Abstract This paper studies management control design of supplier relationships in manufacturing, a supply chain phase currently under-explored. Compared to supplier relations during procurement and R&D, which research found to be governed by a combination of formal and informal controls, supplier relations in manufacturing are more formal, so that they could be governed by more formal and less informal controls. To refine the management control system and influencing contingencies, we propose a theoretical framework specifically adapted for the manufacturing stage.
This framework is investigated by an in depth case study of the supplier management control of a Volvo Cars production facility. We identify three types of suppliers visualizing the associations in the framework and illustrating the framework’s explicative power in (automotive) manufacturing. Furthermore, the case contradicts that supplier relations in the manufacturing phase are governed by little informal control, because the automaker highly values the role of trust building and social pressure.
Most notably, a structured supplier team functions as a clan and establishes informal control among participating suppliers, which strengthens the automaker’s control on dyadic supplier relations. Keywords: Management control; Supplier relationships; Manufacturing; Contingency theory; Case research; Automotive 2 1. Introduction In the current economic environment, characterised by globalisation and enhanced levels of competition, companies require an effective supply chain with inter-organizational relationships (IORs) to strive for sustainable competitive advantage.
Not surprisingly, studies show that IORs have a high potential impact on organization performance (e. g. Anderson & Dekker, 2005). Literature, however, also argues that many IORs do not provide the expected benefits and are often terminated because of managing difficulties (Ireland, Hitt & Vaidynanath, 2002). Academics often propose that lack of coordination and opportunistic behaviour of partners are the two main reasons for the relatively high relationship failure rate (e. g. Dekker, 2004).
Hence, management control systems (MCSs) are argued to play a critical role in preventing such failure, by establishing governance mechanisms to control the relationship (Ireland et al. , 2002). The fundamental goal of MCSs is to influence decision making in attaining strategic objectives (Nixon & Burns, 2005). In an inter-organizational setting, this implies creating bilateral incentives to pursue mutual goals. Already in the mid-nineties, scholars started calling for more attention for this topic (e.g. Hopwood, 1996; Otley, 1994), and have not stopped since (e. g. van der Meer-Kooistra & Vosselman, 2006).
Consequently, inter-organisational MCSs have been studied from several angles, including outsourcing (e. g. Anderson, Glenn & Sedatole, 2000), inter-organizational cost management (e. g. Cooper & Slagmulder, 2004), partnerships (e. g. Seal, Berry, Cullen, Dunlop & Ahmed, 1999), strategic alliances (e. g. Dekker 2004), networks (e. g. Kajuter & Kulmala, 2005) and joint ventures (e.g. Kamminga & van der MeerKooistra, 2007).
Yet, the main emphasis was put on relational collaboration during the first stages of the supply chain, namely procurement, which involves the make-or-buy decision, partner selection and contract design, and R&D. Although this historical focus is certainly justified, management control in a later phase of the supply chain, namely manufacturing, remains relatively under-explored (Cooper & Slagmulder, 2004; Langfield-Smith & Smith, 2003).
However, purchased products and services for manufacturing account for more than 60% of the average company’s total costs (Degraeve & Roodhooft, 2001) and are subject to continuous improvement with suppliers, also requiring adequate management control. Therefore, this study illustrates how manufacturers design the MCS of supplier relations in the manufacturing phase of the supply chain, which we refer to as “manufacturer-supplier relationships” (MSRs).
In other words, we abstract from 3 procurement and R&D influences. 1 Nevertheless, management control research on previous supply chain stages, offers a first theoretical insight into how a MCS for MSRs could look like. In particular, prior empirical research on IORs such as R&D collaboration (Cooper & Slagmulder, 2004), strategic alliances (Dekker 2004) and joint ventures (Kamminga & van der Meer-Kooistra, 2007) found MCSs that combine both formal controls, like outcome controls, and more informal controls, such as trust building.
Also the execution of service outsourcing projects, like industrial maintenance (van der Meer-Kooistra & Vosselman, 2000), IT (Langfield-Smith & Smith, 2003) and accounting (Nicholson, Jones & Espenlaub, 2006) is governed by a combined MCS. So if we assume these findings to hold for other IOR types (external validity) and neglect potential characteristic differences, MSRs could be expected to be governed by a combination of formal and informal control as well. Yet, by taking into account differences between MSRs and other types of IORs, the MCS design could be different.
