Master’s Thesis
Understanding the Interaction between
Control, Trust, and Perceived Risk in
Public Sector Joint Ventures
By Anneke Giliam
Master of Science, Business Administration
Specialization: Organizational and Management Control
University of Groningen
Faculty of Economics and Business
June, 2016
Student number: s2204258
Supervisor: prof. dr. ir. P.M.G. van Veen-Dirks
Co-assessor: drs. D. Tavenier
1
Preface
Before you lies my Master’s Thesis in Business Administration with a specialization in Organizational and
Management Control. With this thesis my study at the University of Groningen comes to an end.
This Master’s Thesis is supported by an internship at KPMG Advisory in Groningen. I have been asked to conduct a study about public sector joint ventures and how local governments can design their management control systems in order to effectively manage their relationships with public sector joint venture companies. Therefore, this thesis is about the role of control, trust, and perceived risk in the vertical relationships between public sector joint ventures and local governments.
I would like to thank my supervisor prof. dr. ir. P.M.G. van Veen-Dirks for her help, guidance, and support. I would also like to thank all the participants for their time and effort. Interviewing people with different skills and expertise resulted in interesting conversations, and I gained valuable information for my thesis. Moreover, the conversations gave me a broader understanding of how public sector joint ventures work, and how municipalities and provinces control the relationships with public sector joint ventures.
I would also like to thank KPMG for hosting my internship over the last 5 months. I have found it to be an interesting learning experience and it has been a pleasure to gain knowledge about KPMG Advisory and to meet the staff who work there. As a result of this study KPMG would like to organize an Around the Table meeting for all the parties concerned. At this meeting we want to discuss the main findings and explain what the added value of this study is. During this Around the Table meeting we want to provide the participants the opportunity to communicate their thoughts.
2
Abstract
There is an increase in cooperation among public sector organizations which is often the result of reform agendas. Public sector Joint Ventures (JVs) are separate administrative entities that undertake public service activities on behalf of local governments. This study focuses on the vertical relationship between local governments and public sector JVs. The first aim of this study is to examine the role of control, trust, and perceived risk in public sector JVs. Based on two case studies, it can be concluded that not only the perceived risk level and the level of trust affect the design of vertical control packages, but vertical control packages, in turn, affect the perceived risk level and the level of trust as well. A vertical control package refers to the combination of control mechanisms used by local governments to control the vertical relationship with their public sector JVs. Furthermore, the vertical control packages are affected by activity diversification and parent size differences. In addition, this study shows that the probability of unsatisfactory business performance does not significantly impact the design of the vertical control packages. The (social) impact when something goes wrong has a greater force to change the design of vertical control packages. Furthermore, it appears that social control mechanisms play a significant role in developing trustworthy relationships, and that behaviour control mechanisms are used to minimize the performance risk, rather than outcome controls. The second aim of this study is to contribute to the discussion on the trust - control nexus. This study supports the interactive perspective, that is that trust is considered to produce control, while control is considered to produce trust. This study also provides evidence which supports the substitutive perspective, which means that trust can be seen as a substitute for formal control.
3
Table of Content
PREFACE ... 1 ABSTRACT ... 2 1. INTRODUCTION ... 5 2. LITERATURE REVIEW ... 62.1PUBLIC SECTOR JOINT VENTURES ... 7
2.2MANAGEMENT CONTROL ... 8
2.3INTER-ORGANIZATIONAL TRUST ... 9
2.4INTER-ORGANIZATIONAL RISK ... 10
2.5TRANSACTIONAL FACTORS ... 10
2.6RELATIONAL FACTORS ... 12
2.7INTERACTION BETWEEN TRUST,CONTROL, AND RISK ... 13
2.7.1 Trust – Control Nexus ... 13
2.7.2 Risk – Trust Nexus ... 15
2.7.3 Control – Risk Nexus ... 15
2.8CONCEPTUAL MODEL ... 16
3. METHODOLOGY ... 18
3.1RESEARCH METHOD ... 18
3.2METHOD OF DATA COLLECTION ... 18
3.3METHOD OF ANALYSIS ... 19
4. RESULTS... 21
4.1PUBLIC SECTOR JOINT VENTURE ROVAN.V. ... 21
4.1.1 Transactional factors ... 21
4.1.2 Relational factors ... 22
4.1.3 Outcome control mechanisms ... 23
4.1.4 Behaviour control mechanisms ... 23
4.1.5 Social control mechanisms ... 24
4.1.6 Risk and control ... 24
4.1.7 Trust and control ... 25
4.2PUBLIC SECTOR JOINT VENTURE WADINKO N.V. ... 26
4.2.1 Transactional factors ... 26
4.2.2 Relational factors ... 27
4.2.3 Outcome control mechanisms ... 28
4.2.4 Behaviour control mechanisms ... 28
4.2.5 Social control mechanisms ... 29
4.2.6 Risk and control ... 29
4.2.7 Trust and control ... 29
4
5.1TRANSACTIONAL FACTORS ... 30
5.2RELATIONAL FACTORS ... 31
5.3CONTROL MECHANISMS ... 32
5.4TRUST –CONTROL NEXUS ... 33
5.5OWNER -CLIENT CONFLICT ... 33
5.6SUMMARY ... 34
6. CONCLUSION ... 34
6.1MANAGERIAL IMPLICATIONS ... 36
6.2LIMITATIONS AND FUTURE RESEARCH ... 36
REFERENCES... 38
APPENDICES ... 44
APPENDIX A–SEMI-STRUCTURED INTERVIEW QUESTIONS TYPE 1 ... 44
APPENDIX B–SEMI-STRUCTURED INTERVIEW QUESTIONS TYPE 2 ... 47
APPENDIX C–SEMI-STRUCTURED INTERVIEW QUESTIONS TYPE 3 ... 50
5
1. Introduction
Nowadays, it is widely recognized that cooperation in the form of strategic alliances can result in significant benefits such as synergies. By sharing ideas, knowledge, or other resources, partners believe that they can obtain something greater than they can achieve on their own (Greasley et al., 2008). Strategic alliances are inter-organizational cooperative arrangements aimed at realizing the partners’ strategic goals (Das and Teng, 2001). There are many different types and forms of strategic alliances, such as horizontal alliances, vertical alliances, joint ventures, or joint production (Nooteboom et al., 1997; Das and Teng, 2001). Increasingly, public sector organizations cooperate in order to exploit economies of scale, minimize risk, gain access to expertise, and so on (Cäker and Siverbo, 2011). Cooperation in the public sector is often the result of reform agendas with the aim of enhancing the efficiency and effectiveness of the delivery of public services (Grafton et al., 2011; Muir and Mullins, 2015).
This paper will study public sector joint ventures in which public sector organizations cooperate. A joint venture (JV) is a separate established entity which is owned by two or more partners. These separate administrative entities undertake public service activities on behalf of the local governments. This paper places emphasis on the vertical inter-organizational relationship between the local governments (owners) and the Joint Venture Company (JV Company). The term Inter-Organizational Relationship (IOR) will be used in this study, even though the local governments and the JV company belong to the same parent organization. This term is applied because it is commonly used in public sector oriented literature (Cristofoli et al., 2010; Kastberg, 2016).
