after they are completed. They state that there is a need to develop KPIs that do not only focus on the results, but also measure aspects of the process, so that KPIs can be used for controlling during the process and stimulate learning and improvements. KPIs should further be focused on many aspects; having KPIs that only measure financial aspects (e.g. cost savings) is not sufficient (Palmer et al., 1996; Shen & Liu, 2003). KPIs should also be focused on aspects such as function analysis and value improvements (Lin & Shen, 2007).

Cecil (2000) on the other hand states that motivating employees by empowering them, giving more responsibilities and decision rights to individuals and teams, job rotation, decreasing the layers of management, and direct reporting to top management will increase the chances of success. Another way to motivate employees is the use of incentive pay by matching pay to performance (e.g. in the form of bonuses for meeting specific goals) (Cecil, 2000). Empowering employees can increase performance by increasing employees’ motivation, their learning, and growth (Kaplan & Norton, 1996). Cecil (2000) further states that universal recommendations can be made based on standardized information. Other authors claim that there is no best practice and that standardized approaches are not suited for such complexities, since there are several factors that could influence which systems are most appropriate, such as knowledge sharing, information management and workflow (Marsden, 2010).

The current literature however does not cover the process of how these control systems develop in these collaborations, although there is some literature regarding factors that influence the development of control systems in general. These factors might also have an effect on control in governmental collaborations. They will therefore be discussed in the next section.

2.5 The development of control systems in general

Overall, earlier studies on management control (e.g. Chenhall 2003) found that the development of control systems within an organization is influenced by four characteristics: strategy, size, organizational structure, and the perceived organization uncertainty (Jokipii, 2010).

12

important role in the design of control systems, since the way control systems are used, the type of information needed, and factors that have to be measured are related to the strategy of an organization.

Earlier studies on management control have further found different ways in which the size of an organization can influence how control methods are designed and used. An increase in size leads to more sophisticated control systems (Hoque & James, 2000). It can further lead to more controls (Duncan et al., 1999) stronger control systems (Wallace & Kreutzfeldt, 1995), and control systems that are more tailored to its environment. This can be explained by the amount of resources owned by large organizations and the higher amount of internal differentiation (Jokipii, 2010) in such organizations.

Chenhall (2003) notes that the organizational structure, which comprises the structure of authority, communication, roles, responsibilities, and relationships within the organization, also affects the design of control systems. For example, in centralized organizations the necessity for control is lower, since most decision making happens at the top (Jokipii, 2010), which influences the design of control systems.

Chenhall (2003) further argues that perceived organizational uncertainty, related to an uncertain external environment, can increase the importance and effectiveness of control systems, and particularly the importance of formal controls and traditional budgets. It further calls for more open, externally focused control systems (Chenhall, 2003).

2.6 Pressures for governmental organizations

The factors that influence control systems in general, as mentioned in the previous section, might influence control in governmental collaborations. There are, however, some characteristics typical for government agencies and the pressures they are subjected to that might also affect control in these types of collaborations.

One of the main differences between governmental organizations and organizations in the private sector is that governmental organizations are normative at their core (Lapalombara, 2001). This distinguishes them from private sector organizations that are typically mainly concerned with efficiency and utility (Lapalombara, 2001). This difference could call for different types of control systems. For instance, there can be ambiguity between social and economic objectives in governmental organizations. This increases the importance of making goals specific (Turcotte, 1974).

Gupta et al. (1994) found that contextual factors are important for the development of control systems and they particularly focused on governmental organizations, which are highly institutionalized and subjected to more institutional pressures. Their results show that a highly institutionalized environment leads to adopting bureaucratic control systems.

13

reliable, error free operations and avoiding unfavourable public criticism on the organization (Merton, 1957), and therefore take error preventing and predictable actions, than being focused on improving the effectiveness of operations and economic performance (Anthony as cited in Turcotte, 1974). This can even be increased when the expectations are not met, which can lead to the adoption of more control systems and specifications of actions for all types of eventualities and thereby create a vicious cycle of controls (Crozier, 1964).

