Strategic Management Of Supplier Manufacturer Relations

Research Policy 31 (2002) 159–182

Strategic management of supplier–manufacturer relations in new product development Maurizio Sobrero a,∗ , Edward B. Roberts b,1 a

b

Department of Management, University of Bologna, Piazza Scaravilli 2, 40125 Bologna, Italy Alfred P. Sloan School of Management, Massachusetts Institute of Technology, E52-535, 50 Memorial Drive, Cambridge, MA 02139, USA Received 7 August 2000; received in revised form 6 November 2000; accepted 1 December 2000

Abstract This paper discusses the role of contractual and organizational arrangements for the governance of supplier–manufacturer relationships in new product development projects. We present cross-sectional project level data from 50 manufacturer–supplier relations in new product development in the European Major Home Appliance industry using a single dyadic interaction as the unit of analysis, to combine institutional decisions driving the governance of inter-firm relations and their organizational implementation. Our results show that (a) relational outcome depends on the type of joint activities, (b) it can be decomposed into short term (efficiency) and long term (learning) effects, (c) less articulate types of joint development activities increase efficiency, while more articulate joint development activities increase partner’s learning, and (d) inter-organizational structuring decisions are significant explanatory variables to understand relational outcome. Implications for the organization of vertical collaboration in new product development projects are derived focusing on the emerging trade-off between short and long term objectives as a central issue in guiding relational strategies. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Innovation; Inter-organizational relationship; New product development; Supply chain management

1. Introduction The attention paid to inter-firm relations has been constantly increasing during the past 15 years. Research has in turn emphasized the strategic appropriateness (Roberts and Berry, 1985; Porter and Baden Fuller, 1987; Harrigan, 1988; Lorenzoni and BadenFuller, 1995), the economic advantages (Williamson, 1979; Contractor and Lorange, 1988; Kogut, 1988), and the learning implications of inter-organizational ∗ Corresponding author. Tel.: +39-051-2098076; fax: +39-051-2098074. E-mail addresses: [email protected] (M. Sobrero), [email protected] (E.B. Roberts). 1 Tel.: +1-617-253-4934; fax: +1-617-253-2660.

relations (Hamel, 1991; Kogut and Zander, 1992; Gulati, 1998). The central argument of this growing literature is that individual organizations can no longer rely on their own resources to compete in today’s world. Rather, they should look for strategic interactions allowing them effectively to leverage internal resources by investing in some core competencies and contracting out other knowledge domains. The analysis of innovation development processes is frequently pointed out as a prototypical example for the need to rethink organizational boundaries (Arora and Gambardella, 1991; Freeman, 1991). The speed of technological progress, the increasing amount of resources needed to control fully the process and the shortening times for investment recovery are common phenomena in many industries. Changes in the drug

0048-7333/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 0 4 8 - 7 3 3 3 ( 0 0 ) 0 0 1 5 7 - 8

160

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

development process (Henderson, 1994), as well as in cars (Clark and Fujimoto, 1991; Nobeoka, 1993), mainframes (Iansiti, 1995; Iansiti and Khanna, 1995), and personal computers (Pine, 1993) are often used as examples. While the key aspect of networking processes is identified in the organization of the interaction among the partners, the empirical analysis is largely focused on the structural context of exchange. Both overdetermined and underdetermined approaches to the study of joint product development activities examine the mutual exchange of rights between the parties involved. On the one hand, the agents involved are modelled as rationally optimizing the contractual form of exchange in a typical market/hierarchy continuum. Different characteristics of the development work and of the partners’ resource set guide the choice among alternative forms such as long term contracts (Monteverde and Teece, 1982; Ouchi and Bolton, 1988; Pisano, 1991), strategic alliances (Roberts and Berry, 1985; Ciborra, 1991; Hamel, 1991), or R&D consortia (Ouchi and Bolton, 1988; Watkins, 1991; Tripsas et al., 1995). On the other hand, higher level institutional conditions idiosyncratic to the locus of interaction determine the predominant form of exchange and the set of norms and rules governing the relationship. Several studies performed in Japan (Dore, 1986; Fruin, 1992; Nishiguchi, 1995) and in localized communities such as Silicon Valley (Saxenian, 1994) or the European industrial districts (Piore and Sabel, 1984; Best, 1990; Harrison, 1994) describe the reliance on trust among the partners as a substitute for more stringent legal protection mechanisms. This paper goes into the “black box” of inter-partner exchanges in innovative activities to develop theoretically and document empirically the complementary role of the organization of the interface. Building on some recent studies at the information processing level of inter-organizational relationships in new product development activities (Clark, 1989; Liker et al., 1995; Bidault, 1998; Sobrero and Roberts, 2001) we analyze 50 manufacturer–supplier relations in new product development projects performed by three different business units, operating in two different countries, of one of the largest European Major Home Appliance manufacturers. We distinguish between the legal definition of the exchange and the operational mechanisms used to integrate the flow of information

between the parties during the development work. Data show invariance in the specificity and extension of the contracts governing the relationships. On the contrary, differences in the type of mechanisms used to exchange information and the extent of their use during the project are significantly related to the degree of supplier involvement. Choices along integration dimensions are associated with different evaluations by the manufacturer of the specific relationship. Less articulate solutions with later involvement of suppliers and their use as pure subcontractors of pre-specified development work are considered by the manufacturer as positively affecting project efficiency, but not as opportunities for learning. On the contrary, higher level of communication, earlier supplier involvement in the project and a higher level of design responsibility assigned to the supplier seem to generate higher coordination costs, but also provide the opportunity for effectively accessing external knowledge whose consequences are spread in future projects. This research offers several theoretical and policy contributions to the understanding and management of inter-organizational relationships in innovative activities. First, institutional decisions driving the governance of inter-firm relations and their organizational implementation are recomposed. The contractual perspective of Transaction Cost Economics is complemented with the information processing interpretation of organizational exchange. Second, the operational mechanisms available to structure inter-unit flows are extended to inter-organizational relations and related to the nature of the joint work, using the dyad as the unit of analysis. Third, the distinctions among different relational outcomes point to the trade-off between short and long term objectives as a central issue in guiding relational strategies. From a strategy perspective, these results suggest the centrality of a planned effort on the organizational implementation of inter-organizational relations, which should synchronically complement contractual negotiations, rather than sequentially supplement them. In addition to that, the empirical results directly suggest a careful examination of partners’ objectives to determine the task-domain in which to interact and how to organize the joint work. Despite all the emphasis on the benefits of networking, in fact, inter-organizational relations are not free, and a rush toward the currently hot boundary spanning strategy

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

might prove costly, especially when technological innovation is at stake. The paper is organized as follows. In Section 2, we review the different approaches to inter-organizational relationships in technology development and the information processing view of the organization of innovative work, deriving the complementarity of the contractual and organizational structuring of the relationship. In Section 3, we introduce the study designed to examine these complementarities in the context of supplier–manufacturer relationships in new product development, describing the methodology used and the industrial setting chosen. In Section 4, we present the results and discuss their implications for the management of supplier relationships in new product development projects. In Section 5, we conclude by interpreting these results for the more general problem of the organization of information flows in inter-organizational relationships, discussing the limits and potentiality for further extension of this work.

2. Theoretical premises 2.1. Inter-firm relationships and innovative processes The empirical observation of the increasing importance of interactions among legally distinguished units challenged the traditional image of an organization as an entity with well-defined boundaries (Contractor and Lorange, 1988). Clearly, one crucial activity influenced by these changes is the development of innovation (Roberts, 1980). Following Kogut (1988), we can identify three theoretical approaches for the analysis of inter-firm relations: a transaction–cost based approach, a strategic approach and an organizational learning approach. Studies based on the assumptions and propositions of transaction cost theory (Williamson, 1975, 1985) focus on the characteristics of the interactions involved in the development of innovation (Ouchi and Bolton, 1988). Given actors’ characteristics (propensity to opportunistic behavior and bounded rationality), task related factors (asset specificity and uncertainty) are used to identify the appropriateness to engage in interactions with other subjects and to determine the optimal governance structure (Pisano, 1990). Accordingly, research has been addressing the

161

conditions under which the sum of production and transaction costs will be minimized through internalization (Pisano, 1991), market driven transactions such as licensing, or other intermediate forms of governance structure such as joint ventures (Hladik, 1988) and consortia (Watkins, 1991; Tripsas et al., 1995). Complementary to the inward focus on efficiency achievements, the strategic perspective on organizational interactions stresses how, by pooling their resources and efforts towards innovation, all the partners involved would improve their competitive positioning (Gulati, 1998). Collaborative activities become a way to overcome asymmetries in the resource endowments and spread the risks associated with innovative activity (Combs, 1990), to control spillovers (Katz, 1986), or to limit the opportunities for new entrants in the market by indirectly raising barriers to entry (Katz, 1986; Katz and Ordover, 1990). Once the determinants for collaboration in innovation processes are set, the possible alternatives (e.g. joint ventures, acquisitions, licensing) are examined through an analysis of either environment specific conditions such as, for example, the “appropriability regime” (Teece, 1986), or the firm resource set (Roberts and Berry, 1985). The organizational learning explanation departs from the economic nature of the determinants of a positive role of inter-organizational relationships in the development of innovation (Kogut, 1988). In particular, it claims that R&D activities are “interacting heuristic search processes” (Nelson and Winter, 1977, p. 52) and that the very knowledge being developed is organizationally embedded. Therefore, its transfer cannot be achieved through a market transaction because it is partly tacit (Polanyi, 1967). However, organizations cannot rely simply on their own internal knowledge base since the routines developed around it become embedded, constraining departures from well-known and explored paths (Nelson and Winter, 1982) and acting as filters to the introduction of innovation (Henderson and Clark, 1990). On the contrary, by widening the sources of knowledge to include external organizations, resistance created by internal structural filters would be overtaken (Kogut and Zander, 1992). Moreover, the use of collaborative arrangements allowing for mutual access to internal processes will facilitate both the development and the transfer of tacit knowledge (Hamel, 1991; Gulati, 1998).