According to Edelenbos and Klijn (2007) such complex inter-organizational networks gain importance in modern network society despite the complexity and unpredictability of these networks. A large amount of research has indicated trust and control as the solution to cope with this complexity (e.g. Edelenbos and Klijn, 2007; Lane and Bachmann, 1998; Nooteboom et al., 1997). Trust is based on positive assumptions regarding the intentions and capabilities of the other partner (Bachmann, 2001). In contrast, control is based on the assumption that the other partner can act opportunistically (Bachmann, 2001). Control mechanisms are used to influence the partner’s behaviour in such a way that it favours achievement of the organizational goals (Bradach and Eccles, 1989; Costa and Bijlsma-Frankema, 2007; Das and Teng, 2001). The Transaction Cost Economics (TCE) framework is used in this study to describe which transactional factors influence the choice of efficient governance structures since TCE is one of the dominant theories to analyse management control packages in inter-organizational relationships (Ding et al., 2013).
Research shows that the interaction between control, trust, and perceived risk plays an important role in IORs (e.g. Das and Teng, 2001; Gulati and Nickerson, 2008; Langfield-Smith, 2008). Perceived risk is defined as the manager’s estimation of objective risk, contrary to objective risk which is based on probabilities that certain consequences will occur (Das and Teng, 2001). Several studies argue that the combination of management control systems and trust can reduce risk in IORs and decrease the probability of failure (Langfield-Smith, 2008; van der Meer-Kooistra and Vosselman, 2000). However, research is not conclusive about how trust and control relate to each other.
6
included in the conceptual model of this study. According to Cäker and Siverbo (2011) parent size has an effect on the vertical control packages in public sector JVs as well as the degree of connectedness between the JV’s activities and the parent’s other activities.Much research has been done about strategic alliances in the private sector (e.g. Langfield-Smith, 2008; Das and Teng, 2001; van der Meer-Kooistra and Vosselman, 2000). However, the research area of cooperation between public sector organizations is rather undeveloped and requires further research (Cäker and Siverbo, 2011; Cristofoli et al., 2010; Greasly et al., 2008; Hall et al., 2005; Kastberg, 2016). According to Cäker and Siverbo (2011) cooperation within the public sector is as common as inter-organizational relationships in the private sector. Therefore it is surprising how little research has been done on public sector JVs. This study will contribute to the closure of this literature gap by providing theory development about how local governments manage their vertical relationships with public sector JV companies. Since there is little known about the phenomenon of public sector JVs a theory-generating research method is needed to study this phenomenon (Aken et al., 2012). It is expected that JVs in the public sector may vary significantly from JVs in the private sector due to differences in goal complexity and ambiguity (Rainey and Bozeman, 2000) and due to the highly institutionalized context in which local governments cooperate (Grafton et al., 2011). As described above, there is a need to understand more about the role of control mechanisms, trust, and perceived risk in these public collaborations. Local governments might use this information in developing more effective and successful management control systems which might minimize their perceived risk level. Therefore, the research question of this study is:
What is the role of control mechanisms, trust, and perceived risk in the vertical relationship between local governments and Dutch public sector joint venture companies?
Data is gathered by conducting a comparative case study in which semi-structured interviews are the main data source. This paper examines two public sector JV companies: ROVA N.V. and Wadinko N.V. which are both joint-stock companies with the provincial and municipal governments as the only shareholders. In addition to interviewing employees and managers of the provincial and municipal governments, managers of the public sector JV companies are also interviewed. The paper is structured as follows: chapter two provides a theoretical framework in which the core literature on control, trust, and risk in IORs will be reviewed. In chapter three the methodology of this study will be explained, followed by chapter four in which the findings will be given and analysed. Thereafter, the results will be discussed and a conclusion will be given. The paper closes with suggestions for further research.
2. Literature Review
7
JVs, the working of these three concepts should be examined. Since little is known about public sector JVs, literature on strategic alliances or JVs in the private sector will be used to guide theory development. This chapter will start by explaining the concept of public sector JVs and the associated vertical coordination problem. Thereafter, the concepts management control, inter-organizational trust, and inter-organizational risk will be explained. Next, transactional and relational factors, which might affect the design of vertical control packages or the perceived risk level, are examined. Thereafter, a review of the literature on the interaction between control, trust, and risk will be given. Finally, a conceptual model will be presented, combining all of the relevant concepts.2.1 Public Sector Joint Ventures
Partner cooperation in public administration is as common as cooperation in the private sector (Cäker and Siverbo, 2011). However, there is an increase in collaboration within and across policy fields, both nationally and internationally (Metcalfe and Lapenta, 2014). There are various types of public cooperation such as partnerships, strategic alliances, outsourcing, or contracting relationships. Table 1 lists various types of public-public partnerships that can be distinguished.
Type Sub-type
Public authority – public authority
Public authority – community
Development partnerships
International partnerships
Inter municipal
Government – municipal
Public authority – NGO Public authority – community Public authority – trade union
High-income country
Public authority – low-income country Public authority
Public authorities from different countries Public authorities from neighbouring countries
Table 1. Typology of Public-Public Partnerships (Hall et al., 2005).
8
example, managing conflicts between individual local governments’ own interests. A vertical coordination problem can be, for example, the alignment of the public sector JV company’s behaviour and the local government’s interests. Another important characteristic is that local governments remain responsible and accountable for service delivery (Narayanan et al., 2007). Therefore this study places emphasis on the vertical management control systems used by local governments to influence the behaviour of the public sector JV company’s behaviour.2.2 Management Control
The concept of management control has been studied thoroughly. Management control can be described as a process that influences the behaviour of employees in such a way that it favours achievement of the organizational objectives (Bradach and Eccles, 1989; Costa and Bijlsma-Frankema, 2007; Das and Teng, 2001). According to Ouchi (1977), control can be conceptualized as a process of monitoring and evaluation of behaviour or of outputs. Caglio and Ditillo (2008) have reviewed management control systems in IORs. They distinguish different breadths of inter-organizational controls such as control archetypes and management controls. Control archetypes such as markets, hierarchies, and hybrids, are largely based on transaction cost theory, trust-based theory and organizational theory. This study will focus on such control archetypes and management control mechanisms in public sector JVs. First, within-firm control systems will be conceptualised in order to understand what control systems partners use separately. Literature on the design of management control systems in IORs will be reviewed, in section 2.7. Section 2.7 will also explain which control mechanisms are expected to be used to mitigate risk in public sector JVs.
Many models have been developed which can be used for studying the design of management control systems within an organization (e.g. Ouchi, 1979; Merchant, 1985; Simons, 1995). Management control systems can be categorised into various approaches. According to Costa and Bijlsma-Frankema (2007) and Langfield-Smith and Langfield-Smith (2003) there are two main approaches to control mechanisms: formal control and informal control.