In that respect, we argue that manufacturing is more formal than procurement and R&D. Indications for that argument and its consequences for management control can be found in the management control framework of Das & Teng (2001). Based on the variables in their framework2, task programmability and outcome measurability, it should be clear that for manufacturing both variable levels are high, or at least higher than in the case of procurement and R&D. Consequently, the framework indicates that formal controls are suited mechanisms to govern MSRs.
This argument is strengthened by the type of knowledge usage in MSRs, for which organization literature provides a clear distinction between knowledge exploration and knowledge exploitation. On the one hand, it is argued that the first supply chain phases, think of procurement and R&D, aim at knowledge exploration, while the later stages, like manufacturing, primarily 1 Obviously, procurement and R&D do impact the manufacturing phase.
Yet, as our aim is refining supplier MCS design in the manufacturing phase, we deliberately exclude these influences. In terms of research methodology, this abstraction is put into operation by studying a MSR between a manufacturer facility and supplier facility only dealing with manufacturing, while procurement and R&D are handled by their respective mother companies (cf part three of this paper “research methodology”). 2 Although this framework was originally developed by Ouchi (1979) for use in MCS design within organizations, Das & Teng (2001) further adapted it for use in IORs.
Task programmability refers to the degree to which managers understand the transformation process in which appropriate behaviour is to take place. Outcome measurability refers to the ability to measure outcome precisely and objectively. When outcome measurability is high/low and task programmability is low/high, formal outcome/behaviour control should be set up to govern the relation. When both dimensions are low, informal control is preferable, but when both measures are high, both outcome and behaviour control are suited control mechanisms (Das & Teng, 2001).
4 aim at knowledge exploitation. On the other hand, research shows that the exploration of knowledge is best governed by informal controls, while knowledge exploitation is most adequately controlled by formal controls (Bijlsma-Frankema & Costa, 2005). Thus, based on the characteristics of high task programmability, high outcome measurability and knowledge exploitation goals, MSRs could be expected to be governed by primarily formal controls with little informal controls. In other words, the literature offers different management control designs for MSRs regarding the informal control level.
Therefore, this study investigates how the MCS of MSRs is designed and how important informal controls are in that design, in particular in IORs between an original equipment manufacturer (OEM) and suppliers of outsourced manufacturing activities in the trend-setting automotive industry (cf Womack, Jones & Roos, 1990). An automobile is a complex product manufactured with thousands of components. Consequently, also this industry increasingly outsourced non-core activities and started relying on suppliers to create lower costs.
To that end, a variety of supply chain management practices has been implemented, such as lean supply and continuous improvement. Yet, these induce the need for appropriate management control structures and bi-directional communication to organize and manage the relation (Carr & Ng, 1995; Scannell, Vickery & Droge, 2000). In that respect, one particular automaker, namely Toyota, is known for partnering with suppliers, transferring its expertise to help suppliers and installing softer forms of control including trust.
To govern the search for continuous improvement in manufacturing, Toyota established the “Toyota Group” by means of a supplier association, an operations management consulting division and voluntary small group learning teams (Dyer & Nobeoka, 2000). However, practitioner literature (e. g. Automotive News/Automotive News Europe) describes several other automakers governing this search by heavily formalized supplier relations. Contrary to cooperation during procurement and R&D, manufacturing is argued to become much more demanding towards suppliers.
Automakers increasingly transfer manufacturing risk and supply responsibility to first-tier suppliers, which results in suppliers delivering to very tight just-in-time and in-sequence schedules (Alford, Sackett & Nelder, 2000). As a result, OEMs install formal controls and supplier improvement techniques, which alert suppliers to the importance of ameliorating supply performance at lower costs. Hence, also automotive practice shows evidence of high and low levels of informal control. Therefore, this study specifically investigates how the MCS of automotive MSRs is designed.
Yet, besides illustrating MCS design, this paper contributes to explaining MCS design of automotive 5 MSRs. To our knowledge, little inter-organizational management control research specifically investigated contingency theory’s explicative power in manufacturing. Naturally, several papers study influences on MCS design in production environments, like the impact of manufacturing flexibility (Abernethy & Lillis, 1995), customization and related interdependence (Bouwens & Abernethy, 2000), profit centre strategy (Lillis, 2002), production strategy, production technology and organization (van Veen-Dirks, 2006).