These institutionalized expectations can be found in rules, standard operating procedures, blueprints for action, impersonal prescriptions, formalization and documentation (Gupta et al., 1994). Governmental organizations therefore may use formalized bureaucratic control practices that have a greater focus on conforming to these rules and procedures than on trying to reach the organization’s objectives (Emch,1954). The related high amount of forms, reports, manuals and directives are often barely used in practice. Governmental organizations apply these control systems, “which support an image of tidiness and clarity, but are often unrelated to work-unit performance, while they control performance through the more social and idiosyncratic personal and group control” (Gupta et al., 1994, p. 277). These bureaucratic control systems can have a mainly symbolic purpose (Gupta et al., 1994).

March and Olsen (1983) have found that negotiations and finding compromises between different parties with possibly conflicting interests play a huge part in the control in governmental organizations.

Whether the governmental collaboration is mandatory or initiated voluntarily can also affect the control systems that are adopted. This is an important distinction since the reasons of the relationship can substantially differ, and it affects the patterns of integration and coordination between collaborating organizations (Oliver, 1990). Voluntary initiated collaborations are associated with mutual adjustments, while mandated collaborations are related to corporate structures of coordination (Whetten, 1981).

2.7 Changing control systems and environments

Once control systems are in place, changes might lead to adjusting them to better fit the new situation.

Evans et al. (1986) found that organizations modify their control systems to adapt to the changing environment. Control systems must further be adapted to changes within the organization. When this does not happen and they are institutionalized, they may become unresponsive over time (Gupta et al., 1994). Control systems further develop through negotiations (e.g. contracts and agreements may evolve over time by modifications agreed upon by the different parties, which could be caused by unforeseen events) (Arenas & Wilson, 2008).

14

changing, with new organizational forms and where knowledge is an important competitive asset.

In these environments, control should be focused on continuous improvements, market-driven strategies, meeting customers’ needs, customization and the empowerment of employees. These types of control, however, should not be chosen over traditional control systems. A right balance is needed between the accountability and empowerment of employees, between direction from the top and room for bottom-up creativity, between efficiency and experimentation, and between constraints that can limit productivity and the freedom to innovate (Simon, 1995).

Helsloot (2008) also emphasizes that command-and-control is not the best suited control method in demanding times, e.g. during a crisis. While managers often use centralized decision making to deal with these types of situations, research has shown that this is not the appropriate method for these kinds of emergencies. Instead, the focus should be on coordination, which can be seen as the opposite of central command-and-control, and it should include a voluntary aspect. In these situations, command-and-control, that can be based on wrong assumption regarding future developments, can hardly be applied well and should be loosened on some aspects (Quarantelli, 1988).

Emch (1954) stressed the importance of the possibility for employees to take initiatives regarding the actions they are responsible for, especially in changing environments. This allows them to undertake action based on their knowledge and deal with the changes, while having the possibility to accommodate to opportunities that arise, which could lead to possible advantages for the organization.

2.8 Concluding

The literature so far has identified a number of factors that influence control systems in general, effects of control systems and pressures governmental organizations are subjected to. These factors might also influence the control systems in governmental collaborations and could therefore be interesting for the research questions of this thesis. There is further some disagreement regarding the amount of control that is best suited for collaborations. While some authors claim that it is important to document and measure detailed agreements, others advocate loosening control and empowering employees to improve performance. It is therefore important to consider why certain types of control systems are adopted by a collaboration and how they affect the collaboration, which will be done in the remaining part of this thesis. In the current literature it is further stated that control systems should be adapted to changes, usually by decentralizing. How governmental collaborations adapt their control systems as a response to changes will therefore be examined as well.

15

3 Research methodology

3.1 Introduction

This chapter will elaborate on the research methods used for this thesis. As previously mentioned, the aim of this thesis is to investigate how control systems develop in a governmental collaboration and how these systems influence the collaboration. Since literature regarding these topics is scarce, and there are no previously developed hypotheses, an initially broad focus is needed. This fits well with exploratory research that starts broad and narrows down as the research proceeds (Adams & Schvaneveldt, 1991). This thesis will therefore be based on an exploratory study.

3.2 Literature review

The first step in finding an answer to the research questions is a literature review. Articles regarding topics that could be of interest for the research questions at hand are reviewed. Although research addressing the research questions of this thesis is scarce, related research can identify factors that are found to be important in other cases and could possibly influence control in a governmental collaboration. These factors should be taken into account during the research. Literature is further used for definitions of concepts central in this research.