9
the behaviour and actions of the JV company and in such a way this supervisory role can serve as a behaviour control mechanism (Dekker, 2004).Informal control or social control reflects Ouchi’s (1979) ‘clan control’ which is based on culture, shared norms, values, and beliefs which encourages certain behaviour. Social or informal controls are unwritten control mechanisms which are based on commitment and consensus about appropriate behaviour (Ouchi, 1979). The internalization of organizational goals by employees results in high commitment and motivation to obtain the organizational goals (Das and Teng, 2001). The effect of social control is strengthened by activities such as frequent interactions, meetings, negotiations, codes of conduct, or rituals (Langfield-Smith and Smith, 2003).
2.3 Inter-organizational Trust
Inter-organizational trust plays a significant role in IORs. Sanchez et al. (2014) even state that trust is a key success factor in IORs. Trust is a complex concept with many different definitions. Trust across organizations can be measured at multiple levels: 1) interpersonal trust, which refers to trust between individuals across different organizations, 2) organizational trust, which refers to individual trust that a member of organization A has in organization B as a whole and 3) inter-organizational trust, which refers to the trust organization A as a whole has in organization B as a whole (Vélez et al., 2008). This study will focus on inter-organizational trust as a multidimensional concept (Das and Teng, 2001; Nooteboom, 1996). According to Inkpen and Curral (2004) inter-organizational trust should be perceived as dynamic rather than static, since the level of trust may change during the partnership.
10
to create goodwill-based trust, joint objectives and interests should be established. Many factors influence the level of goodwill-based trust, examples are: common values and norms, frequent contacts, openness, or the partner’s reputation (Vélez et al., 2008). The concepts social control and trust are similar to each other, but an important difference is that social control is the ability to influence behaviour, based on shared values and norms. In contrast, trust is not about influencing behaviour but about having positive expectations concerning others’ behaviour (Vélez et al., 2008).2.4 Inter-organizational Risk
This section will focus on perceived risk in IORs in general. Section 2.5 will explore the factors which are expected to influence the perceived risk level in public sector JVs. In social sciences much empirical research has been done into risk taking. Research shows that generally interfirm collaboration is associated with high risks which results in a high failure rate (Das and Teng, 2001; Langfield-Smith, 2008; Ding et al., 2013). This study will focus on downside risk, which refers to the possible losses. Das and Teng (2001) explain that risk taking is often related more to the possible losses than the potential gains. Furthermore, downsize risk is often considered to be managerially more important than the associated gains of risk (March and Shapira, 1987). Therefore this study will focus on the possible losses, rather than the potential gains.
Perceived risk is an estimation of objective risk by a manager in contrast to objective risk which is based on probabilities that certain consequences will occur (Das and Teng, 2001). Therefore managers might have different perspectives on risk and this may influence the choice of control mechanisms (Langfield-Smith, 2008). According to Das and Teng (2001) perceived risk can be subdivided into relational risk and performance risk. Ding et al. (2013) refer to both relational risk and performance risk as transaction risk.
Relational risk is the risk of the partner acting opportunistically, which expresses itself in shirking responsibility, deception, and withholding or distortion of information (Das and Teng, 2001). Relational risk in strategic alliances exists when the interests of the partner firms are not congruent. It is expected that these conflicts also exist in the vertical relationships between local governments and public sector JV companies.
Performance risk is defined as the probability of having dissatisfying business performance, which is not unique to interfirm relationships, but exists in all types of business operations (Das and Teng, 2001). Factors that may result in performance risk are increased competition, lack of competences, demand fluctuations, et cetera. Performance risk can be shared between partner firms in IORs, but this does not apply to relational risk. This distinction between performance and relational risk is important in deciding which control mechanisms partner firms should adopt (Das and Teng, 1996).
Nooteboom et al. (1997) distinguish two dimensions of relational risk: “the probability that something will go wrong and the size of the loss incurred when it does” (p. 1). This study will examine these dimensions for both (perceived) relational risk and (perceived) performance risk.
2.5 Transactional Factors
11
van der Meer-Kooistra, 2007; Langfield-Smith and Smith, 2003; Langfield-Smith, 2008). It is expected that asset specificity, behavioural uncertainty, and environmental uncertainty directly affect the perceived risk level and therefore indirectly affect the design of vertical control packages. Output measurability is expected to directly affect the design of vertical control packages. Furthermore, it is expected that especially output measurability and asset specificity play an important role, due to goal complexity and ambiguity in the public sector (Rainey and Bozeman, 2000) and high inter-dependency between the public sector JV company and the parent organizations. A common framework used to determine appropriate governance structures in IORs is Transaction Cost Economics (TCE) (Langfield-Smith and Smith, 2003). Transaction Cost theory can be used to study the IOR between local governments and public sector JV companies since any contracting problem can be viewed from a TCE perspective (Williamson, 1985). TCE can be used in studying how firms govern their operations and which transactional aspects influence the choice of efficient organizational structures or governance structures (Williamson, 1985; Langfield-Smith and Smith, 2003). Governance structures can be categorised into one of three different forms: markets, hybrids (including strategic alliances) and hierarchies (Williamson, 1991). Williamson (1991) distinguishes these three types of governance structures based on three aspects. These aspects are the frequency of the transactions, the related uncertainty with these transactions, and the asset specificity of the transactions.Williamson (1985) defines asset specificity as the “durable investments that are undertaken in support of particular transactions, the opportunity cost of which investments is much lower in best alternative uses or by alternative users should the original transaction be prematurely terminated” (p. 55). In other words, a high level of asset specificity means that investments will lose their value if the public sector JV is terminated. Asset specificity is said to be a key construct in studying the governance structures of IORs (Langfield-Smith and Smith, 2003; Lui et al., 2009), because investments in specific assets can result in opportunistic behaviour by the other party (in this study the public sector JV company) (Lui et al., 2009). TCE predicts that dependency between parties is created when there is a high level of asset specificity, which will result in high switching costs and complicated governance situations (Langfield-Smith and Smith, 2003). Asset specificity can be caused by site specificity, physical assets specificity, human assets specificity, brand name capital, dedicated assets, and temporal specificity (Williamson, 1991). In the context of JVs, Kamminga and van der Meer-Kooistra (2007) explain that parent companies use control mechanisms to safeguard their specific investments made in a JV. However, the type of asset specificity influences the ability to protect their investments. For example, human asset specificity or procedural asset specificity are more difficult to safeguard by contracts, compared to physical asset specificity.