However, these studies investigate characteristics explaining MCS design in one organisation, while our study focuses on inter-organizational relations. To that end, we propose a refined theoretical contingency framework based on recent inter-organizational management control theory, but specifically adapted for the manufacturing stage. This framework proposes several contingencies determining the level of risk, which is governed by different levels of management control techniques.
In order to illustrate the validity of the framework in practice and answer how and why automakers design their MCS, we perform an in depth case study of the relations between a facility (VCG) of the international OEM Volvo Cars and a selection of its first-tier supplier facilities. The case study provides considerable evidence of three supplier types, namely batch, low value-added just-in-sequence and high value-added just-in-sequence suppliers, visualizing the associations in the framework between contingencies, risks and management controls.
These controls include both formal and informal techniques, of which trust building and social pressure are highly valued. Most notably, VCG’s structured supplier team functions as a clan and establishes informal control among participating suppliers, which strengthens control on the OEM’s dyadic supplier relations. As our framework draws on case findings from other less formal IORs, it seems that our case findings offer more evidence of their external validity. That way, the findings contradict that informal controls play a minor role in automotive MSRs.
In particular, VCG’s MCS, combining both formal and informal controls, is argued to be designed specifically to improve supply performance. The remainder of this paper is organized as follows. In the second part, we develop the theoretical contingency framework. The third part describes the case research methodology. The fourth part is the actual case study, which presents VCG, describes three supplier types by means of contingency levels and clarifies how VCG designed the MCS governing them.
In the fifth part, we discuss our findings by comparing VCG’s management control with previous findings and elaborating on the significance of VCG’s supplier team. We conclude the paper with a summary of the main findings and some avenues for further research. 6 2. Theoretical framework In this part, we develop a theoretical contingency framework for MCS design of MSRs, which can be found in figure I. > Contingency theory originated with the aim of explaining the structure of organizations by particular circumstances.
Later, management accounting researchers adopted and further developed the theory in order to explain the shape of MCSs in organizations (e. g. Chenhall, 2003; Luft & Shields, 2003). Therefore, contingency theory suits this study, regarding MCS design of MSRs and its explicative variables. The central concept of the framework is the level of risk a certain MSR runs. Inter-organizational management control theory proposes two types of risk, which result from five different situational antecedents, characterizing the MSR.
Although we clarify both risk types separately, we stress the integrative interpretation of all contingencies jointly determining both levels of risk. Subsequently, this risk is governed by different management control instruments, either with a large or a small role for informal control. 3 2. 1. Performance risk The first risk type is performance risk, defined as the probability of not achieving the MSR objectives, despite satisfactory cooperation (Das & Teng, 2001).
This type of risk is also referred to as “coordination requirements” (Dekker, 2004; Gulati & Singh, 1998) or “the mastery of events” (Tomkins, 2001). As the MSR objective concerns manufacturing as many products of the order book as possible, on time, with good quality at the lowest possible cost, performance risk is the risk of a supply chain interruption disturbing the realisation of this goal. Three contingencies related to technology increase this risk, namely complexity, task uncertainty and task interdependence (Chenhall, 2003).
Yet as complexity and task uncertainty are highly related (Chenhall, 2003), the framework does not include complexity separately (cf Dekker, 2004). 3 According to van Veen-Dirks (2006), all situational characteristics and MCS characteristics are determined jointly instead of sequentially. Also Kamminga & van der Meer-Kooistra (2007) propose that the influence of contingencies is not determined by each antecedent as such, but by their interaction. In addition, they suggest studying control as an integrative concept, in which all control dimensions are incorporated.
Consequently, we do not propose one-on-one associations between one specific contingency, one specific type of risk and one specific type of control, suggested to suit that risk type. Instead, our model simultaneously studies the associations between situational contingencies, risks and management control techniques, as put forward by the three boxes of figure I. The boxes of contingencies and risks are put together to stress their interdependence and joint impact on management control.
7 Task uncertainty relates to variability in transformation tasks and the available knowledge of methods for performing those tasks (Chenhall, 2003). This situational characteristic determines the measurability difficulty of output and activities (Kamminga & van der Meer-Kooistra, 2007; van der MeerKooistra & Vosselman, 2000), which increases with increasing levels of complexity of both the delivered product and its operational processes (Woodward, 1965).