3.3 Interpretative research

This thesis will be based on an interpretative study. An interpretive study is focused on how and why things happen in a specific setting. This type of research seems most appropriate, since the goal of this thesis is not only to describe the developments, changes and effects of control systems, but also how and why they occur.

To be able to describe and explain this, it is important to get a clear picture of the collaboration and its context. Detailed information about the collaboration, its environmental factors, how they influence the development of certain control systems, and how these systems influence the collaboration is necessary. During the research the focus will therefore be on collecting and analyzing rich data, specific to the case.

3.4 Case study

When choosing between quantitative research and qualitative research in the form of a case study, a trade-off has to be made between the generalizability of data and the degree of detailed and accurate data (Hammersly, 1992). Given the detailed and case-specific information needed for this type of research and the limited timeframe, a qualitative in-depth case study with one case, which will be a collaboration within the public sector, seems most appropriate. This type of study makes it possible to collect the type of data needed for this research.

16 Yin (2003, p.1) states:

Each strategy has particular advantages and disadvantages, depending on three conditions: (a) the type of research question, (b) the control the researcher has over actual behavioral events, and (c) the focus on contemporary as opposed to historical phenomena. In general case studies are the preferred strategy when ‘how’ or ‘why’ questions are being posed, when the investigator has little control over events, and when the focus is on a contemporary phenomenon within some real-life context.

Since how and why questions are central in this thesis, and the fact that the research has a mainly contemporary focus and will not influence the control systems, a case study is best suited for this research.

3.5 Data collection method

The data collection methods include interviewing employees and reviewing company documents.

To get the necessary information and different perspectives, employees with various functions are interviewed. The controller can for instance provide specific information on the subject of control methods, while project managers might have more information regarding external factors that influence the collaboration. Further, employees from different organizations within the collaboration are interviewed to avoid possible biases.

The snowball method is used to find suitable interviewees. After discussing the research and research question with a contact person, interviewees that are familiar with these topics were identified and contacted.

The interviews are semi-structured. This type of interview ensures that the relevant topics and questions will be addressed and findings are still comparable, while allowing for the flexibility to discuss and go into relevant topics that come up during the interview. The use of open-ended questions allows for gathering detailed, case-specific information, which is very useful for this type of research.

3.6 Reliability and validity

17

referring to a better representation of the interviewee’s opinion, which he states is key in qualitative study. Lincoln and Guba (1985) define validity (including reliability) in qualitative research as the extent of plausible, credible and trustworthy data, and highlight the importance of consistency and dependability of data and analysis. This research strives for valid and reliable data and conclusions by using and comparing different sources and types of data, collecting rich data to be able to adequately describe important characteristics, and by carefully analyzing the collected data.

3.7 Data analysis

The results of the different interviews will be compared and combined. During the analysis it is important to prevent any interpretation errors. In order to avoid these errors, the main results from the different interviews were also discussed with other interviewees. Overall, this has led to a conformation and further explanation of the results from previous interviews.

The interviews were recorded and transcribed afterwards. According to Berg (2001) data analysis of qualitative data comprises three steps; data reduction, data display and conclusions and verification. Given the often considerable amount, qualitative data should be transformed, reduced and simplified to improve its accessibility, manageability and understandability, which allows for patterns and themes to be identified. This can be realized by summarizing collected data, coding the data or by identifying themes (Berg, 2001). During this research, the collected data was first transcribed, then grouped and summarized. The data was grouped per topic (e.g. effects on control systems, factors), which clearly showed the overlap between the collected data. This overlapping data was reduced, and the data was brought back to only relevant data.

Data display involves presenting the data in an organized manner, e.g. in the form of summaries or tables (Berg, 2001). During this research the summarized data per concept was displayed in a table. While this table only includes the key point of the collected data, it allows for recognizing patterns and it gives a clear overview. This table can be found in the appendices.

The third step in analyzing qualitative data as stated by Berg (2001) includes the conclusion and verification. These conclusions can be made based on the summarized data that is grouped per concept. During this research the summarized and grouped data, that includes the data from different sources, helped to systematically find answers to the research questions and draw conclusions. The verification comprises verifying data (e.g. by carefully executing the analysis) and clarifying the steps taken during the research (Berg , 2001).