12
According to TCE, transaction characteristics such as asset specificity, environmental uncertainty, and behavioural uncertainty, are factors that influence relational risk and performance risk (Williamson, 1985). Langfield-Smith (2008) has assessed the relationships between transactional characteristics and the perceived performance risk and relational risk. Langfield-Smith (2008) argues that a high level of asset specificity results in dependency between partner firms and high switching costs, with an increased level of relational risk as a result. Furthermore, when behavioural uncertainty is high due to difficulties in the alignment of the goals relational risk is expected to be higher. In contrast, environmental uncertainty results in a higher performance risk due to, for example, market volatility or a lack of partner capabilities. It is expected that within the context of public sector JVs a high level of human asset specificity exists. Due to reform agendas local governments provide certain services at a distance with the aim to increase efficiency and effectiveness (Grafton et al., 2011). When local governments do not perform the activities themselves, this may lead to a loss of internal knowledge within the parent organization. Therefore, it is expected that there is a knowledge transfer from the local governments to the public sector JV companies. In such a way the public sector JV companies acquire a high level of human assets which leads to dependency between the partners and high switching costs, and which in turn results in a high level of relational risk (Langfield-Smith, 2008).The fourth transactional factor used in this study is output measurability. Output measurability refers to the extent to which the output can be measured objectively and precisely (Ouchi, 1979; Eisenhardt, 1985). Public sector organizations often face difficulties such as measuring output, interpreting output, and developing outcome measures (e.g. Burgess and Metcalfe, 1999; Slyke, 2007; Smith, 1993). According to Ouchi (1979) and Eisenhardt (1985) outcome controls are most appropriate when output measurability is high. In contrast, behaviour controls or social controls are more appropriate when output measurability is low. Since the output and objectives of public sector organizations are often difficult to define and to measure (e.g. Burgess and Metcalfe, 1990; Slyke, 2007; Smith, 1993), it is expected that local governments do not make extensive use of outcome controls. However, there is an increased interest in the use of outcome performance indicators to enhance control of public sector organizations (e.g. Smith, 1993; Verbeeten, 2008). Verbeeten (2008) explains that clear goals and measurable results are positively associated with the performance in public sector organizations, especially when incentives are linked to performance. Therefore it is expected that local governments will seek to make increased use of outcome controls when possible.
2.6 Relational Factors
This section will focus specifically on relational factors which are expected to have an influence on the design of vertical control packages to be used in public sector JVs. According to Cäker and Siverbo (2011) parent size and diversification of activities have an influence on the vertical control packages used in Swedish public sector JVs. It is expected that these factors also influence the design of management control systems in Dutch public sector JVs. Therefore these factors are included in the conceptual model.
13
to the number of activities which are subjected to control mechanisms. For example, a narrow vertical control package means that only a few activities are controlled by the parent companies. Control tightness refers to the extent to which the parent companies control the actions of the public sector JV company (Cäker and Siverbo, 2011). Loose control means, for example, that the parent companies are less involved in key decisions and that the public sector JV’s behaviour is monitored less frequently. Furthermore, a dominant position by one owner might be the result of differences in parent size. This dominant position might lead to a situation in which the design of vertical control packages is affected by the control packages used in other public sector JVs owned by the dominant partner (Cäker and Siverbo, 2011).When the activities of the public sector JV company are not closely related to the other activities of the local government, which is called diversification, the parent company will apply more loose and narrow control mechanisms. The reason for this is that diversification increases vertical information asymmetry and makes it more difficult for the parent companies to apply broad and tight controls since they often have not the specific knowledge required to control the activities (Cäker and Siverbo, 2011).
2.7 Interaction Between Trust, Control, and Risk
This section will focus on the interaction between control, trust, and risk in IORs. Das and Teng (2001) and Langfield-Smith (2008) have studied the interaction between control, trust, and risk in strategic alliances. According to them, trust and control together determine the perceived total risk. First, an overview of the literature on the relationship between trust and control in interfirm relations will be provided, thereafter the relationship between risk and control will be examined, and lastly the relationship between control and risk will be discussed.
2.7.1 Trust – Control Nexus. Much research has been done concerning the relationship between trust and control in interfirm relations. For instance, the paper of Vosselman and van der Meer-Kooistra (2009) theorizes the accounting-control-trust nexus in interfirm transactional relationships. They view governance structures as formal control structures which function as safeguarding devices as well as incentive systems, and therefore constrain and manage the behaviour of the partners involved. Roath et al. (2002) distinguish two types of governance structures: contractual-based governance and relational-based governance. Contractual-based governance is based on formalized, legally binding agreements or contracts used to govern the IOR. In contrast, a relational-based governance has a governance structure based on mutual trust, commitment, and relational capital. Lee and Cavusgil (2006) argue that relational-based governance enhances the performance of the IOR in its strength, stability, learning, and amount of knowledge transfer. They also state that formalized, legally binding contracts are necessary to provide security.
14
being applied. Pre-specified norms, standards, and detailed rules of behaviour are controls which will be used. These controls are based on supervision and performance measurement and evaluation. In the trust based pattern trust becomes the dominant control mechanism. In such situations no detailed contracts are used, as the relationship is based on competence trust, contractual trust, and goodwill trust. Informal control mechanisms or social controls are often used in these IORs.The paper of Langfield-Smith and Smith (2003) draws on the model of van der Meer-Kooistra and Vosselman (2000), and extends the model by examining the role of specific control mechanisms and trust. Based on an analysis of the characteristics of the transaction, environment, and parties of strategic alliances they examined which control strategy was adopted and which control mechanisms were used. It was found that in the market based pattern there are no specific control mechanisms used and trust is also not necessary to obtain control. In a bureaucratic based pattern the main control mechanisms adopted are outcome controls and behaviour controls. Direct intervention is managed by, for example, performance targets, detailed contracts, and detailed rules of behaviour. Under a trust based pattern outcome controls are also used, but to a more limited extent. Trust and social controls play a more significant role in overcoming opportunistic behaviour and information asymmetry under a trust based pattern. Information sharing and high levels of communication will enhance the trust level (Langfield-Smith and Smith, 2003). Kamminga and van der Meer-Kooistra (2007) explain that high levels of information asymmetry between the parents and the JV company might result in situations in which the parent organizations are placed at a distance. Hence, according to Kamminga and van der Meer-Kooistra (2007) the parents will focus on creating good relationships, building trust, and stimulating open communication. Therefore, an incomplete and flexible contract type is more appropriate. However, prior research is not conclusive about how trust and control relate to each other. Below, three different perspectives are explained regarding this relationship. The first perspective is called the substitutive perspective. This perspective suggests that control and trust are substitutes, which means they are inversely related (e.g. Dekker, 2004; Gulati and Nickerson, 2008; Inkpen and Currall, 2004; Lazzarini et al., 2004). In other words, more control results in less trust and more trust results in less control. Gulati and Nickerson (2008) argue that trust between organizations can lead to substitution of a more formal governance mode (e.g. ally) by a less formal mode (e.g. buy). In such a relationship trust is seen as a substitute for formal contract and can be seen as an alternative control mechanism. For example, when a company has a favorable reputation, less formal control mechanisms may be used compared to a company which is considered less trustworthy. Furthermore, trust minimizes goal conflict and therefore reduces the need for formal control to discourage opportunistic behaviour (Dekker, 2004) Moreover, substantial use of formal control mechanisms may give the impression of a lack of faith (Das and Teng, 1998).