162

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

Regardless of the differences in the approach to the analysis of inter-organizational relations, the three streams of research mainly focus on the conditions under which it would be appropriate for firms to engage in cooperative actions to pursue innovation. Much less attention, on the contrary, is given to the role that alternative organizational forms of coordination and integration among the partners might have in determining the outcome of the relation (Roberts and Berry, 1985). The emphasis is on the definition of the conditions for the interaction to occur, rather than on how the relationship is then managed and the actual contacts among the partners are operationally implemented. To emphasize the necessary complementarity of such aspects required to make the relationship work, in the next section, we discuss the role of organizational design in innovative activity. 2.2. Organization structure and innovative processes Albeit almost neglected in the particular context of inter-organizational relations, the importance of organizational structures and processes and their interaction with task characteristics are strongly rooted in the organizational theory literature. Starting from the seminal works of Burns and Stalker (1961), Woodward (1965), and Lawrence and Lorsch (1967), the contingency school developed both at the macro level, exploring the concept of fit between organizational and environment characteristics (Aldrich, 1972; Duncan, 1972, 1976), and at the micro level, investigating the relationship between nature of the task and structural characteristics (Perrow, 1967; Allen, 1986; Larson and Gobeli, 1988). Burns and Stalker (1961) first proposed that an organization’s characteristics are a function of environmental uncertainty. Lawrence and Lorsch (1967) continued in this direction by showing how environmental conditions, and hence organizational structure, could vary within the same organization among its different units. Organizational design therefore became central, and detailed work was conducted in determining in more depth the effectiveness of different organizational forms (Galbraith, 1973, 1977). Innovation development activities, i.e. activities characterized by ill-defined tasks in uncertain environments, started to attract researchers’ attention. The unit of analysis was therefore shifted to a lower level,

usually the single project, focusing on the differences in effectiveness of alternative solutions in the functional/project spectrum (Marquis and Straight, 1965; Larson and Gobeli, 1988; Roberts, 1988), the influence of patterns of communication on the final outcome (Marquis and Straight, 1965; Allen, 1977; Tushman, 1978; Allen and Katz, 1986; Allen and Hauptman, 1987), and the role of human resource management practices (Katz and Allen, 1985; Allen and Katz, 1990). Only recently have researchers turned their attention to the analysis of the different organizational coordination mechanisms in inter-organizational relationships. At a macro level, for example, Rebentish and Ferretti (1995), studying technology transfer within international joint ventures in the chemical and steel industry, develop a model where the characteristics of the object being transferred (transfer scope) and the characteristics of the organizational structure used to achieve the transfer (transfer method) concur in determining the success of the transfer. Similar arguments are developed by (Hakansson, 1987, 1989) who provide evidence through different case studies about how the strengths of certain Swedish industries are based on specific investments in the organizational coordination of a nexus of relationships at the industry level. Imai and colleagues’ (1985) work on supplier– manufacturer relationships in new product development presents evidence from five in-depth case studies of Japanese firms. Their research covers different industries, from electronics (Epson dot-matrix printer, NEC PC8000, Fuji FX-3500 paper copier) to photographic equipment (Canon Auto-Boy lens shutter camera), to automobiles (Honda City). Suppliers’ early involvement in the development process and the intense patterns of communication flows are indicated as driving forces for faster releases of new devices and prompt responses to competitors’ moves. Similar conclusions are supported by subsequent research in the world auto industry (Clark, 1989; Clark and Fujimoto, 1991), showing lower costs and shorter lead-time associated with higher levels of supplier involvement and higher use of off-the-shelf parts, controlling for industry, project and geographical level characteristics. Overall, greater supplier involvement accounts for roughly one-third of the Japanese advantage in engineering hours and 4–5 months of their lead time advantage.

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

More detailed analyses of the micro-organizational mechanisms used to coordinate suppliers’ participation in new product development extend these results. In two other studies (Sobrero and Toulan, 2000; Sobrero and Roberts, 2001) of supplier–manufacturer relationships in new product development projects, we showed that the type of knowledge being partitioned and its level of interdependency with the rest of the project are important predictors of performance outcomes of the relationship, controlling for contractual differences. Further, the analyses demonstrate a clear trade-off between short term efficiency-increasing and longer term learning-enhancing outcomes. While firm (Dyer and Ouchi, 1993; Dyer, 1996; Nishiguchi and Ikeda, 1996) and country level (Helper, 1991; Liker et al., 1995; Nishiguchi, 1995) differences seem to be time-dependent and to show convergent patterns, all these works independently support the importance of adequate investments in the organizational structuring of exchange flows according to the characteristics of the tasks jointly performed. Choices along the frequency of communication between the partners, the type of communication channels used, the timing of partners’ involvements in the problem solving process and the assignment of specific design responsibility must be accurately balanced to have a positive impact on project outcomes. 2.3. The research framework These two different research approaches to the analysis of inter-organizational coordination mechanisms stress separate albeit complementary aspects of the same problem (Sobrero and Schrader, 1998). On the one hand, strategic approaches examine the institutional governance mechanisms necessary to define the rule set agreed upon by the partners. The focus is on the mutual exchange of rights among the parties involved. On the other hand, organizational approaches highlight the informational characteristics of the interaction. The focus is on the possible alternatives to put into place the mutual exchange of information among the partners. The choice along both dimensions is related to the characteristics of the tasks performed in the relationship. Combining the different perspectives reviewed, the task domain can be articulated focusing on three aspects: asset specificity, means uncertainty, and goal

163

uncertainty. Asset specificity determines the extent to which the activities performed in the relationships have some economic value per se or not (Williamson, 1979). The higher the asset specificity, the lower the chances for the partners to benefit from their activities outside the relationship. The level of task uncertainty can be referred to the action or to the goal domain (Thompson, 1967). In the first case, the partners have agreed on the objective of their relationships but have multiple options to achieve the goal. Ring and Rands (1989), for example, document how NASA and 3M were able to specify up front their goal for common projects on microgravity experiments and subsequently worked through the implementation of their collaboration. In the second case, the goal itself is unclear. These situations have been documented in many studies of collaborative R&D projects, when the partners might initiate the relationship for some generic strategic reason but lack an operational objective (Brockhoff, 1992; Kleinknecht and Reijen, 1992; Tripsas et al., 1995). Given different characteristics of the tasks to be performed within the relationship, the partners can structure the interaction by (a) articulating the legal conditions governing the transaction and (b) identifying the mechanisms to transfer information among them to implement the transaction. Contractual coordination mechanisms are used to define the legal boundaries of the relationships. They involve the choice of the legal form governing the agreement (e.g. joint venture or strategic alliance, equity and nonequity joint venture), the length of the agreement, the extent to which partners are bound to the agreement (e.g. exclusivity clauses, penalties) and the degree to which these choices are specific to the agreement or not (e.g. standardized versus personalized contracts). Research shows the impact on relationship performance of these choices. Studying strategic alliances, for example, Parkhe (1993) documents how the definition of contractual safeguards positively affected the fulfillment of partners’ strategic objectives. Similarly, in a sample of Scandinavian collaborative, R&D ventures Hakanson (1993) finds that a clear contractual definition of ex-ante safeguards decreases the likelihood of anticipated termination of the venture. Information transferring mechanisms are used as an alternative way to put into place the relationships within the institutional boundaries defined by

164

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

means of the contractual coordination mechanisms chosen. They are targeted to the structuring of information flows between the partners. Decisions on the frequency, timing, directionality of information flows, as well as the means through which these flows occur identify the operational dimensions of procedural coordination mechanisms. For a given institutional setting defined by a specific combination of contractual coordination mechanisms, alternative choices of information transferring coordination mechanisms are going to impact the outcome of the relationship differently. Carter and Miller (1989), for example, show how frequent and bilateral communication between vendors and buyers sensibly limited the occurrence of quality problems in the materials delivered. Dyer (1996) reports that higher levels of face to face communications between suppliers and manufacturers contribute to reduce new model cycle time. Hakanson (1993) documents the positive impact of the coordination procedures used by the partners on the survival of collaborative R&D ventures, provided that managerial autonomy is granted by avoiding detailed specification of implementation procedures within the contractual arrangement. To examine fully how partners should organize their relationships to achieve the intended goals, both

dimensions must therefore be considered (see Fig. 1). First, choices on both dimensions are linked to the characteristics of the tasks jointly performed. Second, they both independently account for the performance of the relationship. Third, their combination might in and of itself be a strategic lever to be used by the partners. In the next sections, we present a study of supplier–manufacturer relationships designed to include in the analysis all these aspects, to deepen our understanding of the implications of structuring decisions in innovation processes.