3.8 Concluding

18

4 Case introduction

4.1 Introduction

In this chapter the organization that will be central in this thesis, which is Regio Groningen-Assen, will be introduced. It is a public sector collaboration between the two provinces Groningen and Drenthe, and the municipalities Assen, Bedum, Groningen, Haren, Hoogezand-Sappemeer, Leek, Noordenveld, Slochteren, Ten Boer, Tynaarlo, Winsum en Zuidhorn. The collaboration contains four regional programs; housing, region park (nature & landscape), business areas & economy, and accessibility. Within these programs there are several projects that need to be executed. These projects are (mainly) financed with a contribution that is deposited yearly by the participants to the region fund (in some cases supplemented by a contribution from the State), and are executed by the different participants, often in cooperation with other organizations such as educational institutions and housing cooperatives.

The collaboration started in 1996, and it differentiates itself from other public sector collaborations in two ways. First, it is a collaboration between two types of governmental organization; provinces and municipalities, over the borders of provinces. These different organizations are subjected to different rules, which add to the complexity of the collaboration. Another important aspect of the collaboration is that it is initiated voluntarily by the fourteen participants. The participants saw the importance of the regional interests and the necessity of cooperating to obtain the regional goals.

4.2 Structure of the organization

The collaboration basically has three important layers. The first is the steering committee, which consists of members of the city councils and provincial executives of the participants. This committee is the highest steering group and is occupied with strategic decision making (e.g. financial decision making), the development of board assignments, and monitoring the strategy as documented in the “Regiovisie” and the agreements made between the participants. Part of this steering committee forms the executive committee.

The second layer consists of the portfolio holders, which are the aldermen and delegates of the participants. They are involved with steering and monitoring the execution of the four programs as documented in the board assignments. They decide on what needs to happen per program within the frameworks of the board assignments. The four programs of the collaboration (housing, region park, business areas & economy, and accessibility) have their own portfolio holder.

Program teams are the third layer. Each program has its own program team, consisting of the civil servants of the participants and members of external organizations in some cases.

19

portfolio holders, with process support, steering projects, and measuring if the progress on projects is according to plan.

Another important party is the program management, which consists of the program director (the chairman) and the directors of the collaborating participants. Their tasks include, amongst others, giving advice to the steering committee, preparing the meetings of the steering committee and executive committee, communicating with governments, citizens and societal organizations, and guiding the preparation and execution of essential regional projects.

Further, there is an independent quality team that evaluates designs and plans and gives advice.

The municipal councils and states provincials of the participants are, although not directly part of the collaboration, ultimately responsible. They are the eventual control mechanism. The aldermen and the mayors also influence the collaboration, and some of them are members of the steering committee.

Figure 1: Organization chart Regio Groningen-Assen

municipal councils and states provincials

20

5 Case descriptions

5.1 Introduction

In this chapter the case will be discussed. It will elaborate on control systems within Regio Groningen-Assen. In the following section the control systems are discussed. The third section is focused on how these control systems came into being, and it involves characteristics of the collaboration that influenced this type of control, and the effects of the control systems. The changes in the control systems will be described in the fourth section, including how different factors have influenced these changes. Finally, there will be a brief conclusion in the fifth section.

5.2 Control systems

In this section the control systems used within the collaboration and the parties responsible for the type of control are discussed to give an overview of the types of control. These control systems include agreements, plans, measurement systems and feedback processes.

5.2.1 Covenant

The provinces and municipalities have documented the foundations of the collaboration, the organization, the execution, the financial aspects, the evaluation, monitoring and future activities in a covenant. This covenant is the starting point of the collaboration. It is agreed upon and signed by all participants and they can decide to adjust, extend or end it on the basis of every evaluation. This evaluation (actualization) takes place every four years. The participants have deliberately chosen to keep the covenant relatively abstract regarding the specific programs and activities to prevent the need for continue adjustments. It is mainly oriented on the agreements regarding the collaboration itself.

5.2.2 Multiannual investment program

Future project and contributions of the participants are documented in a multiannual investment program, which is composed yearly by the steering committee. This program contains a collection of the projects that are already financed by the region, the projects that are about to be financed, and projects still considered. The types of projects selected depend on the regional goals and interests. The multiannual investment program further specifies conditions that have to be met before a project is eligible for a contribution.