Another perspective which is distinguished in the literature is the complementary perspective. According to this perspective formal governance and trust are complementary, which means that control and trust are mutually reinforcing (Costa and Bijlsma-Frankema, 2007; Das and Teng, 1998; Sitkin, 1995; Zucker, 1986). This relationship implies that more control leads to more trust (Costa and Bijlsma-Frankema, 2007). The usage of formal control mechanisms may result in a trustworthy relationship, by reducing the risk and by recording the behaviour and performance of both partners (Das and Teng, 1998; Poppo and Zenger, 2002).
15
governance structures. They argue that control and trust have the same objective, namely the absorption of behavioural uncertainty. Therefore control and trust interact with each other, meaning that control needs trust, and trust needs control. It also means that control produces trust and vice versa. This interactive perspective on the trust - control nexus is based on the ‘duality perspective’ of Möllering (2005), which assumes that trust and control create each other. According to Vosselman and van der Meer-Kooistra (2009), a central element in the process of trust building is relational signalling, which refers to behavioural signals the partners produce in order to show their willingness to not act opportunistically. The study of Vosselman and van der Meer-Kooistra (2009) focuses on governance structures as formal control structures. In their article they distinguish thin trust and thick trust in which the latter is developed by positive behavioural expectations with regard to the behaviour of others. An effective governance structure compensates for potential opportunism, resulting in thin trust. So thin trust is produced by formal control structures and is called ‘thin’ because no positive behavioural expectations are produced. However, in order to produce thin trust and to increase the effectiveness of the governance structure, institutional trust from the environment is necessary. Thick trust is the result of trust building by relational signalling. This process of trust building is accelerated by the governance structure. This means that an effective governance structure strengthens the trust development process. Subsequently, the governance structure is reinforced since the trust building process enhances the credibility of the formal structures of the governance structure. In other words, an embedded governance structure is required for successful cooperation and can compensate for opportunistic behaviour. Governance structures therefore need to find a balance between control and trust and both partners should agree upon this structure.2.7.2 Risk – Trust Nexus. According to Das and Teng (2001) trust can reduce the perceived risk. More specifically, they argue that goodwill trust minimizes relational risk and thus opportunistic behaviour of the partner firm. Perceived performance risk will be reduced by competence trust, rather than goodwill trust. Since competence trust is the confidence that the partner firm is able to perform certain tasks, this does not necessarily mean they have the intention to do so.
2.7.3 Control – Risk Nexus. Management Control is another mechanism which can be used to reduce perceived risk in strategic alliances, as explained in section 2.2. Behaviour control mechanisms can regulate the behaviour of the partner firm by applying rules, standards, and procedures to prevent unacceptable behaviour. In such a way behaviour controls can reduce relational risk according to Das and Teng (2001). Similarly, outcome controls might be used to minimize perceived performance risk since performance risk is more result oriented. While outcome controls place the emphasis on key performance measures and thus on performance risks, behaviour controls do not focus on specific goals or achievements and are therefore less useful to reduce performance risk. Finally, Das and Teng (2001) argue that social control is applicable in reducing both performance risk and relational risk. Social control reduces perceived relational risk since social controls are based on shared norms, values, and beliefs which increases goal congruence between partner firms. Simultaneously, social control may reduce performance risk since social controls stimulate partner firms to establish common goals, which is according to Hatfield and Pearce (1994) a critical requirement for successful cooperation.
16
parent companies will make extensive use of behaviour- and social control mechanisms to compensate for this relational risk, or by a high level of goodwill trust (Das and Teng, 2001). Furthermore, behavioural uncertainty and environmental uncertainty is expected to differ among public sector JVs. When environmental uncertainty is high, for example, due to market volatility or a lack of partner capabilities, it is expected that the perceived level of performance risk will be high (Langfield-Smith, 2008). According to Das and Teng (2001) this perceived level of performance risk can be minimized by competence trust or by outcome controls, since outcome controls focus on key performance measures. Since outcome controls might not always be applicable in public sector organizations, it is expected that competence trust plays a more important role in minimizing performance risk. And when, for example, behavioural uncertainty is high, it is expected that goodwill trust, social controls, and behaviour controls play an important role in minimizing relational risk, caused by this behavioural uncertainty (Das and Teng, 2001).Dekker (2004) presents in his paper a framework of inter-organizational control which builds on TCE. Dekker (2004) also explains that control mechanisms can mitigate both relational risk and performance risk. In his paper he identifies specific control mechanisms that are used in alliances to minimize relational and performance risk. Table 2 provides an overview of these control mechanisms.
Outcome control Behaviour control Social control
o Goal setting
o Performance monitoring o Rewarding structures o Open book accounting
o Planning
o Standard Operating Procedures
o Rules & Regulations o Alliance Board o Behaviour monitoring:
-Cost & Quality control -Reporting & checking devices
-Arbitration clauses
o Interaction o Social networks o Interactive goal setting o Interactive decision
making
o Joint problem solving o Partner development o Joint history & cultural fit
Table 2. Control mechanisms used in IORs to mitigate risk (Dekker, 2004).
2.8 Conceptual Model
17
After determining the level of trust, the applied control mechanisms, and the level of perceived risk, the relationship between these three concepts will be assessed. This means, for example, that the proposed framework not only examines how the perceived risk level affects the design of vertical control packages, but also examines how the design of the vertical control packages affect the perceived risk level. This study might result in useful information for managers about how trust and control mechanisms can be used to minimize the perceived risk level. The relationships which are specifically important to public sector JVs will be briefly explained again below.Figure 1. Proposed integrated framework of control, trust, and perceived risk in public sector JVs (based on Cäker and Siverbo (2011), Das and Teng (2001), Dekker (2004), Langfield-Smith (2008), and van der Meer-Kooistra and Vosselman (2000)).
Due to reform agendas local governments place certain activities at a distance, with the aim to increase efficiency and effectivity (Grafton et al., 2011). Providing services at a distance might result in a loss of knowledge within local governments. Therefore, it is expected that there is a knowledge transfer from the local governments to the public sector JV companies. In such a way the public sector JV companies acquire a high level of human assets which results in dependency between the partners and high switching costs. According to Langfield-Smith (2008) high levels of assets specificity result in a relational risk. When asset specificity is relatively high it is expected that the parent companies make extensive use of behaviour control mechanisms and social control mechanisms to compensate for this relational risk, or by a high level of goodwill trust (Das and Teng, 2001).
As explained, it is expected that two relational factors (parent size and activity diversification) influence the design of the vertical control packages used by local governments. If the public sector JV is not equally owned by the local governments, it is expected that the largest shareholder occupies a dominant position. It is expected that this dominant position affects the design of the vertical control packages of the other local governments. The second relational factor is activity diversification. Particularly if the public sector JV company’s activities are politically uninteresting, it is expected that when the activities of the public sector JV company are not closely related to the other activities, the parent company will use less control mechanisms, because they will not have the expertise to do so.