3. The study We studied the whole population of 50 different supplier–manufacturer relationships for the development of new components or new equipment in three development projects performed between 1993 and 1995 by three business units (SBUs) of one of the five largest European groups in the Major Home Appliances industry. The existence of an established body of research on supplier–manufacturer relationships provides a valuable opportunity to anchor the empirical analysis to a policy domain where the strategic relevance of the problem examined can be easily

Fig. 1. The theoretical framework.

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

appreciated. The Major Home Appliance industry is characterized by complex assembled products made of parts and subsystems requiring the combination of different knowledge domains, from mechanics to electronics to plastic molding. The evolutionary patterns of this industry in Europe show how cyclical interactions between product and process innovation have converged towards highly fragmented and temporarily unstable patterns of consumption, generating at the same time the constraints of large scale manufacturing and the opportunities for creative product differentiation (Sobrero, 1995). Firms’ successful responses to these challenges have been a constant shaping of internal structures around the technological base. Failures of these strategies resulted in exits at different times (Goshal and Haspeslagh, 1990). The contemporary need for size and scope has not been achieved merely by functional or geographical separation, but rather fostered through increased inter-unit coordination (Baden-Fuller and Stopford, 1991). Specific activities, such as product development, where different and frequently contradictory task characteristics are conveyed, became the institutional environment to shape the internal and external structure of the organization and to adapt it to the changed nature of the external environment. Within this context, suppliers are being called upon to play an ever increasingly important role. The strategic relevance of product–process interaction, the shortening product life cycles, and the need to balance volumes with product differentiation require home appliance manufacturers to focus on processes and product architecture knowledge. Such knowledge, though, cannot successfully be exploited without also utilizing components-related knowledge. External sources are used more and more to complement internal competencies, subcontracting the control of parts-related knowledge domains. The internal reorganization is hence accompanied by the need to appropriately structure the supplier chain within new product development activities. 3.1. Methodology The study uses a multi-case and multi-unit of analysis approach to explore if (a) the characteristics of the object of the relationships affect structuring decisions,

165

and if (b) the relationship structuring decisions affect the relationships outcomes. This design is particularly advantageous to approach these research questions (Yin, 1994) because (1) the phenomena investigated are recent or even contemporary, (2) the phenomena investigated are analyzed within a specific context, (3) the research is focused on “how” and “why” aspects, and (4) the researcher has limited control over the phenomena. Multiple sources of information are used to avoid typical potential sources of bias deriving from organizational role, individual’s memory failure, protection mechanisms and the like. The a priori identification of multiple units of analysis allows the structured collection of a larger amount of information. Finally, the stratification of the units of analysis and sources of information allows mapping the within-case sources of variance and determines a base of comparison for multi-case analysis. Specific actions were taken to avoid threats to validity (Yin, 1994; Stake, 1995). To control for the construct validity of the results, we triangulated different sources of information and confronted the fit between the analytical categories selected and the data representation derived. This process is also suggested to control for internal validity biases, as introducing possible alternative sources of variance and forcing the analysis to resolve incongruencies between sources. To further strengthen the internal validity of the study, data were collected at three different levels of analysis: the business unit, the project, and the relationship between a supplier and the SBU manufacturer. Multiple cases were selected to strengthen the external validity of the results. It is important, however, to define properly the generalizability domain (Eisenhardt, 1989). Within case-studies, in fact, generalizability refers to the analytical level and not, as is more usual, to the statistical level. In the first case, the discriminant criterion lies in the adherence of the selected observation to the logical categories defined up front, and it suffices to define certain dimensions and find matches between them and the selected cases. In the second case, instead, representativeness is defined by the attributional properties of the sample with respect to the larger population of which it is a part, which has to be defined up front to devise the appropriate sampling strategy. The goal here was to achieve analytical representativeness of

166

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

inter-organizational relationships in the development of innovation. 3.2. Case selection Three separate SBUs of one of the five largest European Major Home Appliance groups agreed to participate in the study. The group has facilities located all around Europe and operates in different countries both with multiple brands and as the OEM for large retailers. It produces the whole product line of major home appliances, from refrigerators (“cold” line) to washing machines (“wet” line) and kitchen stoves and ovens (“hot” line). The SBUs selected operate in two different countries. Two SBUs produce and develop different ranges of products in the “wet” line and the third one develops and produces the whole “cold” line. All of them produce for all group brands. Within each unit, we selected one new product development project with the aid of top managers, namely the chief executive officer (CEO), the group managing director and the group R&D director. Top managers were considered the most appropriate informants because of their overall knowledge of the

project portfolio, the impact of each project on the SBU and the group marketing strategy, and an indirect involvement in the project. The selection process was guided by the following three criteria (see also Table 1). 3.2.1. Degree of innovativeness Focusing on a new product development project, the first necessary step is to define the extent to which the project represents major or minor departures from the SBU’s existing competencies. Innovativeness is therefore defined here as change with respect to the established knowledge base of the observed unit (Henderson and Clark, 1990). The knowledge base itself is articulated into a technical and market knowledge base (Abernathy and Clark, 1985; Roberts and Berry, 1985). Working with the top management, three projects were selected to span across the innovativeness domain, assessing the extent to which the specific project represented a departure from existing market and technological competencies at the given SBU level. One case is about a radical departure from existing competencies, another one is about incremental departure, and the third one lies somewhere in the middle.

Table 1 Informants, selection criteria and rationale for the identification of product development projects Informant

Selection criteria

Rationale

Top management

Degree of internal market and technological innovativeness (one radical, one incremental and one in between the two)

To partition ex-ante the projects within homogeneous classes To identify an internal benchmark with respect to an external one in order to circumstantiate change within a specific reality To map the widest possible range of innovative projects To recognize the complementary nature of market and technological newness

Recency (finished not more than 1 year prior to beginning of research)

To avoid memory retrieval problems To limit ex-post rationalizations embedded in the shared organizational memory To take into account typical development times in the industry

Representativeness (prototypical of certain class of projects)

To minimize between case variance within homogeneous classes To avoid rare cases or outliers

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

3.2.2. Recency The second requirement for project selection was the timing of its completion. In addition to considering typical memory retrieval problems in the analysis of past events, the issue of timing needs to account for industry specific development cycles (Baden-Fuller and Stopford, 1991; Sobrero, 1995). Following both methodological and context specific considerations, the selection criteria required the project not be ended earlier than 1 year prior to the beginning of the study. 3.2.3. Representativeness The third criterion is related to the first and reflects the intention, within the study, not to sample outliers. Although this is considered a common research strategy whenever using case-design, the intention here was to avoid either rare or peculiar cases. Rather, to ensure analytical representativeness, the project selected must have been considered typical of its own innovativeness category. 3.3. Data collection To address the usual concerns on the reliability of the data gathering process in qualitative research, we used a priori developed data protocols (Yin, 1994). These protocols identified a set of items to be covered in the data collection and their logical relationships.

167

They were used to guide the field research, to provide a standard format for data coding, and as a device to be challenged with new evidence emerging from the analysis. Whenever new elements not initially considered were included, old notes were re-elaborated to address the new aspects and further data collection followed if data were not available. Data were collected between March 1995 and January 1996 through direct interviews, the analysis of internal records, and public sources such as industry specific newsletters, consumer reports and business press. After selecting the projects, a first round of interviews was conducted with “top management” to gather general data at the group level (see Table 2). In addition to an historical account of the group evolution and growth during the years, and its current positioning in the European markets, some specific issues were covered with the group R&D director and the group purchasing director. With the first, together with each unit R&D manager, we examined the evolution of the product families at the “unit” level. With the second, the general purchasing strategies and objectives were discussed at length. The goal here was two-fold. First, we wanted to place each project within the proper product and market related context. Second, we wanted to identify the general environment in which the relationships with suppliers we were going to study were being activated.

Table 2 An account of the interviewing activity: number of interviews by round, unit and informanta Round

Purpose

Informant

Unit A

1

To To To To

2

select the projects obtain an historical account of the company and of each unit define the technical and market evolution of the company’s products determine the company’s purchasing strategies

B

C

Group CEO, group CTO, group CPO

5

5

4

To confirm the evidence collected in the previous round To reconstruct the history of the project examined

Unit CTO

7

7

6

3

To confirm the evidence collected in the previous round To reconstruct the history of the project examined To identify specific instances of interactions with suppliers

Unit CPO, PM, designers, engineering, director

7

7

8

4

To confirm the evidence collected in the previous round To analyze each relationship with suppliers

Unit CPO, PM, designers, engineering, director

7

6

5

26

25

23

Total a

CEO: chief executive officer; CTO: chief technology officer; CPO: chief purchasing officer; PM: project manager.