5.2.3 Board assignments, execution programs and rules framework

Board assignments are documents composed by the steering committee that contain a concise description of the intended goals, investigations, and actions that have to be performed, the planning, budget, approach, and intended outcomes. It further elaborates on possible risks during the execution. The steering committee, which is responsible for the strategic decision making, yearly composes these board assignments on a strategic, tactical and operational level, and monitors its progress.

21

can be developed by the portfolio holders. Portfolio holders are responsible for the execution of the programs and their tasks include steering, monitoring and providing insight into the progress of the programs. The progress is documented and presented to the steering committee every six months for evaluation and, if needed, adjustments.

On a more detailed level, there is a framework of rules. The projects that will be executed within the programs have to fit within this framework. This project framework describes the criteria that have to be met by projects in order to be considered and the conditions that have to be met by the participant before receiving a contribution from the region fund.

The participants that are responsible for the execution of the projects have their own rules and procedures (e.g. development plans) that they have to take into account, this is separate from the control within Regio Groningen-Assen.

5.2.4 Barometer

Once in every six months, a member of the program bureau makes a round to evaluate all projects and verify if the progress and finances are according to plan. This will be summarized and reported in a barometer. Colors are used to display the progress of the projects in the barometer. Green indicates that the project goes according to plan (both time wise and financially) and there are no obstacles. Yellow is used to display that a project should be monitored more carefully and attention is needed, e.g. because of unforeseen circumstances, but no further action is required yet. Red indicates that a projects deviates from the plan, previously determined results cannot be achieved anymore, and action is needed. If a deviation is identified, the reason behind this (e.g. higher complexity or getting a license takes longer than expected) should be found and documented in the barometer. These barometers are presented to the portfolio holders during meetings, and solutions and new action are discussed and approved by the portfolio holders for the projects in red. The barometer also includes a summary of the completion of actions that were agreed upon during the discussion of the previous barometer, which will also be discussed during these meeting.

When projects are not going according to plan, the program bureau first contacts the person responsible for the project to verify the cause and a possible solution. If a solution cannot be found, the program bureau will report it to the portfolio holders. If it concerns a problem with a large impact, it will eventually be presented to the steering committee. Eventually,

solutions or alternatives are found through negotiations. Projects are further evaluated, to examine what went wrong and how this can be prevented in the future.

5.2.5 Monitoring the external environment

22 5.2.6 Quality team

Besides the earlier mentioned systems, Regio Groningen-Assen has a quality team that monitors the quality of the designs and plans within Regio Groningen-Assen. This team consists of independent experts. Participant can ask the team for advice, but it also gives unsolicited advice. In principle, participants should act upon this advice. However, participants can choose to apart from it, in which case a motivation is required. The quality team further aims to coach and stimulate the participants.

5.2.7 Administrative organization

Another important control method within Regio Groningen-Assen is the administrative organization. It gives an overview of, amongst others, the financial developments, and documents such as the multiannual investment program, the annual account and budgets are checked. These documents will also be provided to the municipal councils and states provincials for examination.

5.2.8 Personnel control

The control systems mentioned in the previous paragraphs can be listed under either action or result control; there are rules specifying actions that should be undertaken and plans and measures that focus on results. Personnel control is less present within the region. There are team days, once a year, that are focused on the development of employees (e.g. how they can improve their effectiveness or deal with certain situations) to improve performance. Employees could further undergo training, but as one of the interviewees mentioned: “there is a possibility to undergo training, but this should usually be on one’s own initiative”, although this can also be suggested by the director if he observes possible room for improvement. The main focus is, however, on result and action control.

5.2.9. Control overview

As discussed in the previous paragraphs, different parties have different roles in the control process within the collaboration. To give a global overview of the control, the main tasks of the different parties are represented in figure 2. Regio Groningen-Assen uses this Deming circle to improve the quality of (planning) processes during the execution of the different programs.

Figure 2: Deming circle (source: Regio Groningen-Assen company documents)

23

5.3 Development of control systems

This section focuses on the characteristics of Regio Groningen-Assen that have influenced the development of control systems. It elaborates on how and why these factors contributed to the development of these control systems and their effect on the collaboration.