The first complexity is related to the added value of the product and gradually increases depending on whether the supplier delivers a standard component or an important customized module (Cooper & Slagmulder, 2004). The second complexity regards the added value of the production process and reflects the complexity of the supplier’s manufacturing processes needed to effectively produce and deliver products as required. Task interdependence refers to the degree to which subactivities of the value creation process have been split up and made dependent on each other (Dekker, 2004).
In MSRs, this interdependence is sequential (Thompson, 1967)4, because the relation involves transferring the supplier’s output to the manufacturer’s input process. The level of sequential interdependence is impacted by the dependence level of the manufacturer’s operational performance on the supply quality (timeliness and product quality). Moreover, the interdependence level of a specific MSR is influenced by the production flexibility required from both parties and the manufacturer’s lack of precise knowledge to perform activities previously done in-house.
2. 2. Relational risk The second type of risk is relational risk, implying the probability of not having satisfactory cooperation because of opportunistic behaviour of the supplier, exemplified in shirking, cheating, distorting information and appropriating resources (Das and Teng, 2001). This type of risk is also referred to as “appropriation concerns” (Dekker, 2004; Gulati & Singh, 1998) or “the generation of trust” (Tomkins, 2001).
Transaction cost economics (TCE) theory5 proposes three contingencies that influence relational risk and subsequently determine appropriate control: asset specificity, environmental uncertainty and transaction frequency (Williamson, 1979). Yet, as the manufacturer possesses no specific assets related to a certain supplier, at 4 Thompson (1967) identifies three levels of task interdependence from low to high, which influence the level of inter-organisational coordination and communication: pooled, sequential and reciprocal interdependence.
5 TCE argues that parties are only boundedly rational and behave opportunistically. Therefore, the total cost of outsourcing is the sum of both the supplied component costs and the transaction costs, including costs for negotiation, drawing up contracts, coordination, control and risk of opportunistic behaviour (van der Meer-Kooistra & Vosselman, 2000). 8 least not in the manufacturing phase of the supply chain, there is no lock-in to supplier opportunistic behaviour.
6 Hence, unlike uncertainty and transaction frequency, asset specificity does not influence supplier opportunistic behaviour in MSRs and is not included in our theoretical framework. Consistent with being a central contingency research concept, environmental uncertainty also forms a powerful characteristic of MSRs (Chenhall, 2003). In particular, this contingency relates to general market uncertainties and uncertainty about unknown future contingencies (Kamminga & van der Meer-Kooistra, 2007; Langfield-Smith & Smith, 2003; van der Meer-Kooistra & Vosselman, 2000).
Because manufacturer and supplier interact under these uncertainties, both parties face changes over time, which require detailed contracts (Dekker, 2004). However, incomplete contract theory argues that there exist limitations in drawing up complete contracts, because all future contingencies can not be foreseen, are too expensive to foresee or are too expensive or impossible to contract upon (Gietzmann, 1996). Consequently, the combination of uncertainty and incomplete contracts leads to potential opportunistic behaviour of the supplier.
According to TCE, more frequent interactions lower the possibility of opportunistic behaviour (Williamson, 1979). So, to preserve a positive relation between contingencies and relational risk, we could utilize infrequency as contingency variable (e. g. Anderson & Dekker, 2005). Yet, as we study MSRs with no connection to commercial negotiations determining the contract term, we include the antecedent relational stability aim. This contingency relates to the manufacturer’s aim of continued future interactions with the supplier and serves to build bilateral commitment (Cooper & Slagmulder, 2004).
We argue that MSRs, in which relational stability is considered necessary and thus aspired by the manufacturer, are subject to higher relational risk. For example, if supplier switching costs are high due to high interdependence, high commitment from the manufacturer could incite the supplier to accept lower quality or delivery performance. Besides including a transaction environment characteristic and a transaction characteristic, we also incorporate a transaction party characteristic (Langfield-Smith & Smith, 2003; van der Meer-Kooistra & Vosselman, 2000).
In particular, we include supplier knowledge importance, which encompasses the degree of importance for the manufacturer to know the supplier and to be able to assess characteristics, such as management competence, trustworthiness and willingness to share proprietary knowledge. Usually, this kind of assessment is done by means of first-hand or second-hand experience. Hence, we argue that when the 6 Obviously, suppliers do have specific assets in place, rendering them vulnerable to opportunistic behaviour from the part of the manufacturer.