168

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

The second round of interviews was conducted at the “unit” level. With the R&D manager, we reconstructed an historical account of each project, both directly through his descriptions, and indirectly by accessing internal reports, budgets and product planning documents. Typically, after a first meeting organized along these lines, we were granted access to archival documents that we used to triangulate the information collected during the interview. For each project, we identified the different phases, their timing and costs, an articulation of the different dimensions of the problem solving activities, which organizational units were involved and responsible during the different phases, major difficulties encountered and unresolved issues, and a judgment on the overall performance of the project both financially and organizationally. We then prepared a summary of the evidence collected and of the discrepancies that had emerged. This summary was used as the basis of a subsequent discussion, after which the process was repeated one more time, for a total of three interviews which lasted between 2 and 4 h each. During these interviews, in addition to deepening the understanding of the project, particular attention was given to the identification of other key informants, with whom a third set of interviews was conducted. The purpose was to confirm the evidence collected on the project and to go into greater depth about the specific interactions with suppliers. The interviewees were the Project Manager, some designers, key people from the Engineering Department, and the Unit Purchasing Director. The interviews with each person lasted between 1 and 2 h and covered their individual role in the project and any account of specific interactions with suppliers. Working with the Project Manager and the Unit Purchasing Director, we developed a list of all the suppliers that were involved at some stage of the project. This list was used as a basis for further discussion with the project members about each single relation with the supplier through a fourth round of direct interviews. First, we discussed at length the characteristics of each single interaction (e.g. what was the supplier called to do, at which stage was it involved, what type of communication channels were used), and the responses collected were reviewed with the informants to resolve any major inconsistencies. Then, we distributed to the project manager

a structured questionnaire to obtain a judgmental indication of each single relationship’s performance.

4. Data analysis 4.1. The context The three SBUs selected can be considered representative of strategic players operating in today’s mature mass markets. They are focused on specific product segments within which they support all group brands. The export share is therefore quite high (see Table 3), even for unit C, which offers a geographically-specific product. Productivity levels, measured in number of pieces per employee, are within industry average (Paba, 1991), and so are the number of different product codes recorded in the production plan. Overall, these data offer more specific evidence for the general characteristics of fragmentation of consumer preferences and excess of installed manufacturing capacity, pushing for technical separation of product responsibilities into separate units, able to serve the whole set of brands used by the group both in local markets and all around Europe. Suppliers participate intensively in this organization of production. Total purchases account for between 26 and 65% of each SBU’s sales. Even the lower 26% value for unit A is fairly high, considering the level of vertical integration of refrigerator plants. Despite this key role, purchasing strategies are quite different from one unit to another. Unit C uses formal vendor rating procedures and has initiated a program to reduce the number of suppliers and introduce a quality certification system. Units A and B, in contrast, lack any strategic action on supply management. The Table 3 Main characteristics of the three SBUs: mean yearly values 1990–1994

Product segment Productivitya Export (%) No. of models produced Purchasing costs (%)b a b

A

B

C

Cold 564 40 600 26

Wet 613 30 600 60

Wet 587 14 150 65

Number of pieces per employee. As a percentage of total SBU sales.

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

purchasing office is organized in two second level units, one for components and materials and the other for manufacturing equipment. There are some attempts to monitor suppliers’ capability to respect some given quality levels, but attention is primarily given to manufacturing activities and does not take into consideration a more integrated vision of vertical relationships to include new product development activities as well. Within these specific contexts, the three projects were selected by design to represent different innovation efforts. The first project, developed in unit A, was targeted to the introduction of new models in an existing product family (different sizes in a line of refrigerators). The second project, developed in unit B, was centered around the re-engineering of a key subsystem (the washing group in a washing machine). The third project, developed in unit C, was targeted to the complete redesign of a new product platform and the corresponding changes in the manufacturing plant. Due to confidentiality reasons, projects cannot be compared using financial data. Yet some other quite informative dimensions can be used instead (see Table 4). Total development time from the official launch of the project to the beginning of production is 13 months for project A, 30 months for project B, and 50 months for project C. The targeted cost reduction, measured in real values with respect to the corresponding previous product generation is about 2.5% for project A, 3% for project B and a high 10% for project C. The expected product life is more sensitive to the specific product–market conditions, with a low of 3 years for project B, a high of 10 years for project C, and 5 years for project A. The percentage of new components with respect to previous products is in general fairly high, despite

Table 4 The three projects: main characteristics

(%)a

New components Development time (months) Components reduction (%)a Cost reduction target (%)b Expected product life (years) a b

A

B

C

90 13 25 2.5 5

55 30 27 3 3

95 50 20 10 10

In comparison with previous product generation. Real values, relative to previous product generation.

169

between-project differences, indicating how central is the overall product architecture engineering to harmonize the technical subsystems included in the product. For this reason, all projects are also targeted to high levels of components reduction, reflecting the type of innovative efforts performed. In project A, the value is the highest, with a target of 27% fewer components. In project C, on the contrary, the value is the lowest, with a target of 20%, and project B somewhere in the middle, with a target of 25%. The more the innovative activity is limited in its scope, the more the efforts are focused on more stringent product rationalization plans, with rather limited direct opportunities on the manufacturing side. Product simplification, in fact, potentially determines at the same time a reduction of material costs and of direct costs through the elimination of manufacturing activities. In contrast, more innovative projects like project C can specifically incorporate in the new design radical intervention on the process side, and leave to future projects more focused actions on the overall product architecture. 4.2. Characteristics of the tasks jointly performed Within all three projects, suppliers played a central role in the development of new solutions. Overall, 11, 18, and 21 suppliers of components or dedicated equipment were involved at some stage of the development process and assigned specific design responsibilities in projects A, B, and C, respectively. Their activities accounted for about 25% of the final manufacturing cost for project A, 20% for project B, and 27% for project C. These are quite large values, if one takes a closer look at their economic implications. In the case of washing machines, for example, the full manufacturing cost is around US$ 170 per unit. If the project target is to reduce such cost by 3%, this means to save in real terms US$ 5.40 per unit. A clear distinction of the activities performed by suppliers and how these relationships are structured therefore becomes a key strategic issue to be managed by the manufacturer, not only ex-post on the production side, but also ex-ante during the development process itself. Extensive interviews were conducted with each Project Manager to distinguish what kind of development work was assigned and what kind of design autonomy and responsibility was granted to each supplier. After an initial round of interviews, a

170

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

first classification of the relationships discussed was provided to the interviewee, together with an account of the evidence that emerged during previous meetings. Possible disagreements were resolved during an additional round of interviews and through the use of internal documents. Four typologies of interactions were identified through this process (Sobrero and Roberts, 2001). In the first case, all design work and related problem solving was done internally. The supplier received detailed drawings and technical specifications to be met for a component which was not considered to critically affect other parts of the overall project. The supplier’s action domain was therefore limited to meet these technical specifications. This group represents the traditional subcontracting approach in new product development and is therefore labelled “traditional subcontracting” parts. In the second case, suppliers were involved in the design of components, which might critically impact other parts of the project. The control over suppliers’ design freedom was imposed indirectly by the up-front definition of the interface specification provided. In this type of relationship, suppliers were recognized as an important source of knowledge, but the characteristics of the component suggested limiting their freedom in problem solving activities. In a way, they are integrated into the manufacturer’s problem solving logic and process. We label this group “integrated subcontracting” parts. In the third case, suppliers worked on components with low interdependency with the rest of the project. The manufacturer here limited its own activities by selecting an area of the project and delegating it entirely to the supplier. The supplier domain of action was therefore widened, albeit limited to areas of potentially low influence on the overall project. In this type of relationship, the manufacturer tried to access a specific knowledge domain of the supplier, without limiting its potential outcome by a predetermined set of solutions. We called this group “advanced subcontracting” parts. In the fourth case, suppliers were responsible for the whole problem solving activity for highly critical components. Despite the potentially high influence on the overall project, suppliers were given freedom to define the solution starting from the concept design domain and then moving to the functional parameter domain (Alexander, 1964; Clark, 1985). This practice

has been highly documented in Japanese firms as a way to reduce the overall product development time and costs (Imai et al., 1985; Clark, 1989). Components developed according to these arrangements have been identified as “black box” parts which term we therefore adopted as our label. Fig. 2 reports data on the economic impact of each category of relationships for the three projects studied, measured as the percentage of full manufacturing cost accounted for by externally developed components. Suppliers’ involvement goes well beyond the mere design of already well-defined engineering solutions typical of “traditional subcontracting” parts. Relationships requiring specific interactions to integrate external solutions in the overall product architecture are at the core of the economic impact of suppliers on the project. Most of the outside involvement is centered on “integrated subcontracting” and “advanced subcontracting” parts, while reliance on “black box” parts seems fairly limited with respect to what has been observed in other studies (Clark and Fujimoto, 1991; Liker et al., 1995; Dyer, 1996). This can partly be explained by the nature of the technology involved. Innovation in major home appliances comes from investments in product architecture and process flexibility, rather than from the inclusion of sophisticated components or subsystems alone. On the one hand, it would be hardly perceived by the consumer; on the other hand, the different technologies used tend to be rather mature. Even the most recent efforts towards the production of “intelligent” machines relying on fuzzy logic programmed microprocessors are based on the integration and adaptation of knowledge already used in other industries. External sources therefore become potentially useful in the process as long as the manufacturer controls and masters their coordination and integration in the process. In the next paragraphs, we therefore turn the attention to the contractual and information transfer mechanisms used in the sample observed. 4.3. Contractual coordination mechanisms To evaluate the extent to which each one of the multiple interactions activated within the three projects was governed by a set of contractual coordination mechanisms explicitly articulated for that specific relationship, we asked each unit’s purchasing manager

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

171

Fig. 2. Total value of supply by type and group (% of full manufacturing cost).

two different questions: first, if the contract negotiated with the supplier included elements regulating a set of 10 items 2 derived from previous research on European purchasing activities (Zanoni, 1984); and second, for each of these items, to indicate on a 5-points Likert scale their positioning between relying on standard contracts (1) and negotiating a specific agreement (5). In this way, information on the breadth of the contractual arrangement and on its specificity was collected for each one of the 50 dyadic relationships (Cronbach’s α = 0.78). The scores on each item were then added up and divided by 10 (for the 10 items), generating a final index between 0 (no element included in the contract) and 5 (all elements explicitly negotiated for that specific relationship included in the contract). Fig. 3 reports the values of this index measuring the extent of use of contractual coordination mechanisms for each project, distinguishing different types of supply. Clearly, the attention paid to the contractual definition of several contingencies is limited. 2 The items were price, quantity, quality levels, lead time, volume elasticity, order activating procedures, penalties and fines, exclusivity, control mechanisms, renewal.