5.3.1 Structure and size

Regio Groningen-Assen is a complex collaboration. With fourteen participating governmental organizations, including provinces and municipalities that are subjected to different sets of rules and policies, and are concerned with different themes, the structure of the collaboration becomes quite complicated. Besides the participants, other stakeholders that indirectly influence the collaboration, such as the municipal councils and states provincials, should also be taken into account. Further, by far most governmental collaborations in the Netherlands consist of only two or three municipals, and the number of governmental collaborations decreases as the number of participants increases (Postma et al., 2012). Based on the number of collaborating municipals and provinces, Regio Groningen-Assen can be seen as a large collaboration.

This complicated structure and the large number of participants created a need for a transparent control method to provide a clear overview for all stakeholders. The board assignments, that concisely describe the intended goals, actions and resources needed, expected output, planning, risks, and the organization, are used by the steering committee to deal with this complexity by increasing the transparency for the different parties and thereby answer to this need. It supports the communication with the municipal councils and states provincials regarding the goals and results of the collaboration. It further gives the portfolio holders and program teams an accessible framework for the execution of programs and clarifies the roles of the different parties.

Besides adding to complexity, the structure of the collaboration has a strong direct influence on the development of control systems. The different layers within the collaboration have their own decision rights, tasks and responsibilities. As stated by a member of the program bureau: “the municipal councils and states provincials are eventually in charge, since they are democratically chosen”. For example, the steering committee has the authority to make strategic decisions. These levels of authority and centralized decision making are partly responsible for the control systems used, such as the earlier mentioned board assignments and reporting and monitoring between the different levels. To give an example, the program bureau that monitors the progress of the projects is not an administrative body. It therefore does not have the authority to decide on certain actions when they signal projects deviating from the plans. Instead they monitor and report developments to portfolio holders using the earlier mentioned barometer. In this way, the structure (authority levels) of the collaboration directly influences the control systems.

24

continuously, and preceding a decision there is a lot of negotiation. As one of the employees of the program bureau said: “the biggest part of our job consists of deliberating, finding compromises, considering the interests of the participants, and maneuvering between this”. This can be a long process. Other stakeholders and societal interests should also be taken into consideration. Although formally all participants have an equal say in the collaboration, in practice, participants with more persuasiveness, and the larger participants (often with a higher financial interest, more knowledge and expertise) could have a stronger influence on the decision making.

5.3.2 No legal entity

A characteristic of Regio Groningen-Assen that further influences the control systems is that the collaboration is initiated voluntarily and is no legal entity. The provinces and municipalities therefore retain their decisive say. As a consequence, Regio Groningen-Assen does not have the same obligations as a legal entity, e.g. to justify to the State on set times. It only reports to the participants. So in that sense it is restricted by less rules and formats than governmental collaborations that are legal entities. As one of the interviewees stated:

There are two sides of control. Within the region we are concerned with our own control systems, such as the framework of rules that is determined within the region, to be able to justify the investments that are made with money from the region fund to the participants. If we decide to invest in a certain project executed by one of the participants, that participant is subjected to many other rules and control systems. But that is their side, and it is separated from the control within the collaboration.

To give a concrete example, the fund of Regio Groningen-Assen is not subjected to the number of rules that the funds of other governmental organizations are usually subjected to. There are many conditions related to the spending of municipalities (e.g. development plans have to be made). Regio Groningen-Assen is less constrained by these rules since it is not a legal entity and the participants are contributing their own money. The fourteen participants have set the rules and standards regarding the spending themselves, which are less complex and more efficient. These are negotiated and decisions are made unanimous in the steering committee. However, the participants themselves are subjected to the rules and policies of governmental organizations (e.g. making development plans), and they apply them as they should. The region does not apply or monitor these rules and procedures. Therefore, they have no direct effect on the region, although the region has to deliver some documents and information to their participants, so that they could justify their spending to the State. In that sense, these rules do have an indirect effect. The region is further subjected to other rules, such as the European procurement framework.

25

programs. As opposed to other types of governmental organizations that are more bound to rules, policies and expectations that lead to adopting formal methods, which are rarely used in practice, the control within the region is relatively compact.