18
When, for example, the performance risk is high, it is expected that local governments minimize this risk by competence trust, rather than by using outcome controls.3. Methodology
3.1 Research Method
In this study a qualitative research method was used to collect data. Since the phenomenon of public sector JVs is not well explained in academic literature a case study research method is appropriate to study this phenomenon (Aken et al., 2012). Qualitative research enables the researcher to create deeper insight into individual perceptions and provides a better understanding of the interaction between control, trust, and risk in public sector JVs. Empirical data for this study was collected by conducting two exploratory case studies. According to Yin (1994) case studies are especially useful when: “a how or why question is being asked about a contemporary set of events, over which the investigator has little or no control” (p. 9). Therefore case studies are not only useful in exploratory stages of a research project, but also for descriptive and explanatory research purposes (Eisenhardt, 1989; Yin, 1994). Yin (1994) and Eisenhardt (1989) argue that well designed case studies can be helpful for theory development because of the robust nature of the empirical data. An important aspect of case studies is that data collection methods are being combined such as interviews, observations, or archival data (Eisenhardt, 1989). Examining two case studies enables the researcher to make comparisons between both cases. The two case studies which were developed are Wadinko N.V. and ROVA N.V. which are both joint-stock companies with the provincial and municipal governments as the only shareholders. Wadinko N.V. is a venture capital company which is jointly owned by the province of Overijssel and 24 municipalities in Overijssel, Noordoost Polder, and Zuidwest Drenthe. Wadinko N.V. stimulates business growth and therefore encourages employment at local and regional level. The second case study is ROVA N.V. which is jointly owned by 21 municipalities in the region Ijssel-Vecht, Achterhoek and Eemland. ROVA N.V. is a waste management company and was established with the aim of enhancing the efficiency and effectiveness of household waste disposal.
3.2 Method of Data Collection
19
useful information which can be used in answering the research question (Saunders et al., 2009). Formal contracts and reports were gathered and analysed which provided supplementary information about how formal control was exercised. Appendix D provides an overview of the secondary data sources used in this study.Based on theoretical sampling, which means cases were selected based on theoretical reasons (Eisenhardt, 1989), two public sector JV companies in the Dutch public sector (Wadinko N.V. and ROVA N.V.) were chosen. This selection was based on the differences in activities performed by Wadinko N.V. and ROVA N.V. and the different sectors in which they operate. Due to these differences the associated perceived risk level and the vertical control packages used by the local governments might also differ. These differences enabled the researcher to analyse several factors that might affect the perceived risk level and the related vertical control packages’ design.
Both companies are jointly owned by more than twenty local governments. Three of the five selected local governments are shareholder of both Wadinko N.V. and ROVA N.V. This enabled the researcher in asking the respondents to compare both relationships. Interviewees were selected based on their functions at the local governments or at the public sector JV company. All interviewees were involved in the control of the public sector JVs, either from a financial or policy point of view. Some interviewees were approached after the first interviews were conducted, based on references of initial respondents. Furthermore, also a manager and controller of Wadinko N.V. and ROVA N.V. were interviewed to find out how they experience the relationship with their shareholders. Moreover, three concern controllers of municipalities in the northern provinces of the Netherlands and a strategy manager of a comparative waste disposal company were approached. These four interviews are irrespective of the case studies of Wadinko N.V. and ROVA N.V. During these interviews more general questions regarding the control of government-related entities1 have been asked (see Appendix C). Therefore, these interviews provide a
more general view on the control mechanisms used by local governments. In one interview three people were interviewed together and in three interviews two people were interviewed together, which was preferred by the interviewees. These interviews provided a broad range of information and resulted in relevant discussions. Table 3 provides an overview of the interviewees. The interviews lasted between 45 and 90 minutes. In order to guarantee privacy the names of all the respondents have been anonymised.
3.3 Method of Analysis
The interviews were held in Dutch and were audio-recorded and fully transcribed, which enabled the researcher to re-listen the interviews several times. A full transcription ensures that everything what has been said can be analysed and therefore increases reliability (Aken et al., 2012). The method of Eisenhardt (1989) was applied in analysing the data which means that first a within-case analysis was done and thereafter a cross-case analysis. This means that first for each case study the related interviews and documents were coded and analysed separately. Codifying is a way to systematically categorize data which enables researchers to organize data and to uncover patterns (Mack et al., 2005). The Atlas.ti. program was used to support the analysis of the qualitative data. Atlas.ti. is a computer-assisted qualitative data analysis software which enables the researcher to code and categorize documents and to search within and across documents and categories. The use of computer-assisted qualitative data analysis software makes the analytic process more transparent (Friese, 2014). Open coding was the initial step in the coding procedure, which is defined as “the process of breaking down, examining, comparing,
20
conceptualizing, and categorizing data” (Strauss and Corbin, 1990: 61). In this process all the documents were read properly and different concepts were coded. Thereafter, similar concepts were divided into the following identifiable categories: 1) environmental uncertainty, 2) behavioural uncertainty, 3) asset specificity, 4) competence trust, 5) goodwill trust, 6) outcome control, 7) behaviour control, 8) social control, 9) output measurability, 10) parent size, and 11) activity diversification. During the initial process of open coding a few extra categories were added: 12) financial risk, 13) social impact, 14) owner relationship, 15) client relationship, 16) personal interests, and 17) internal knowledge level.After the initial open coding process the different categories were connected with each other, which is called axial coding (Strauss and Corbin, 1990). According to Strauss and Corbin (1990) axial coding means exploring relationships between categories. During this process relationships between control mechanisms, trust, and perceived risk were analysed. After this within-case analysis both case studies were compared with each other on the above described points. Since this research is supported by an internship at KPMG Advisory Groningen, the criterion of ‘recognition of results’ plays an important role. The recognition of research results refers to reasonable, plausible, or at least possible results (Aken et al., 2012). This recognition is important for the development of management control systems which can enhance the governance of the vertical relationships between local governments and public sector JV companies, in such a way that the perceived risk level can be minimized.
Organization Interviewees Number of interviews
Wadinko N.V. ROVA N.V. Province of Overijssel Municipality of Zwolle Municipality of Amersfoort Municipality of Steenwijkerland Municipality of Raalte Municipality of Groningen Municipality of Smallingerland Municipality of Heerenveen HVC N.V. Groep - Controller
- Manager administrative and legal affairs - Policy developer government-related entities /
Account manager Wadinko N.V. - Alderman for finance
- Policy developer government-related entities - Coordinator government-related entities - Contract manager ROVA N.V. - Policy advisor waste disposal
- Account manager, Economic Development - Contract manager ROVA N.V.