While there are no statistically significant differences among supply types (F3,44 = 0.32, n.s.), the use of contractual coordination mechanisms varies among projects (F2,44 = 11.26, P < 0.001). A closer look at the use of each of the items considered for a contract, however, limits the substantial implications of inter-project differences along the contractual coordination dimension (see Fig. 4). Only quantity and price items are included in all contracts, while more sophisticated items such as some kind of agreement governing volume elasticity, how orders are activated, the possibility to rely on penalties or fines, and ex-ante options to renew the contract are never present in any of the relationships observed. The detected differences are explained by the inclusion of specific clauses regulating suppliers’ quality levels, the mechanisms to control such levels and exclusivity agreements in 5 contracts out of 11 in project A. Contracts are here used as a substitute for more elaborated form of supplier management such as the one being introduced in unit C, where the certification and definition of specific vendor rating procedures make individual negotiations on these aspects redundant. Previous research suggested the importance of repetitive interaction among the partners to foster

172

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

Fig. 3. Contract specificity by supply type and project.

mutual reliance, favor inter-partner knowledge and limit the need for leveraging on contracts (Hamel, 1991; Heide and Miner, 1992; Powell and Brantley, 1992). The percentage of suppliers involved in the project which had already interacted with the different units in the past is quite high (see Table 5), ranging from 72% for project A to 86% for project C and with no statistically significant difference among the

projects (F2,47 = 0.41, n.s.). This evidence for a repetitive interaction effect to explain the low investment in contractual coordination mechanisms, however, is challenged by the absence of a significant correlation between the existence of a previous relationship and the level of contractual specification used to govern the interaction observed. There is inconsistency in the sign, the size and the significance level of the

Fig. 4. Contract items by cases of use.

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

173

174

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

coefficients computed for each project, with apparently supporting evidence emerging only in the case of project B. A comparison of the three correlation coefficients standardizing the magnitude differences and taking into account the unequal sample sizes, however, does not detect any statistical heterogeneity among them (χ 2 -test = 3.72, d.f. = 2, n.s.). Similar results are also obtained when linking the economic value of the activity performed within the relationship and the level of contractual specification. Different authors have used cost-based measure of dyadic interaction to operationalize asset specificity, indicating that high levels of asset specificity justify investments in contractual coordination mechanisms (see, for example, Joskow, 1987). Yet the correlation between the value of the component developed entirely or in part by the supplier, measured as a percentage of the cost of the component over the total cost of the final product, and the level of contract specificity does not support this conclusion. While this seems to be true in the case of project B, the correlation is negligible for project C and reversed for project A. Moreover, these apparent differences are not statistically confirmed by the numerical comparison of the three coefficients (χ 2 -test = 4.07, d.f. = 2). Finally, we examined the examined the relationship between supply type and the level of contractual specification, controlling for possible inter-project differences. A two-way ANOVA did not show any significant supply type effect (F3,40 = 1.1, n.s.) or any significant interaction between supply type and inter-project differences (F4,40 = 0.651, n.s.). In general, therefore, despite quite substantial differences in the type of activities performed within the relationships examined, there seem to be no particular investments in contractual coordination mechanisms. Contrary to previous evidence, neither a trust-based explanation focusing on repetitive contact among the partners, nor a transaction cost-based explanation focused on the economic value of the interaction can be exclusively used to account for these choices. Possible influences of a group-level policy towards the form of legal agreements with external actors are also not supported by the recorded differences in the three SBUs’ purchasing strategies and procedures described before. Differences in geographical location of both the units and the suppliers are not helpful either to assess a predominant role of the external context,

and underdetermined explanations should rather rely on some general patterns of interaction at the whole industry level (Granovetter, 1985; Cusumano and Takeishi, 1995). Contrary to what we expected, however, in the sample studied, there are no significant differences in the use of the contractual leverage according to the object of the exchange. This absence of a strong link between two variables that are oftentimes presented as critical areas for strategic decision making whenever approaching potential collaboration is particularly interesting if we consider that the presence of previous relationships between the partners does not emerge as a significant intervening variable in our data. Previous research suggested that information transferring mechanisms can be used as substitutes for contractual coordination mechanisms (Piore and Sabel, 1984; Granovetter, 1985; Best, 1990). Repetitive interactions generate inter-partner learning, fostering trust and facilitating communication processes. The investments necessary for these changes may already sufficiently commit the relationship partners to each other (Heide and Miner, 1992). Consequently, additional contractually based commitment mechanisms might not be necessary anymore. In other words, while the task characteristics may suggest that the partners should establish a high level of contractual coordination, it may actually be unnecessary to do so if the partners have already committed to each other through previous investments in procedural coordination mechanisms, which contribute to steer decisions on the legal and organizational arrangements chosen to govern subsequent interactions. This evidence suggests the limits of contractual coordination mechanisms as the main leverage for structuring inter-organizational relationships, and the need for a complementary analysis of the information transfer mechanisms used in the different cases. 4.4. Procedural coordination mechanisms For each one of the 50 relationships identified we gathered data to determine (a) how early the supplier was involved in the project, (b) the extent to which the manufacturer and the supplier worked sequentially or in overlap during the project, (c) how often they exchanged information, and (d) what kind of information media they used to exchange information. The first two dimensions take into consideration

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

175

Table 6 Use of Information transfer mechanisms by type of supply: mean values and inter-group comparisons Dimension of evaluation

Early involvement Degree of overlap Communication frequency Richness of information media ∗

Kruskal–Wallis χ 2

Supply type Traditional

Integrated

Advanced

Black box

0.62 0.28 1.20 1.80

0.57 0.64 2.28 2.61

0.61 0.50 1.92 2.38

0.71 0.44 1.75 2.75

2.68 7.99∗ 16.08∗∗ 8.30∗

P < 0.05. P < 0.01.

∗∗

the results of numerous studies on effective ways to accelerate product development (for a review, see Brown and Eisenhardt, 1995). To measure early supplier involvement, we computed the difference in days between the date of first contact with the supplier in the project and the end of the project, normalized by the total length of the project. The index varies between 0 (latest involvement) and 1 (earliest involvement). To measure the degree of overlap in development activities, we computed the difference in days between the date the contract was signed and the date of first contact with the supplier, normalized against the longest time span within each of the three projects. The longer this span, the more both parties had a chance to overlap their activities. The index therefore varies between 0 (perfectly sequential) and 1 (completely overlapping). The other two dimensions tackle the role played by frequent and rich information transfer for faster and more effective problem solving activities (Allen, 1977; Bastien, 1987; Orlikowsky, 1993). In the first case, we measured how often the supplier and the manufacturer interacted during the project. Communication frequency was coded using a 3-points Likert scale (1 = less than once a month, 3 = information more than once a month), determining the appropriate anchoring lag through several in-depth interviews. In the second case, we measured what kind of information media was used predominantly in the relationship. Five options ordered hierarchically following a Gutman scale logic were derived from previous research on communication (Rogers and Agarwala-Rogers, 1983) to form a “media type” index for each relationship ranging between 1 and 5: (1) drawings, (2) faxes, (3) meetings, (4) personal communication among the project members and (5) electronic data exchange.

Table 6 reports the average values of the four indicators of procedural coordination mechanisms, distinguishing by type of activities. In general, all classes of suppliers are involved fairly early in the project. The values of the “early involvement” index are all above 0.5, indicating that initial contacts with suppliers were always established in the first half of the project life. The difference among the various classes are not statistically significant and the largest value is recorded for “black box” parts, signalling an earlier involvement of suppliers in the project for these type of activities. The analysis of the timing of information transfer within the relationship provides statistical support for differences among relationships types, suggesting higher levels of overlapping for those activities requiring a more direct involvement of suppliers in the design phase. The largest value (0.64), however, is recorded for “integrated subcontracting” parts, indicating that the higher level of overlap occurs for activities where the supplier is still quite strongly bound to specific engineering indications communicated by the manufacturer. Considering how the index is constructed—i.e. the normalized difference between the date when the contract was signed and the date of first contact—the results might be suggesting that in these cases it takes longer to “freeze” the design of the component. With this interpretation in mind, larger values could also signal more intensive information exchange processes needed to jointly define the preferred solution. These intuitions are confirmed by a closer look at the values of the indicators of frequency and richness of information transfer. In general, differences among the four type of relationships identified on the basis of the design work performed by the parties involved are statistically confirmed in both cases. Moreover,

176

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

relationships with suppliers for the development of “integrated subcontracting” parts require on average more frequent communication and use more elaborate type of information transfer mechanisms. The increasing level of investment in the type of media used to transfer information is also observed in general for more sophisticated type of relationships. Differences in communications frequency, on the contrary, are less evident when focusing on specific comparisons between pairs of relationships. “traditional subcontracting” parts are clearly characterized by the lowest value, but the differences between “advanced subcontracting” parts and “black box” parts are negligible. The analysis of procedural coordination mechanisms offers some elements to unveil the interaction processes behind the different types of relationships. In general, we observe that whenever suppliers are assigned problem solving responsibilities, they tend to work in overlap, focusing on a part of the problem which then needs to be linked through more frequent interactions, using more sophisticated ways of transferring information. However, if this freedom is limited to the identification of a solution within well pre-specified boundaries, as is the case with “integrated subcontracting” parts, the interaction process seems to require more resources and a larger coordination effort. In these cases, the access to an external knowledge pool seems to be nominal, as it is confined within boundaries pre-specified internally by the manufacturer. The manufacturer is constraining the development of functional parameters, with the identification of a design concept developed internally using an inappropriate knowledge base. These differences in the interaction patterns, reflected in alternative use of procedural coordination mechanisms, are likely to be reflected in different relationship outcomes. More sophisticated types of interactions might be establishing the conditions to leverage external knowledge sources, but certainly require higher levels of organizational investments, which are reflected in more intensive and frequent interactions. Similarly, outsourcing decisions might speed up the whole process by efficiently using external resources to subcontract part of the development activity, but a lack of clarity on the goals or a mispecification of the initial problem might generate a costly trial and error adjustment process. It is therefore

important to conclude this analysis by looking at the performance implications of the use of different types of relationships. 4.5. Relationship outcomes Inter-organizational relationships can be activated for different reasons (Sobrero and Roberts, 2001). Efficiency-increasing goals are based on the assumption that economic activities could indifferently be organized through internal integration or through the combination and coordination of the work of separate economic agents, and the outsourcing decision is cost driven. Learning-enhancing arguments, on the contrary, focus on the characteristics of the knowledge being transacted. While explicit knowledge can easily be codified and transferred among different actors, tacit knowledge is embedded in its owner. Inter-organizational relationships therefore become the necessary structuring solution to overcome the difficulties of trading knowledge-based assets, by providing the opportunities to establish direct contacts with the sources of knowledge and their development environments. To distinguish between these two possibly contrasting relationship outcomes, we distributed a eight-item questionnaire to the three Project Managers, asking them to evaluate each supplier involved in the development project by expressing a judgment on a 5-points Likert scale associated with each item. Six of the eight items were typical operationalization of efficiency such as the adherence to time, cost and quality levels, respect for initial agreements and the expected level of competencies, and a general judgment of satisfaction (efficiency Cronbach’s α = 0.91). The other two items asked whether the supplier developed some new solutions that were used in other occasions (e.g. other product lines, other projects), and whether they felt they had learned something during the interaction (learning Cronbach’s α = 0.72). To further support the results of the reliability analysis, we ran a “principal components” analysis using the whole set of items. The first two eigenvalues accounted for 75% of the variance (50 and 25%, respectively). The efficiency items all loaded strongly and positively on the first eigenvalue (component loadings varied from 0.75 to 0.93) and not on the second (component loadings varied from −0.20 to 0.20), while the “learning”

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

177

Table 7 Evaluation of relationship by type of supply: mean values and inter-group comparisons Relationships outcome

Efficiency Learning ∗∗∗

Kruskal–Wallis χ 2

Supply type Traditional

Integrated

Advanced

Black box

0.181 −0.853

−0.455 0.072

0.993 0.737

0.246 0.961

14.411∗∗∗ 20.338∗∗∗

P < 0.001.

items loaded strongly on the second eigenvalue (component loadings 0.85 and 0.91) but not on the first (component loadings −0.04 and 0.08). We used the results of the “principal component” analysis to create the two output indicators. For the “efficiency” indicator, for each relationship, all the responses to the six items described above were summed, weighted by their corresponding component loadings on the first eigenvalue. For the “learning” indicator, the same procedure was applied, using the other two items and their corresponding component loadings on the second eigenvalue instead. Both indicators were then standardized to make between-case comparison more meaningful and more easily interpretable. Several elements of interest emerge comparing the average standardized “efficiency” and “learning” outcome levels focusing on the type of development work characterizing the interaction (Table 7). First, in both cases, there are clear differences in the outcome depending on the characteristics of the activity performed within the relationship. “Integrated subcontracting” parts show the lowest value, clearly indicating an impact on the final outcome of the relationship of the repetitive interaction patterns highlighted before, mainly generated by the need to correct through trial-and-error processes the problems generated by initial mispecifications. On the contrary, relationships based on “advanced subcontracting” parts and “black box” parts are characterized by above average levels of efficiency, showing a more appropriate combination of the type of activity performed by the partners and the organization of their interaction. The leverage of supplier competencies is best achieved for “advanced subcontracting” parts, when the knowledge specificity associated with the design of the new solution suggests granting complete freedom to the partner involved, but the limited need of

integration of the component in the overall project does not require particular investments in the design of its interface. On the contrary, these specific design requirement lower on average the overall efficiency of interaction focused on “black box” parts, where the interaction must be more intense to co-ordinate the integration of the subsystem in the product architecture, requiring specific joint efforts in the design and problem solving activities dedicated to the definition of the technical interface. Average levels of efficiency result from interactions on “traditional subcontracting” parts, where suppliers’ contributions are reflected in asymmetric cost structures. While there seem to be some opportunities to benefit from the cost advantages provided by external partners in the engineering phase of the development process, the internalization of the detailed definition of components design specifications clearly limit the opportunities to compress component development costs. The same conditions favoring higher levels of efficiency in the relationships outcome, however, seem to be acting in a reverse manner towards the emergence of learning effects. The “learning” indicators associated with the different types of relationships show increasing values for more complex and articulated interactions. Suppliers’ involvement in the project for “traditional subcontracting” parts and “integrated subcontracting” parts show below average or average opportunities to foster the accumulation and transfer of new knowledge through the relationship. On the contrary, relationships with suppliers responsible for “advanced subcontracting” parts, and even more so relationships with suppliers responsible for “black box” parts, provide more opportunities to stimulate internal learning processes. Once more, the results reflect the differential investments in procedural coordination mechanisms. A higher level of involvement of external actors in the process generate higher coordination costs, as

178

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

noted in the analysis of the “efficiency” indicator, but also create the conditions for an effective information transfer process to occur. The theoretical requirements for facilitating the exchange of invisible assets such as knowledge and competencies are operationalized in specific structuring arrangements chosen to govern the interaction between the two partners. An earlier involvement in the project, coupled with more elaborated communication structures and an overlapping approach to problem solving activities are necessary conditions to foster inter-partner learning. These investments on the procedural coordination side, however, are costly. Clearly, they are more demanding in terms of coordination time, as well as resource allocation to the management of the relationship. A comparison of the trends of “efficiency” and “learning” indicators is informative in this sense. On the one hand, there seems to be a trade-off between the two different types of outcome. On the other hand, a closer look at the data suggest that we rethink the whole problem of inter-organizational relationships in new product development as a set of complementary decisions involving coherent choices on the characteristics of the interaction and its structuring arrangements to achieve the desired outcome. Unless the involvement of external actors goes beyond a mere subcontracting of part of the development process, any potential effect can at best be limited to a partial reduction of overall development cost. Yet it can also generate severe sources of inefficiency if the nature of the general definition of the problem solving domain cannot be effectively approached using internal resources. More articulated types of relationships, on the contrary, offer the opportunity for limiting these risks by a proper allocation of the design responsibilities to the owner of the relevant knowledge. The apparently higher coordination costs faced during the project to structure and manage these relationships are compensated in the end by the avoidance of negative loops generated by initially weak definitions of the problem to be solved, and by the emergence of opportunities to generate variance in the internal resource set with the exposure to new ideas and solutions. Initially greater investments are therefore to be compared to the possible alternatives focusing on the characteristics of the information exchange process that will emerge from the use of different procedural coordination mechanisms.

5. Conclusions This paper contributes to the debate on the analysis of inter-organizational relationships in the development of innovations with the recomposition of institutional decisions driving the governance of inter-firm relations and their organizational implementation. Building on previous works in economics, strategy and organizational design, the framework developed considers as fundamental elements for the effectiveness of inter-firm relations both status variables (e.g. determinants of cooperation) and process variables (e.g. procedural mechanisms to coordinate the relation). It links inter-organizational structures to the nature of the task being performed and articulates such structures into contractual and procedural coordination mechanisms, distinguishing between the institutional definition of the legal boundaries of the transaction, and the information transfer mechanisms needed to achieve an effective transfer. Finally, it shows the theoretical relevance of distinguishing between these two complementary dimensions to ascertain the implications for individual partners’ outcome of the activation of external relationships. Using a multi-case multi-level analysis approach to the study of 50 supplier–manufacturer relationships in new product development, the empirical results show that, indeed, the leverage on contractual coordination mechanisms can be fairly limited, and so will be its impact on the relationship outcome. On the contrary, procedural coordination mechanisms offer more opportunities to differentiate the structuring mechanisms used to manage the interaction among the partners. In addition to that, with different formulations of the outcome variable, similar choices along the structuring dimensions generate different results, depending on the characteristics of the joint problem solving. Higher investments in procedural coordination mechanisms are costly to set up and offer opportunities to pay back as long as they are used to nurture the manufacturer’s knowledge base, rather than merely reduce internal development costs through sub-contracting. On the contrary, whenever short term efficiency is the goal of the interaction, suppliers should be given full responsibilities, limiting the interactions with the manufacturer and its investments in procedural coordination mechanisms.