Another effect of this type of control is that it influences the attitude of the participants towards the collaboration. Since decision making and setting standards happen collaboratively and unanimous, the participants see the activities performed within the collaboration less as an obligation. As mentioned during an interview: “things only happen if the participants want it to happen”. This makes participants more open to collaborating, which positively influences the collaboration. It makes them more willing to meet each other’s requests.

Further, the fact that the collaboration is not a legal entity leaves the final say and responsibilities with the municipal councils and states provincials. They are eventually responsible for the final control and the “principals” of the region. They receive documents, such as the annual accounts and budgets and check if these comply with the agreements.

Finally, the fact that the participants make contributions to the region fund also affects the consequences of not meeting certain plans or agreements. As an interview explained:

Regional means should be spend within a certain amount of time, but since the money assigned to the participants is ultimately their own money, there are no real sanctions in the sense of lowering the money assigned to a participant for instance, when this does not happen.

If the program bureau monitors that a project does not progress as it should, solutions are sought to get the desired result after all, instead of sanctions, such as lowering the contribution that the participant receives from the region fund.

5.4 Changes in control systems

This section is focused on how several factors led to changes in the control systems and it explores the internal and external changes, and their effect on control systems within the collaboration.

5.4.1 Governmental collaborations

26 5.4.2 Recession, budget cuts and decentralization

External factors that have strongly influenced the control systems within the collaboration are the recession and budget cuts. These changes have led to participants lowering their contribution to region fund.

This decrease in the amount of money available has led to a greater emphasis on the financial aspects, efficiency, and developing a more realistic view. When money is scarce, investments have to be discussed more extensively. However, these aspects remain less dominant than in commercial organizations, since the many societal interests cannot be ignored. While these aspects are still important, the recession and budget cuts have created the need to pay more attention to financial aspects.

This focus on finances is reinforced by the trend of decentralization within the government. More and more tasks are being delegated to provinces and municipalities. This increase in tasks makes controlling the finances even more challenging.

These developments have changed the control systems in several ways. For one, projects within some programs are now being divided into smaller phases. As a member of the program bureau explained:

Dividing projects into smaller phases increases the controllability. Each of these phases has to be valuable separately too. It gives the certainty that there is enough money to finish a phase, there is no waste of money due to halting in the middle of projects.

This new approach increases the flexibility and the ability to keep projects and finances under control.

Another change caused by the recession is a change in the long term planning. The collaboration used to work more with set long term plans (e.g. projects that should be taken on in the next ten years). Even though long term planning is still important, not everything is determined several years in advance anymore. More is left open to improve the possibility to adapt to the opportunities that come along.

27

collaboration can be improved. Another proposed change is reforming the program structure of the collaboration.

Finally, the recession and budget cuts have created a need to lower the number of employees within the collaboration itself. Some of the controlling activities, e.g. monitoring activities, which are now performed within the region, will have to be executed by the participants themselves in the future.

5.4.3 Societal changes

Developments within the environment of Regio Groningen-Assen are monitored closely. To give an example, the developments in the housing sector are monitored every three months in a collaboration with housing cooperatives. They assess, among others, the need for building new houses and the realization of new houses to prevent under- or overproduction. Based on these developments they examine the actions, that should be undertaken by Regio Groningen-Assen, in a joint meeting with housing cooperatives. Other societal changes that are noticed (e.g. by the program bureau, project teams or societal organizations) could be incorporated in initiatives that will be presented for approval to the portfolio holders (or other parties, dependent on the effect) in the form of actions.

Execution programs are developed once a year. This increases the flexibility and makes it possible to transform ideas into action that will be executed on a relatively short term, and incorporate them into the programs quite frequently.

By influencing the projects that are executed, societal changes indirectly influence the control systems. They indicate the importance and support base in a certain area or within certain groups, and adapting to these changes improves the link with societal needs. An example of a change caused by societal changes is making housing a separate program, as an effect of its growing importance within society.

5.4.4 Politics

Control methods within Regio Groningen-Assen are further affected by politics. During the elections, new representatives are chosen, which changes the composition of the steering committee. These new leaders have a new approach and a different outlook on the collaboration that could translate itself into changing controlling methods. The elections in March 2014 were further partly the motivation behind anticipating actualization to 2012. As stated in the covenant, the actualization should take place every four years. This would, however, mean that it would still be in process during the elections. The anticipation makes it possible for the actualization to be completed before the elections.