- Strategic financial advisor - Policy & advice team leader - Division controller
- Policy advisor waste disposal - Account manager Wadinko N.V. - Concern controller
- Concern controller - Concern controller
- Coordinator government-related entities - Program manager - Strategy manager 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Total number of interviewees 21
21
4. Results
4.1 Public Sector Joint Venture ROVA N.V.
Public sector JV ROVA N.V. (ROVA) is a Dutch waste management company that mainly collects household waste (ROVA, 2014). For a few municipalities ROVA also provides the management of public spaces. ROVA was established in 1996 and it is now jointly owned by 21 municipalities in the region Ijssel-Vecht, Achterhoek and Eemland. ROVA not only focuses on efficient waste collection, but also on innovative waste collection such as ‘reverse collecting’. In 1996 ROVA was only an execution-service company for the associated municipalities. But throughout the years ROVA has developed itself towards a professional advisor in the field of waste disposal and durability. As a manager of ROVA explained:
“We started as an execution-service company, but we always had the ambition to become a full-service company. After twenty years, we do not just collect waste, but we are the ‘back-office’ for the municipalities which means we also give policy advice to our municipalities. We are ourselves fully aware of the fact that we are a public company. We focus on the constantly changing environment and we try to analyse how these changes will affect us, and our municipalities.”
The most important consideration in the establishment of a jointly owned company was to enhance the efficiency and effectiveness of household waste disposal. According to a policy advisor at the municipality of Zwolle the establishment of ROVA was the result of reform agendas with the aim to decrease expenditures and downsize services. The expectation was that ROVA could collect waste disposal more effectively and efficiently due to economies of scale. Furthermore, it was expected that by establishing a public sector JV waste disposal techniques could also be improved.
22
account manager at the municipality of Amersfoort there is a need to actualize the service contract and the associated conditions if a party wants to exit.Difficulties in predicting the intentions and activities of the partner firm might result in behavioural uncertainty. ROVA has individual service contracts with each of the municipalities, which minimizes behavioural uncertainty. Municipalities decide how they want to design their waste disposal policy and which activities should be carried out by ROVA. A manager of ROVA explained that standardization is not necessary to obtain economies of scale, but ROVA tries to create harmonization between its municipalities and therefore advises to adopt certain waste management structures. The flexible contract is highly preferred by the managers of the municipalities since this flexibility rarely results in difficulties in predicting the intentions and activities of ROVA.
Environmental uncertainty refers to the environmental dynamism and complexity. All interviewees acknowledged the highly dynamic environment in which ROVA operates. An account manager at the municipality of Amersfoort explained that first the focus was on the collection of waste. But in recent years, the environment in which ROVA operates has become highly dynamic. Waste and recycling is a hot item on the government’s agenda. Waste is no longer a waste, but a raw material. This complexity is the result of the continuous changes in rules and regulations. Since the establishment of ROVA many new laws have been developed that describe how municipalities should take care of waste. According to a manager of ROVA the following questions are upcoming: how do we deal with waste, how can we improve waste separation, how can we motivate society to increase waste separation, and how should local governments, the business community and citizens interact with each other? There are also technological developments in the waste disposal industry, but these changes take place more gradually.
The fourth transactional factor included in the conceptual model is output measurability, which refers to the extent to which the output can be measured objectively and precisely. The output measurability of ROVA’s activities is very high. A manager of ROVA explained:
“Our goals are highly measurable, we know exactly how many kilograms we have collected or the number of complaints that have been made. Regarding the management of public spaces the goals are less measurable, but we developed a systematic approach for this.”
4.1.2 Relational factors. Parent size appears to have an important effect on the control mechanisms used by municipalities. Smaller municipalities have less frequent contact with the public sector JV company and they apply fewer control mechanisms. For example, the two largest municipalities are currently in dialogue with ROVA about the development of performance indicators. The smaller municipalities are not (yet) in conversation with ROVA about such developments. Particularly, the level of development of the internal organization plays an important role. Four interviewees, of relatively small municipalities, indicated that due to the continuous change of personnel they are not able to effectively control public sector JVs. Often, certain tasks are not covered well, with the result that these activities are being performed poorly or not performed at all.
23
related to the other activities. Interview responses showed that there is no activity diversification. An account manager at the municipality of Amersfoort explained that waste disposal is an important core activity of municipalities and therefore the interest in the management control and performance of ROVA is substantial. Since waste disposal is one of the core activities of local governments they place more emphasis on controlling these activities. This will be discussed in the following section.4.1.3 Outcome control mechanisms. During the annual shareholders meeting the upcoming budget of ROVA is approved. During this meeting the shareholders also approve the dividend payment. Furthermore, ROVA reports on a quarterly basis to its municipalities. According to all the interviewed managers these reports are very detailed and provide sufficient information. For each type of waste it is reported how many kilograms are collected, what the recycling percentage is, and what the waste collection is per citizen. ROVA has defined internal organizational objectives, such as an acceptable complaint rate. Regarding the management of public spaces there are a few performance indicators established. Quality sheets are used by the municipalities to check whether public spaces are clean. However, no quantitative performance agreements have been made between ROVA and the municipalities. Three interviewees, of relatively small municipalities, argued that they believe that performance indicators are not required since ROVA performs well. But the majority of the interviewees acknowledged that they should establish key performance indicators. Currently, ROVA and the two largest municipalities are trying to develop such performance indicators. An account manager of a relatively large municipality argued:
“There is a need to develop key performance indicators. For us it is not relevant to know how often a drain is being cleaned, but it is more relevant to know what percentage is working effectively during a year. This means that we do not have to interfere in their business, but only focus on the results, and if necessary address the problem. I think we can gain control by applying outcome controls.”
However, managers of the municipalities do not know (yet) which performance indicators are relevant and if there should be any consequences linked to these objectives. Another issue is about who should establish these performance indicators and whether these indicators should be established at shareholder level, or at municipality level. A manager of ROVA supports the idea of developing performance indicators, however he emphasized that there is no need for the local governments to control everything. A policy advisor at the municipality of Zwolle explained they would also prefer to use an open book accounting system in order to create more insight into the financial performance of ROVA.
24
municipalities. However, for specific projects ROVA tries to conform to the desires of the municipalities. Such decisions are taken in consultation with the relevant municipalities. Finally, ROVA is a two-tier board company which consists of a Management Board and a Supervisory Board. The Supervisory Board is appointed by the shareholders and is closely involved in the development of the strategy of ROVA. An alderman for finance explained that the Supervisory Board plays an important role in supervising the policy of the Management Board.4.1.5 Social control mechanisms. Social controls are unwritten control mechanisms which are based on commitment and consensus about appropriate behaviour. Social controls play the most important role in the vertical relationship between the parents and ROVA. All the respondents explained they have regular, informal contact with ROVA. Both ROVA and the municipalities emphasized the importance of this interaction, since they believe that frequent contact strengthens the relationship. All the respondents perceive ROVA as a close partner. Important shortcomings or politically sensitive issues are always discussed with the municipalities, even though this is not based on contractual terms. A manager of ROVA explained:
“Frequent contact with our municipalities is important, because otherwise we risk alienating our municipalities. We need to understand each other and speak the same language. Therefore we have an internal account manager who is responsible for maintaining the contact with our municipalities. But this is not only our responsibility, but also of the municipalities themselves. I encourage them to acquire more knowledge about our work.”