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

Several implications can be derived from these results for an analysis of the strategic impact of relational activities. First, planning and implementing ties and connections with other actors goes well beyond the definition of the contractual coordination mechanisms chosen to govern the relation. Although frequently considered as a fundamental aspect, the definition of the legal boundaries of the agreement is not sufficient to obtain the desired outcome. External conditions such as local systems of norms and practices might even question the necessity of specific investments in the contractual coordination side. While still a key component of the decisions over alternative structuring arrangements of inter-organizational relations, the research presented in this paper suggests that manufacturers not overlook the procedural coordination side by placing an excessive emphasis on the contractual coordination side, at the expenses of the final outcome of the relation. Second, the impact on the performance of the actors is strictly linked with the structuring decision taken with respect to the procedural coordination mechanisms used to make the relationship work. The effect of alternative choices on how to implement the exchange among the parties attracts the attention to the relevance of inter-organizational structuring decisions. Similar structures, however, might be economically justified only in certain cases, considering the coordination costs associated with more articulated relational options. Indeed, the empirical results show that for investments in relational activities to generate attractive returns they need to be congruent with the type of task jointly performed. Depending on the characteristics of the task at hand therefore different procedural coordination mechanisms can have very different impact on the outcome of the subjects involved. On the one hand, a pure financial evaluation based on short term observations of such impact might underestimate the effect deriving from changes in the competence set which need longer time spans to become visible. On the other hand, leaner structures might generate a positive impact on the overall cost structure of the focal firm, but could not be exploited to affect its competence base. Differences in the observed outcome of alternative procedural coordination mechanisms reflect their appropriateness to achieve different outcomes. Invest-

179

ments in dense and differentiated relational arrangements are certainly increasing the opportunities for exchanges, but the related costs are also increasing. Depending on the expected outcome therefore the choice of the corresponding set of procedural coordination mechanisms should be made accordingly. In this paper, we have discussed how inter-organizational structuring decisions can impact the innovative outcome of new product development projects, while we have not collected data on process innovation. Closer and denser interactions with suppliers, however, have emerged in different studies as powerful sources of continuous improvements also on the shop floor. More generally, it can be argued that such opportunities could be multiple, depending on all the opportunities that the partners have to interact. Future research could be directed to mapping and assessing the connections between the different relational sets activated by the partners, and the impact of such spillovers on innovation. Finally, while we have decided to focus on one specific actor along the value system (the manufacturer) we believe that the issues raised in the paper could be useful and transferable all along the chain. The well-established concept of organizing suppliers in multiple level tiers, which as been widely explored in the automobile industry, is quickly migrating to other industries, from aerospace to packaging machinery. Managing inter-organizational relationship therefore becomes a key aspect of more advanced approaches to the extended supply chain, requiring the development of specific relational capabilities not only at the manufacturer level, but also at the supplier level. Acknowledgements Partial support from MIT Industrial Performance center, MIT International Center for Research on the Management of Technology, and MURST ex-40% #9913628915-003. References Abernathy, W.J., Clark, K.B., 1985. Innovation: mapping the winds of creative destruction. Research Policy 14, 3–22. Aldrich, H.E., 1972. Technology and organizational structure: a reexamination of the findings of the Aston group. Administrative Science Quarterly 17 (1), 26–43.

180

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

Alexander, C., 1964. Notes on the Synthesis of Form. Harvard University Press, Cambridge, MA. Allen, T., Katz, R., 1990. The treble ladder revisited: why do engineers lose interest in the dual ladder as they grow older, The International Center for Research on the Management of Technology, Massachusetts Institute of Technology. Allen, T.J., 1977. Managing the Flow of Technology. MIT Press, Cambridge, MA. Allen, T.J., 1986. Organizational structure, information technology, and R&D productivity. IEEE Transactions on Engineering Management EM-33 (4), 212–217. Allen, T.J., Hauptman, O., 1987. The influence of communication technologies on organizational structure. Communication Research 14 (5), 575–587. Allen, T.J., Katz, R., 1986. The dual ladder: motivational solution or managerial delusion? R&D Management 16 (2), 185–197. Arora, A., Gambardella, A., 1991. Complementarity and external linkage: the strategies of the large firms in biotechnology. Journal of Industrial Economics 37 (4), 361–379. Baden-Fuller, C., Stopford, J.M., 1991. Globalization frustrated: the case of white goods. Strategic Management Journal 12, 493–507. Bastien, D.T., 1987. Common patterns of behavior and communication in corporate mergers and acquisitions. Human Resource Management 26 (1), 17–33. Best, M., 1990. The New Competition: Institutions of Industrial Restructuring. Harvard University Press, Cambridge, MA. Bidault, F., 1998. The drivers of cooperation between buyers and suppliers for product innovation. Research Policy 26 (7/8), 719–733. Brockhoff, K., 1992. R&D cooperation between firms: a perceived transaction costs perspective. Management Science 38, 514–524. Brown, S.L., Eisenhardt, K.M., 1995. Product development: past research, present findings, and future directions. Academy of Management Review 20 (2), 343–378. Burns, T., Stalker, G.M., 1961. The Management of Innovation. Free Press, New York. Carter, J.R., Miller, J.G., 1989. The impact of alternative vendor–buyer communication structures on the quality of purchased materials. Decision Sciences 20 (4), 759–775. Ciborra, C., 1991. Alliances as learning experiments: cooperation, competition and change in hightech industries. In: Mytelka, L.K. (Ed.), Strategic Partnerships: States, Firms and International Competition. Farleigh Dickinson University Press, Rutherford. Clark, K.B., 1985. The interaction of design hierarchies and market concepts in technological evolution. Research Policy 14 (5), 235–251. Clark, K.B., 1989. Project scope and project performance: the effect of parts strategy and supplier involvement on product development. Management Science 35 (10), 1247–1263. Clark, K.B., Fujimoto, T., 1991. Product Development Performance: Strategy, Organization, and Management in the World Auto Industry. Harvard Business School Press, Boston, MA. Combs, K.L., 1990. Cooperative research and innovative activity. In: Link, A.N., Smith, V.K. (Eds.), Advances in Applied Microeconomics, Vol. 5. JAI Press, Greenwich, CT, pp. 47–67.

Contractor, F.J., Lorange, P. (Eds.), 1988. Cooperative Strategies in International Business. D.C. Heath and Company, Lexington, MA. Cusumano, M., Takeishi, A., 1995. What have we learned and what have we yet to learn from manufacturer–supplier relations in the auto industry. WP#3840, Alfred P. Sloan School of Management, Massachusetts Institute of Technology. Dore, R., 1986. Flexible Rigidities: Industrial Policy and Structural Adjustment in the Japanese Economy. Athlone, London, pp. 1970–1980. Duncan, R.B., 1972. Characteristics of organizational environments and perceived environmental uncertainty. Administrative Science Quarterly 17, 313–327. Duncan, R.B. (Ed.), 1976. The Ambidextrous Organization: Designing Dual Structures for Innovation. The Management of Organization Design: Strategies and Implementation, I. North-Holland, New York. Dyer, J.H., 1996. Specialized supplier networks as a source of competitive advantage: evidence from the auto industry. Strategic Management Journal 17 (4), 271–291. Dyer, J.H., Ouchi, W.G., 1993. Japanese style business partnerships: giving companies a competitive edge. Sloan Management Review 35 (1), 51–63. Eisenhardt, K.M., 1989. Building theories from case study research. Academy of Management Review 14 (4), 532–550. Freeman, C., 1991. Networks of innovators: a synthesis of research issues. Research Policy 20 (5), 499–514. Fruin, M.W., 1992. The Japanese Enterprise System, Oxford University Press, New York. Galbraith, J., 1973. Designing Complex Organizations. AddisonWesley, Reading, MA. Galbraith, J.R., 1977. Organization Design. Addison-Wesley, Reading, MA. Goshal, S., Haspeslagh, P., 1990. The acquisition and integration of Zanussi by Electrolux: a case study. European Journal of Management 8 (4), 414–433. Granovetter, M., 1985. Economic action and social structure: the problem of embeddedness. American Journal of Sociology 91 (3), 481–510. Gulati, R., 1998. Alliances and networks. Strategic Management Journal 19, 293–317. Hakanson, L., 1993. Managing cooperative research and development: Parter selection and contract design. R&D Management 23 (4), 273–285. Hakansson, H. (Ed.), 1987. Industrial Technological Development. A Network Approach. Croom Helm, London. Hakansson, H., 1989. Corporate Technological Behavior: Co-operation and Networks. Routledge, London. Hamel, G., 1991. Competition for competence and inter-partner learning within international strategic alliances. Strategic Management Journal 12 (Summer special issue), 83–103. Harrigan, K.R., 1988. Joint ventures and competitive strategy. Strategic Management Journal 9, 141–158. Harrison, B., 1994. Lean and Mean: The Changing Landscape of Corporate Power in the Age of Flexibility. Basic Books, New York.

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182 Heide, J.B., Miner, A.S., 1992. The shadow of the future: effects of anticipated interaction and frequency of contacts on buyer–seller cooperation. Academy of Management Journal 35 (2), 265–291. Helper, S., 1991. Strategy and irreversibility in supplier relations: the case of the US automobile industry. Business History Review 65 (4), 781–824. Henderson, R., 1994. The evolution of integrative capability: innovation in cardiovascular drug discovery. Industrial and Corporate Change 3 (3), 607–630. Henderson, R.M., Clark, K.B., 1990. Architectural innovation: the reconfiguration of existing product technologies and the failure of established firms. Administrative Science Quarterly 35, 9–30. Hladik, K.J., 1988. R&D and international joint ventures. In: Contractor, F.J., Lorange, P. (Eds.), Cooperative Strategies in International Business. Lexington Books, Lexington, MA. Iansiti, M., 1995. Technological development and integration: an empirical study of the interaction between applied science and product development. IEEE Transactions on Engineering Management 42 (3), 259–269. Iansiti, M., 1995. Technology integration: managing technological evolution in a complex environment. Research Policy 24 (4), 521–542. Iansiti, M., Khanna, T., 1995. Technological evolution, system architecture and the obsolescence of firm capabilities. Industrial and Corporate Change 4 (2), 333–362. Imai, K., Nonaka, I., Takeuchi, H., 1985. Managing the new product development process: how Japanese learn and unlearn. In: Clark, K.B., Hayes, R.H., Lorenz, C. (Eds.), The Uneasy Alliance: Managing the Productivity-Technology Dilemma. Harvard Business School Press, Boston, MA, pp. 337–376. Joskow, P.L., 1987. Contract duration and relationship-specific investments: empirical evidence from coal markets. American Economic Review 77 (1), 168–185. Katz, M.L., 1986. An analysis of cooperative research and development. Rand Journal of Economics 17 (4), 527–543. Katz, M.L., Ordover, J.A., 1990. R&D cooperation and competition, Brookings Papers: Microeconomics, pp. 137–203. Katz, R., Allen, T.J., 1985. Project performance and the locus of influence in the R&D matrix. Academy of Management Journal 28 (1), 67–87. Kleinknecht, A., Reijen, J.O.N., 1992. Why do firms cooperate on R&D? An empirical study. Research Policy 21 (4), 347–360. Kogut, B., 1988. Joint ventures: theoretical and empirical perspectives. Strategic Management Journal 9 (2), 319–332. Kogut, B., Zander, U., 1992. Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science 3 (3), 383–397. Larson, E.W., Gobeli, D.H., 1988. Organizing for product development projects. Journal of Product Innovation Management 5, 180–190. Lawrence, P.R., Lorsch, J.W., 1967. Organization and Environments: Managing Differentiation and Integration. Irwin, Homewood, IL. Liker, J.K., Kamath, R.R., Nazli Wasti, S., Nagamachi, M., 1995. Supplier involvement in automotive component design: are there really large US Japan differences? Research Policy 25, 59–89.

181

Lorenzoni, G., Baden-Fuller, C., 1995. Creating a strategic center to manage a web of partners. California Management Review 37 (3), 146–162. Marquis, D.G., Straight, D.L., 1965. Organizational factors in project performance. Sloan School of Management, Massachusetts Institute of Technology. Monteverde, K., Teece, D., 1982. Supplier switching costs and vertical integration in the automobile industry. Bell Journal of Economics 13, 206–213. Nelson, R., Winter, S., 1982. An Evolutionary Theory of Economic Change. Harvard University Press, Cambridge, MA. Nelson, R.R., Winter, S.G., 1977. In Search of a useful theory of innovation. Research Policy 6, 36–76. Nishiguchi, T., 1995. Strategic Industrial Sourcing: The Japanese Advantage. Oxford University Press, New York. Nishiguchi, T., Ikeda, M., 1996. Suppliers’ process innovation: understated aspects of Japanese industrial sourcing. In: Nishiguchi, T. (Ed.), Managing Product Development. Oxford University Press, New York, pp. 206–230. Nobeoka, K., 1993. Multi-project management: strategy and organization in automobile product development. Sloan School of Management, Massachusetts Institute of Technology. Orlikowsky, W., 1993. From memo to dialogue: enacting genres of communication in electronic media. WP#3525-93, Alfred P. Sloan School of Management, Massachusetts Institute of Technology. Ouchi, W.G., Bolton, M.K., 1988. The logic of joint research and development. California Management Review 30 (3), 9–33. Paba, S., 1991. Reputazione ed Efficienza: Crescita e Concentrazione Nell’industria Europea Degli Elettrodomestici Bianchi. Mulino, Bologna, IL. Parkhe, A., 1993. Strategic alliance structuring: a game theoretic and transaction cost examination of inter-firm cooperation. Academy of Management Journal 36 (4), 794–829. Perrow, C., 1967. A framework for comparative organizational analysis. American Sociological Review 32 (2), 194–208. Pine, J.B., 1993. Mass Customization: The New Frontier to Business Competition. Harvard Business School Press, Boston, MA. Piore, M., Sabel, C., 1984. The Second Industrial Divide. Basic Books, New York. Pisano, G.P., 1990. The R&D boundaries of the firm: an empirical analysis. Administrative Science Quarterly 35 (1), 153–176. Pisano, G.P., 1991. The governance of innovation: vertical integration and collaborative arrangements in the biotechnology industry. Research Policy 20, 237–249. Polanyi, M., 1967. The Tacit Dimension. Doubleday Anchor, New York. Porter, M.E., Baden Fuller, C. (Eds.), 1987. Coalitions and Global Strategy. Competition in Global Industries. Harvard Business School Press, Boston, MA. Powell, W.W., Brantley, P., 1992. Competitive cooperation in biotechnology: learning through networks? In: Nohria, N., Eccles, R.G. (Eds.), Networks and Organizations: Structure, Form, and Action. Harvard Business School Press, Boston, MA, pp. 366–394.

182

M. Sobrero, E.B. Roberts / Research Policy 31 (2002) 159–182

Rebentish, E., Ferretti, M., 1995. A knowledge–asset based view of technology transfer in international joint ventures. Journal of Engineering and Technology Management 12 (1/2), 1–25. Ring, P.S., Rands, G.P., 1989. Sensemaking, understanding, and committing: emergent interpersonal transaction processes in the evolution of 3M’s microgravity research program. In: Van de Ven, A.H., Angle, H., Poole, M.S. (Eds.), Research in the Management of Innovation: The Minnesota Studies, Vol. 13. Ballinger/Harper & Row, New York, pp. 483–498. Roberts, E.B., 1980. New ventures for corporate growth. Harvard Business Review 58 (4), 134–142. Roberts, E.B., 1988. What we’ve learned: managing invention and innovation. Research Technology Management 31 (1), 11–29. Roberts, E.B., Berry, C.A., 1985. Entering new businesses: selecting strategies for success. Sloan Management Review 26 (3), 3–17. Rogers, E. M., Agarwala-Rogers, R., 1983. Communication in Organizations. Free Press, New York. Saxenian, A.L., 1994. Regional advantage: culture and competition in Silicon Valley and Route 128. Harvard University Press, Cambridge, MA. Sobrero, M., 1995. The European home appliance industry: history, structure and technology 1945–1992. WP#3801, Alfred P. Sloan School of Management, Massachusetts Institute of Technology. Sobrero, M., Roberts, E.B., 2001. The trade-off between efficiency and learning in inter-organizational relationships for product development, Management Science, 47 (4) in press. Sobrero, M., Schrader, S., 1998. Structuring inter-firm relationships: a meta-analytic approach. Organization Studies 19 (4), 585–616. Sobrero, M., Toulan, O., 2000. Task partitioning, communication activities, and the performance of supplier relations in new

product development: an empirical analysis. In: Birkinshaw, J., Hagström, P. (Eds.), Capability Management in Network Organizations. Oxford University Press, London, pp. 21–43. Stake, R.E., 1995. The Art of Case Study Research. Sage, Thousands Oaks, CA. Teece, D.J., 1986. Profiting from technological innovation: implications for integration, licensing and public policy. Research Policy 15 (6), 285–305. Thompson, J.D., 1967. Organizations in Action. McGraw-Hill, New York. Tripsas, M., Schrader, S., Sobrero, M., 1995. Discouraging opportunistic behavior in R&D consortia: a new role for the government. Research Policy 24 (3), 367–390. Tushman, M.L., 1978. Technical communication in R&D laboratories: the impact of project work characteristics. Academy of Management Journal 21 (4), 624–645. Watkins, T.A., 1991. A technological communications costs model of R&D consortia as public policy. Research Policy 20, 87–107. Williamson, O., 1975. Markets and Hierarchies. Free Press, New York. Williamson, O., 1979. Transaction cost economics: the governance of contractual relations. Journal of Law and Economics 22, 233–262. Williamson, O., 1985. The Economic Institutions of Capitalism. Free Press, New York. Woodward, J., 1965. Industrial Organizations: Theory and Practice. Oxford University Press, London. Yin, R.K., 1994. Case Study Research: Design and Methods. Sage, Thousands Oaks, CA. Zanoni, A., 1984. Gli Approvvigionamenti. ETAS Libri, Milano.