5.4.5 Uncertain and changing environment

28

Governments cannot steer everything anymore these days, which has led to a reconsideration of the degree of detail and fixed projects in the multiannual investment program. Flexibility became more important, which can improve the possibility of taking initiatives related to unanticipated societal developments and adapt to opportunities that arise. This has translated itself into a more flexible organization. An interviewee explained: “investment programs are not completely assigned to specific projects anymore, there is more room for flexibility now”. A specific example of this development is money reserved for unforeseen circumstances, which was not there in the past.

The possibility to take initiatives has a motivating effect. It gives employees a “we feeling” and contributing ideas can make them identify more with a project.

The changing and uncertain environment has further made scenario planning more important. Since the changes in the environment can influence the collaboration in several ways, it is valuable to consider the impact of different scenarios on projects before committing to them.

5.4.6 Desire for more control

Another factor that influenced the control within the collaboration is the extra attention paid to finances within the participating organizations, and a clear justification became necessary for them. For this reason they decided in 2012 to also improve the control within the collaboration. This led to an improved administrative organization. In the past the administrative organization of Regio Groningen-Assen contained some baselines, but it was not sufficient. Not enough attention was paid to the control systems. The numbers were checked and there was a check to control if money was spent as intended, the accountant performed a legitimacy test. The need for more control in 2012 led to adopting stricter and more control systems, including a legitimacy tests. This legitimacy test is focused on more aspects and also checks if the activities performed remained within the frameworks of the collaboration (e.g. the covenant, multiannual investment program and European procurement framework). These adjustments and additions make it possible to adequately display the current situation to the participants and give them a clear overview.

5.5 Concluding

This chapter has elaborated on the control systems used by the collaboration. It described how internal and external factors affected the development of control systems within the collaboration and how these factors led to adapting the control systems. Also, some main effects of these control systems were mentioned. The next chapter will go further into this.

6 Case implications

29

6.2 Case implications

Overall, the factors that influenced the development of control systems in the governmental collaborations are comparable with factors that influence control systems in general, which are the size, structure, strategy, and uncertainty of an organization (Joikipii, 2010). All of these factors had an effect on the control systems, however, some affected the control in a different way than expected.

The case under study is, based the fourteen participants, a relatively large governmental collaboration, considering that most of the governmental collaborations within the Netherlands consist of two or three participants (Postma et al., 2012). This size has influenced the control by creating a need for clarity, e.g. regarding the tasks and responsibilities of the different parties. Rather than applying extra control systems, there was an emphasis on using clear, concise control systems that gave a clear overview, to achieve this result. The size of the collaboration further increased the importance and amount of negotiation, since finding compromises can be more time consuming in a collaboration with many participants. While March and Olsen (1983) noted the importance of negotiations and finding compromises between the different parties in governmental collaborations, they did not relate this to the size of the organization. Other authors (e.g. Wallace & Kreutzfeldt, 1995; Duncan et al., 1999; Hoque & James, 2000) did find that size can influence control systems, however, they found different effects such as more sophisticated control, more control systems, stronger control systems, and control systems that are more tailored to their environment. These effects are enabled by the high amount of resources that large organizations possess and the differentiation within those organizations (Jokipii, 2010).

The case study further showed that the structure can directly influence control systems through the different authority levels that determine the amount of measurements and reporting between the different layers (as was also found by Chenhall, 2003). Besides this effect, the complex structure of the collaboration, which is caused by the different types of participants and the different layers within the collaboration, also affected the control systems indirectly by creating a need for transparency for the participants, e.g. regarding frameworks, tasks, and responsibilities. This led to adopting clear, concise control systems, such as the board assignments that give a clear overview of these aspects.

The strategy of the collaboration determines, amongst others, the types of control systems, what information is needed, and which factors should be measured (Miles & Snow, 1978; Porter 1980; Miller & Droge, 1986). For example, when the strategy calls for a fast response to changes in the environment, flexible control is best suited. An important objective of Regio Groningen-Assen is responding to societal needs (e.g. the demands in the housing sector) and several control systems are therefore focused on monitoring and adjusting to these needs (e.g. the control systems that focus on measuring the developments in the housing sector, and making housing a separate program as a consequence of the growing importance).