However, this study and the developed theoretical framework only focus on supplier opportunistic behaviour. 9 importance of supplier knowledge rises, the risk for insufficient or erroneous assessment and subsequent supplier opportunistic behaviour increases. 2. 3. Management control system Although MCSs have been conceptualised and categorised in various ways, the current management control literature has reached a consensus on two types of management controls, namely formal and informal control instruments (Langfield-Smith & Smith, 2003).
Obviously, studying the usage of informal controls compared to formal controls requires both control types to be included in the theoretical framework. Formal controls are explicitly set up to coordinate the MSR and include outcome controls and behaviour controls. Outcome control involves the measurement and evaluation of the outcomes of operations against pre-defined outcomes or targets, by using several performance measurement techniques (Ouchi, 1979; Dekker, 2004). The most important outcome metrics for MSRs are percentage of defects, quality of delivered goods and on time delivery of goods (Gunasekaran, Patel & McGaughey, 2004).
Behavioural control concerns the specification and actual surveillance of behaviour, by means of rules and standard procedures (Ouchi, 1979). Additionally, behaviour control includes evaluating compliance with pre-specified planning, procedures, rules and regulations (Dekker, 2004). Informal controls (also called social controls) are not explicitly designed, but are grown out of shared norms and values, shaped by frequent interaction, meetings and management attitude (Ouchi, 1979; Merchant, 1998). Especially trust building7 has emerged as a very important informal control instrument in inter-organizational MCSs (e. g.Dekker, 2004).
While formal controls reduce the risk by altering the incentives for underperformance and opportunistic behaviour, trust mitigates risk by minimizing the fear of underperformance and opportunistic behaviour to occur (Das and Teng 2001). Therefore, we include three types of inter-organizational trust building, namely building contractual trust, competence trust and goodwill trust (Sako, 1992). 8 Contractual trust results from previous contractual relations or grows during the MSR 7 Rousseau, Sitkin, Burt & Camerer (1998, p. 394).
Define trust as “a psychological state comprising the intention to accept vulnerability, based upon positive expectations of the intentions or behaviour of another”. According to them “trust is not a behaviour (cooperation), or a choice (e. g. taking a risk), but an underlying psychological condition that can cause or result from such actions” (Rousseau et al. , 1998, p. 395; italics added).
As such, trust in itself can not be a control instrument in the MCS of MSRs. Instead, the control techniques are the actions the manufacturer performs to create and build trust in the supplier. 8 Contractual trust is based on the expectation that the supplier will keep promises and comply with agreements made, whether these10 (Sako, 1992).
Competence trust is increased by previous good performance, i. e. good quality and delivery results. Moreover, competence trust results from buying activities from reputable suppliers or transferring competences to the supplier. Additionally, product and/or process certification and process standardisation enhance competence trust (Sako, 1992). To develop goodwill trust, Sako (1992) identifies shared values and norms as necessary, but insufficient, as transaction parties also need to show the willingness to be indebted to each other.
Gulati (1995) stresses creating and growing an inter-organizational bond of friendship to trigger goodwill trust (Gulati, 1995). Other possible goodwill trust initiators are interactive goal setting, trustworthiness reputation and a long term relationship (Dekker, 2004). Next to these specific trust building mechanisms, the literature also proposes an important overall trust building technique, namely close interaction, based on mutual interests and established by means of joint decision making and joint problem solving via a joint relationship board and/or joint task groups (Das & Teng, 2001; Dekker, 2004).
9 Besides trust building, MSRs can be governed by another type of informal control, which Ouchi (1979) refers to as clan control. Based on shared norms, values and a common inter-organizational goal, supplier behaviour in the interest of the MSR will be reinforced, because suppliers are motivated to achieve the goal (Das & Teng, 2001). This incentive results from inter-organisational social pressure (Spekle, 2001) exerted by the manufacturer, which we believe is social control in its literal meaning.
Because of high interdependence between manufacturer and supplier, below standard results of the supplier directly impact the manufacturer’s performance. Consequently, supplier management is unpleasantly confronted with manufacturer management and faces personal humiliation because of the error. Additionally, supplier management runs the risk of their reputation and personal relationship with interacting manufacturer management getting injured. Also Dyer & Singh (1998) mention reputation and personal relations as social control mechanisms, besides norms and trust.
By acting as negatively valued social sanctions (Bijlsma- are contractually stipulated or not. Competence trust concerns the expectation that the supplier possesses the necessary technical and managerial competences to deliver the order as agreed. Goodwill trust regards the expectation that the supplier shares an open commitment, with the willingness to perform activities beneficial to the MSR, but possibly neither in the supplier’s interest nor required by the contract (Sako, 1992). 9 Other potential overall trust building techniques in a MSR are communication via regular inter-organizational meetings (Chalos &
O’Connor, 2004; Das & Teng, 2001), information sharing of problem areas (Chalos & O’Connor, 2004), supplier development activities (Carr & Ng, 1995), networking (Das & Teng, 2001), training (Chalos & O’Connor, 2004) and the extent to which the employees of both parties understand the factors ensuring the collaboration’s future success (Chalos & O’Connor, 2004). 11 Frankema & Costa, 2005), these social consequences create incentives for satisfactory supplier performance and render supplier opportunism hard to sustain (Spekle, 2001).
If we assume operational snags to be day-today business in MSRs, this social pressure creates an informal means to mitigate risk in MSRs. 3. Research methodology 3. 1. Case study research The empirical part of this paper is based on an in depth case study, which is an investigation of a real life phenomenon, relying on multiple sources of evidence and benefiting from prior development of theoretical propositions (Yin, 1994). This research method suits our research that concerns refining existing interorganizational management control theory for the relatively under-explored manufacturing phase of the supply chain.
10 According to Keating (1995), such theory refinement needs a clear theoretical starting point, supplemented with openness to the discovery of unexpected findings. To balance these theory attachment and detachment requirements, we developed a theoretical framework to guide the data collection, but at the same time used data collection techniques allowing sufficient openness. Furthermore, several interorganizational management control case studies (e. g. Cooper & Slagmulder, 2004; Dekker, 2004; Kamminga & van der Meer-Kooistra, 2007;
Nicholson et al. , 2006) strengthen the argument that cases allow investigating in detail the structure and influencing variables of IORs (Sartorius & Kirsten, 2005). These studies show that theory refinement of MCS design can be adequately investigated by means of qualitative research. The social meaning of inter-organizational MCSs, especially regarding the use and interpretation of informal controls, and the subsequent behaviour of companies and employees is very complex.
So if we only skim the surface, we will never discover how different parties interpret certain IORs and whether the MCS is designed accordingly. This argument not only justifies the choice for a case study, but also forms the reason 10 Our research corresponds to investigating a complex phenomenon within its real life context of which empirical evidence is rather limited, and answering how and why questions about this phenomenon, for which case study research is most suited (Eisenhardt, 1989; Yin, 1994).
Furthermore, Keating (1995) argues that case studies suit three goals and that our theory refinement goal represents the middle ground between theory discovery (describing novel phenomena) and theory refutation (disconfirming well specified theories by bringing in negative evidence). More specifically, our case research is of the theory illustration type, documenting “previously unappreciated aspects of management accounting practice” and identifying “aspects of the illustrated theory that require reformulation or more rigorous specification” (Keating, 1995, p.71).
Indeed, the goal of this study is to illustrate how manufacturers design supplier MCSs, to what extent this design differs from designs in other IORs and how the design can be explained by means of a specifically adapted theoretical framework. 12 why more of this research is requested (e. g. Langfield-Smith & Smith, 2003; Dekker, 2004; van der MeerKooistra & Vosselman, 2006). 3. 2. Unit of analysis In most inter-organizational studies, the unit of analysis is one dyadic relation between two independent parties (van der Meer-Kooistra & Vosselman, 2006).
Since there exist different dyadic MSRs within one manufacturer and we study MCS’s dependence on relationship contingencies, our unit of analysis consists of specific MSRs. Dyer & Singh (1998) explicitly propose the “relational view”, focusing on the buyer-supplier dyad, as opposed to the “industry structure view” and “resource based view”, when analyzing cooperative strategy and sources of inter-organizational competitive advantage. In order to answer the proposed research questions concerning MSR MCS design, we analyzed all relations after the manufacturer had decided to outsource the manufacturing activities.
In other words, we addressed neither the make-or-buy decision nor related commercial negotiations, but collected data from the start of production onwards. Furthermore, we only gathered data on standard MCSs for MSRs with good operational performance. 3. 3. Case company selection The selection of the case company and its suppliers was influenced by two selection concerns: theoretical sampling (Eisenhardt, 1989), and open and flexible access to.