The interviewed municipalities rarely have conflicts with ROVA. The reason for this is that they use joint decision making and when they experience problems, joint problem solving is used. A policy advisor at the municipality of Zwolle explained that first conflicts are discussed at operational level. When these conversations do not result in the desired outcomes, it will be discussed at managers level, or with the alderman. But the interviewees explained that most problems are being resolved at operational level. Furthermore, the cultural fit between the municipalities and ROVA also plays an important role here. The municipalities and ROVA share the same view on the development of ROVA and on the objectives regarding waste disposal. Four interviewees, of relatively large municipalities, explained that the main difference between the cultures of the public sector JV company and its parents is the decision-making process. ROVA places more emphasis on its business interests and has a stronger focus on time efficiency. However, due to the frequent meetings and informal interactions convergence is created between the interests of the parents and ROVA.
25
A manager of ROVA explained that they are aware of the fact that they are a public organization, which is funded by public money. Therefore they closely involve the municipalities in their risk management process. Every year ROVA conducts a risk analysis which is approved by the shareholders during the shareholders meeting. Many parents adopt this analysis without adapting it to their own situation. An account manager at the municipality of Amersfoort explained that they currently adopt the risk analysis of ROVA in their budget. He explained that they should improve their risk management process by assessing and adapting the risks to their own situation. One interviewee at the municipality of Zwolle explained:“Currently, we write down our risks associated with our relationship with ROVA. An important point for improvement is to actually analyse these risks and understand what these risks mean. In such a way we can act on the basis of what we say and make real progress.”
A policy advisor at the municipality of Raalte explained that they first describe which risks their municipality is exposed to, thereafter they examine the impact of these risks, and then they assess the probability that these risks will occur. Based on this analysis they can implement their control mechanisms. However, none of the municipalities has developed concrete control mechanisms which should be implemented when the risks increase. The interviewees explained that if the risk increases, they will increase the frequency of meetings. When these interactive meetings are not sufficient they would consider using internal control mechanisms (personnel surveillance) and increase their decision making authority. These actions can be categorized as behaviour control mechanisms.
4.1.7 Trust and control. Both goodwill trust and competence trust play a major role in controlling the vertical relationship between ROVA and the municipalities. All the interviewees have great confidence in the work of ROVA, which is based on the very close cooperation over the years. A controller at the municipality of Steenwijkerland explained that ROVA honours their commitment they have entered into. Besides delivering high quality services, ROVA provides very detailed reports which are provided in time. According to all the interviewees providing information on a timely basis strengthens the trust relationship between the parents and the public sector JV company. The alderman of finance at the municipality of Zwolle explained:
“Trust is based on the quality of the people, the quality of the organization, the quality of the administrative culture, the internal control, and the fact that they deliver fairness, stability, and results.”
26
“I understand that the municipalities want to use control mechanisms, and we accept this.However, when municipalities constantly ask for extra information and want to control everything, this will affect our relationship. We do not have to justify every decision, there is a limit.”
When the level of confidence decreases, the first step undertaken by the municipalities is an increase in the use of behaviour control mechanisms. An alderman of finance uses the following scale: “trust me, tell me, show me, involve me”. This means that if the performance risk or relational risk increases, the parent company will start asking questions and monitor the behaviour of the public sector JV company. When this does not lead to satisfactory results, they will move on to ‘involve me’, which means the parent company wants to be involved by creating an audit commission, developing standard operating procedures, or other rules.
4.2 Public Sector Joint Venture Wadinko N.V.
The second public sector JV company which was examined is Wadinko N.V. (Wadinko), which is jointly owned by the province of Overijssel and 24 municipalities in Overijssel, Noordoost Polder, and Zuidwest Drenthe. Wadinko is a venture capital company that participates mainly in manufacturing and service industry companies (Wadinko, 2014). Wadinko invests in this sector since companies in this industry are often highly innovative, which presents many development and growth opportunities. Wadinko invests in companies by providing venture capital, knowledge, or management support. They focus on three different types of problems: 1) the problem of succession in businesses, 2) fast growth of businesses, and 3) businesses with incidental problems. A controller of Wadinko explained that they yearly receive approximately 140 requests from entrepreneurs. Wadinko visits these entrepreneurs and in 80% of the cases, they decline the request within one hour. This is because Wadinko strongly focuses on three C’s: charisma, competence, and consistency. If one of these characteristics is missing, they will not invest any time or effort into these requests. The overall aim of Wadinko is to stimulate economic activity and thus the employment in Overijssel, Noordoost Polder, and Zuidwest Drenthe. This mission reflects the personal conviction of a controller of Wadinko:
“I believe that everyone matters, and that everyone who can work should work since working results in structure, self-awareness, and income.”
27
their specific region. However, the local council did not agree because of the large dividend payments. None of the interviewees could explain the exact consequences of withdrawal.Behavioural uncertainty is relatively low according to an account manager at the province of Overijssel. He explained that Wadinko is aware of its public shareholders and knows what is happening in the public sector and which issues are important. Therefore he does not experience difficulties in predicting the intentions and activities of Wadinko. Furthermore, the smaller municipalities have less contact with Wadinko and therefore find it more difficult to predict the intentions and activities of Wadinko.
According to a controller of Wadinko they operate in a relatively stable environment since it is predictable what will occur year on year. This enables Wadinko to develop a fairly reliable five-year plan. However, the parent companies define the environment in which Wadinko operates as dynamic and complex, due to the risks which are associated with a venture capital company (i.e. high risk of failure of investments).
The main objective of Wadinko is to stimulate employment. With regard to this objective Wadinko developed two performance indicators; the amount of FTE generated and the Return on Equity. Both objectives can be categorized as highly measurable.
4.2.2 Relational factors. The province of Overijssel owns approximately 25% of the shares of Wadinko and is thereby the largest shareholder (Wadinko, 2014). The influence of parent size on the control mechanisms used is clearly visible since the largest shareholder has frequent contact with a controller of Wadinko (four times a year), while at municipalities (which are smaller) only the alderman has contact with Wadinko once a year. A controller of Wadinko explained:
“The province is in the lead and the smaller municipalities are often only interested in the dividend payment. But more important are the differences in personal interests between the local governments. Some aldermen have an entrepreneurial background and are therefore more interested in Wadinko compared to other aldermen.”
According to an account manager at the province of Overijssel smaller municipalities are not aware of what Wadinko actually does. A few municipalities explained that Wadinko does not invest in their region and therefore they do not recognize the benefits of being a shareholder of Wadinko, except for the dividend payment.
Interview responses showed that the activities of Wadinko are not closely related to the other activities of the local governments and therefore it can be said there is activity diversification. All the interviewees explained that the aim of Wadinko – stimulating employment – is one of the core activities of local governments. However, owning a venture capital company does not belong to the core activities of local governments. A controller of Wadinko described his thinking:
