Long Range Planning 42 (2009) 216e233

http://www.elsevier.com/locate/lrp

Organising R&D Projects to Profit From Innovation: Insights From Co-opetition Bruno Cassiman, Maria Chiara Di Guardo and Giovanni Valentini1

Increasingly, technological innovation results from the joint creation effort of different players in the value chain, such as suppliers, customers, research centres and universities. Balancing co-operative and competitive forces in the innovation process to co-create value and to capture part of this value has become crucial to profit from innovation. In this article, we show that this tension between value creation and value capture is present in each R&D project. Drawing on the case of STMicroelectronics, we show that the balance of co-operative and competitive forces in R&D projects is made through the careful alignment of three variables: project knowledge attributes, project governance structure (internal development, co-operation or contracting), and project partner selection (firm or university). The capability to match these three elements explains the success of the innovation process of STMicroelectronics. Building on the experience of this firm, we provide some practical guidance on how managers should trade off these co-operative and competitive forces in organising their R&D projects. Ó 2009 Elsevier Ltd. All rights reserved.

Introduction A recent McKinsey Global Survey of top executives found that 70 per cent of respondents considered innovation one of their companies’ top three strategic priorities for driving growth.1 The same survey found that fewer than half of the executives were satisfied with the financial returns on their investments in innovation. While innovation is now considered key for creating a sustainable 1 The authors would like to thank Giuseppe Ferla and Cristina Di Gesu´ of STMicroelectronics for their time and the access provided to data on R&D projects. The authors are grateful for the comments received from Giambattista Dagnino as well as from the participants at the EIASM Workshop on ‘‘Co-opetition strategy: Toward a new kind of interfirm dynamics?’’ (Catania) on previous drafts of this paper. Bruno Cassiman acknowledges partial funding from the Spanish Ministry of Education and Science and Technology through the project n SEJ2006-11833/ECON and the Fundacion Ramon Areces.

0024-6301/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.lrp.2009.01.001

competitive advantage, many firms struggle with its successful organisation. In this paper we show why balancing the concerns of creating value through innovation on the one hand, and capturing returns from innovation on the other is such a complicated task. This task requires aligning different elements at the R&D project level, i.e. each R&D project and activity needs its particular organisation, complicating the innovation management task considerably. Successful innovation depends on the development and integration of new knowledge in the innovation process. Some have argued that technological knowledge is typically broadly distributed, so that no single firm has all the internal capabilities and resources needed for success. In such an environment, the locus of innovation is found in a network of inter-organisational relationships. However, managing these inter-organisational relationships requires skilful selection of which external agents to involve in the innovation process and when to involve them. The choice of opening the firm’s innovation process assures some gains through accessing complementary knowledge sources, but at the same time exposes the firm to the risk of opportunistic behaviour from external partners. Inter-organisational R&D agreements are, therefore, characterised by a combination of competitive and collaborative actions.2 In previous literature, the nature of the interfirm interdependences where both competitive and collaborative issues co-exist is called co-opetition.3 Co-opetition is a synthesis between two opposite paradigms: the competitive paradigm, postulating that firms interact on the basis of a fully divergent interest structure, and the co-operative paradigm, assuming that firms interact on the basis of a completely convergent interest structure. Co-opetition provides a powerful framework to better understand how firms combine e or should combine e internal and external innovation activities in their R&D process. Firms may display co-operative interest structures at the time of co-creating value through an R&D project, but they may also undergo competitive pressures at the time of capturing the value created. Received theory provides some interesting results on the factors that explain the establishment and management of co-opetitive relationships at the firm level.4 However, most innovation managers will be confronted with a varying degree of these issues at the R&D project level. We believe it is important to understand which factors are relevant at the project level when different projects within the same firm will line the spectrum from highly co-operative to highly competitive pressures, trading off value creation for value capture elements to a different extent. As literature has made only small inroads into understanding the co-opetition strategy at this level of disaggregation, we base our analysis on a quantitative case study. We adopt the R&D project as the key level of analysis, and explore the organisation of 52 projects carried out in one of the most important research centres of STMicroelectronics (ST) e the largest microelectronics company in Europe e between 1998 and 2003. ST is a successful and innovative firm that has made the coopetitive approach a cornerstone of its innovation strategy. We build on the findings from this case study to discuss the relevant variables an innovation manager should take into account when designing a suitable organisation of an R&D project. Our analysis establishes three key findings. First, the balance of co-operative and competitive forces in the organisation of R&D projects is made through the careful alignment of three variables at the project level: project knowledge attributes, project governance structure (i.e., co-operation vs contracting), and project partner selection (firm, university or research institute). Depending on the nature of the project, the firm decides on the suitable governance form and the partner with the appropriate resources and incentive structure in order to balance the value creation and value capture forces at stake. In essence, with whom the firms establish a link depends on co-operation issues e linked with the capabilities and skills it potentially needs e and competitive pressures related to the risks and the opportunities involved in the specific activity. Second, the complex relationship between value creation and value capture considerations is relevant for the organisation also within a single project: different partners and different capabilities, along with different organisational structures, are sought in different phases of an R&D project. Finally, through the analysis of the firm’s balancing of co-operative and competitive forces at the project level, this paper also sheds new light on the important question of how Long Range Planning, vol 42

2009

217

do firms build valuable capabilities to achieve superior and sustainable performance? While the key argument of the paper is related to the management of co-opetitive forces at the level of individual R&D projects that lead to innovation, we also provide useful insights on how firms can manage critical knowledge cogeneration that forms the basis for their competitive advantage. Managing this variety of R&D projects along this trade-off between value creation and value capture can lead to a management capability that constitutes a sustainable competitive advantage. However, the fact that we study only one successful case does not allow us to easily generalise our results. The remainder of the paper is organised as follows. First, we discuss the relevant characteristics of the innovation process and illustrate the drivers of the organisation of R&D projects, as well as their expected influence on the organisational form chosen. Second, we present our research setting and research methods, which is followed by the results of our empirical analysis. We finally provide some concluding remarks and develop a simple heuristic for managers when deciding on the organisation of their R&D projects.

The innovation process and the drivers of R&D projects organisation Today even the largest and most technologically self-sufficient organisations require knowledge from beyond their boundaries. Accordingly, recent studies provide mounting evidence about the potential for combining internal and external sourcing modes as complementary innovation activities and the firms that combine the two activities outperform those firms that are active in only one single activity.5 As firms increasingly use external relationships to acquire new knowledge, they need to develop the capability for governing these relationships. Hence, as prior research suggests, two forces shape firm decisions concerning the organisational boundaries of R&D. On one hand, firms need to look for external productive resources in their developmental efforts and collaborate in the creation of valuable innovations. On the other hand, this external search should be balanced against the risks of encountering opportunistic behaviour of partners that compete in the appropriation of the returns of these innovations. The R&D project organisation should, therefore, reflect the opportunities and threats of both co-operative and competitive forces.6 To clearly represent these forces to balance, we propose to adopt the co-opetition framework. Three factors that condition the balance of competitive and co-operative forces stand out from the literature: 1. The knowledge attributes of the project 2. The form of governance at the project level 3. The characteristics and the nature of the objectives of the partners. We now briefly expand on each of these factors and explain how a correct alignment of these three elements is crucial to put the firm in the best position to manage the value creation and value capture trade-off. Specifically, in the discussion section, we will focus on how the project knowledge attributes influence the governance of R&D projects and the nature of partners selected, as we believe they provide an interesting starting point for constructing a managerial heuristic to optimise the R&D project.

As firms increasingly use external relationships to acquire new knowledge, they need to develop the capability for governing these relationships 218

Organising R&D Projects to Profit From Innovation

Knowledge attributes of R&D projects Knowledge has been shown to be an important contingent variable influencing organisational decisions in different technological settings.7 In particular, as recent research has pointed out, the knowledge structure of the firms involved in a co-operative relationship affects the relative balance in co-operative and competitive elements and, hence, is an important determinant of a co-opetitive relation.8 Therefore, we claim that the characteristics of the knowledge involved in an R&D project, that is, its knowledge attributes, significantly influence its organisation and the propensity of the firm to engage in different innovation activities and with different partners. To characterise these attributes, we need to take into account the fact that each project is embedded in a broader context. Specifically, while there are objective characteristics of the knowledge involved in a project, other characteristics acquire full meaning only with respect to the relative situation in which a project is developed and, in particular, to the firm’s current and future positioning. Meaningful knowledge attributes should thus include both objective and relative dimensions. We argue that four knowledge attributes are particularly relevant and influence the organisation of R&D projects: basicness, novelty, ease of industrialisation and strategic importance.9 Basicness is an objective characteristic of the project, which indicates its relatedness to basic research (as opposed to applied). Novelty and ease of industrialisation are project knowledge characteristics relative to the firm’s current positioning. Novelty refers to the extent to which the knowledge involved in a project is novel relative to the existing knowledge base of the firm. The ease of industrialisation refers to the ability of the firm to produce and commercialise the potential output of the R&D project. These two measures e novelty and ease of industrialisation e consider how a project relates to the two most important resources for creating and appropriating value from innovation: a firm’s knowledge base and complementary assets. Finally, the fourth knowledge attribute e the strategic importance of the knowledge involved in the project e relates to the future positioning of the firm. The strategic importance of the project knowledge mirrors and acquires significance only with respect to the choices the firm has made for its future. Forms of governance The possible forms of governance the inter-organisational agreements may assume at the project level constitute the second aspect to be considered. There are several types of hybrid arrangements that firms can find themselves in aside from in-house work (hierarchy) or simple contracting. This is evident more specifically within the field of research and technology relationships.10 Earlier literature on R&D projects has seldom analysed hybrid forms of governance, that is, co-operation, as an alternative to contracting. In this study, we decided to focus on formal links between organisations in the R&D projects considering and distinguishing both co-operation and contracting. As researchers note, an outside organisation is defined to co-operate with the firm when it is formally co-responsible for project goals and results with the firm.11 Conversely, contracting implies that the outside organisation simply commits to deliver a contractually specified output to the firm for some activities. These two different modes of the relationship have a significant effect on how R&D projects are organised and co-ordinated, notably in terms of the (unequal) distribution of dependency and power. More important, these different forms of governance result in a different relative level of co-operation versus competition between the partners. Characteristics and nature of objectives of partners Finally, the third dimension that we take into account is the characteristics and the nature of the partners that are selected for the potential relationship at the project level. Specialisation gives each partner an idiosyncratic set of properties (including its capabilities, but also its own internal structure, communication codes and social practices) and objectives. Different partners provide access to different resources, knowledge and capabilities. For example, a large literature suggests that universities and industrial firms have complementary resources and skills. While universities and Long Range Planning, vol 42

2009

219

research centres have access to intellectual resources and a world-class basic research infrastructure, industrial firms usually have practical expertise, financial resources, internship opportunities for students and employment opportunities for graduates and students.12 An R&D agreement with another firm has different implications with respect to value appropriation issues compared with an agreement with a university. Universities and business firms are in fact characterised by different structures and incentives, the former being traditionally more interested in free knowledge diffusion.13

How knowledge attributes affect governance and partners’ choices We argued that each of the factors discussed e the relevance of project knowledge attributes, the forms of governance of the relationship between agents and the characteristics and the nature of objectives of the partners e affects the balance of co-operative and competitive elements in an R&D relationship and, hence, has a direct effect on how one manages the balance between the value creation and value capture forces at stake. We start by focusing on the project knowledge attributes that are particularly relevant for the organisation of R&D projects and relate these to possible governance forms and the nature of partners. Basicness e or relatedness to basic research e is a first relevant dimension of a project’s knowledge. On one hand, the basicness of the project should favour the recourse to co-operation. Firms may have little economic incentive for investments in basic research. Uncertainty about the results and appropriability hazards are two relevant causes for this reduced incentive. The same factors may also negatively affect the creation of a ‘‘market for knowledge.’’14 Co-operation in R&D may mitigate these problems. First, by co-operating with other organisations (for example, with a competitor), firms may share risks and costs.15 Second, co-operation may facilitate the internalisation of knowledge spillovers. Moreover, by co-operating, firms may acquire or build capabilities they would not have by simply contracting out their needs. Opportunities to learn require firms to be willing to share their knowledge with partners. As a consequence, the more basic a project is, the lower the weight of competitive forces in a co-operative relationship. The novelty of a project’s knowledge relative to the existing knowledge base of the firm represents a second important dimension. The co-opetition literature suggests that the knowledge profile of the firms has a role in generating the basic features of a co-operative context and consequently, it may affect the interest structure that emerges within a co-operative relationship (i.e., co-opetition). The novelty of a project increases the propensity to cross the R&D boundaries of the firm. Firms are more likely to look for complementary external resources when they are moving away from their knowledge domain, looking for partners with more productive resources given a specific task. While the higher basicness and novelty of a project provide strong co-operative impulses for the firm, competitive pressures are still present as these projects typically face higher uncertainty and the risks affecting the interaction between the partners. Technically novel projects need creative problem solving and thus may also cause unwanted delays and cost overruns. The resulting increase in outcome uncertainty may require hierarchical governance to guard against these hazards. Internal development could provide a better means to respond to unanticipated contingencies (or opportunities) over the course of the project. In addition, internal development relieves firms from fully specifying contractual arrangements, whose terms are less obvious and known when outcomes

Different partners provide access to different resources, knowledge and capabilities

220

Organising R&D Projects to Profit From Innovation

are new and uncertain. Moreover, the more basic and novel a project is, the more difficult it becomes to assess its outcomes. Uncertainty on performance measurement creates a higher incentive for opportunistic behaviour of partners, and thus, increases the strain on any co-operative endeavour. The third knowledge attribute we consider measures the ease for the firm to industrialise the knowledge developed in the project. Potential or current customers frequently suggest simple new ideas often leading to R&D projects with outcomes that are easily transferable to manufacturing. Such R&D projects are likely conducted in co-operation. However, when a project’s knowledge is easily transferable to manufacturing, it might also be easier to ‘‘steal’’, leading potential partners to experience a higher incentive to behave opportunistically due to its closeness to immediate results. From a co-opetition perspective, this implies a higher likelihood of competitive issues in such a co-operative relationship. Co-operation with competitors, for example, would be less likely in this situation. Finally, the strategic importance of the project is an essential driver for understanding the organisation of the R&D project. Projects that provide knowledge of strategic value to the firm increase the incentive for partners to behave opportunistically, engaging in a ‘‘learning race’’ and avoiding sharing the knowledge developed. In addition, projects of strategic importance are more likely to imply commitments and specific investments.16 The more specialised a resource, the lower its value in alternative uses, and the higher the probability of being held up by a partner. Therefore, internal development should be preferred, in particular to co-operation with a competitor.17 Nonetheless, in highly strategic projects it becomes apparent how no single firm has all the capabilities necessary for success, and how important the adoption of a co-opetition strategy is. Therefore, R&D contracting in some specific tasks may provide access to the required productive, external resources, guarding at the same time against the hazards of opportunistic behaviour. As such, the firm combines the need to look for external productive resources while minimising of the risk of encountering opportunistic behaviour of partners. Figure 1 clusters the factors e knowledge attributes, partners and governance forms e according to how they affect the trade-off between value creation e co-operative forces e and value capture e competitive forces. Following the analysis of this section, we expect empirically that external knowledge acquisition activities are organised by combining factors that are consistent with each other and, therefore, ‘‘close’’ to each other in Figure 1. For example, starting from the different knowledge attributes, we expect that more basic R&D projects will be organised as a co-operative agreement with the university as these projects are more geared towards value creation and, hence, co-operative forces dominate. On the opposite side of the spectrum, in strategically important R&D projects it will be important to control carefully the participation of partners as competitive forces are Value Creation Value Capture Knowledge Attributes

Basicness

Novelty

Ease of Industrialisation

University

Partner

Governance Form

Co-operation

Strategic Importance

Firm

Contract

Internal (Hierarchy)

Figure 1. Framework for Co-opetitive R&D projects’ organisation Long Range Planning, vol 42

2009

221

more prominent. When internal organisation of the R&D project would not generate the desired results, contracts on specific activities of these projects are better able to delineate each partners’ contribution. The type of partner included in the agreement, university or firm, will depend on the value-creating potential of the project. Actually, as we argue later, we would expect universities to intervene in earlier stages of such a project, while firms should enter in the later, more applied, stages. The box-insert highlights three real projects at STMicroelectronics showing how project managers struggle to balance the value creation and value capture forces in each of them.

Project 1

Project 2

Project 3

Type of innovation Knowledge Attributes

Organisation and Application Partner Selection

Emphasis is on

Platform project. Knowledge exploitation

Contracting out to university

Immediate application in terms of final product

Value appropriation

Collaboration with firms

Target is application in three to five years

Value creation/value appropriation

Collaboration with universities

Focus on Value creation long-term goals

Basicness: low Novelty: high Strategic importance: high Ease of industrialisation: low System technology Basicness: low innovation. Novelty: high Knowledge Strategic creation importance: medium Ease of industrialisation: high Advanced Basicness: high technological Novelty: high development. Strategic Knowledge importance: medium creation Ease of industrialisation: low

We report here on three selected interviews with project managers of STMicroelectronics on three different R&D projects. The projects differ in their knowledge attributes, governance structure (co-operation or contracting), and partner type (firm or university); all projects are subject to a non-disclosure agreement. The above table highlights the main features of each project. Project 1 aims at improving the existing technological capabilities and is focused on knowledge exploitation rather than on knowledge creation. It is characterised by high novelty and strategic importance. The innovative ideas arose from the collaboration with specific customers. While the project manager recognises that the clients are a very important source for innovative ideas, he also points out: ‘‘We are treating them simply as a source of information,’’ thus choosing not to directly involve them in the organisation of the innovative process. To balance the value creation and value capture forces, some activities were contracted out for this project rather than developed under collaboration. Given the high degree of novelty, contracting out preferentially involves universities rather than other companies. Contracting out specific activities to the university allows controlling the competitive aspects of this knowledge and universities are much more focused on value creation as opposed to value capture, and, hence, less threatening. However, in order to facilitate the value capture, our interviewee underlines the need for formulating specific requests and precisely defining work tasks when contracting with universities when mentioning: ‘‘It is a strange process. We work a lot with universities, but you have to have in mind in advance what you want from the university to get something back.’’

222

Organising R&D Projects to Profit From Innovation

Project 2 concerns knowledge exploitation. The project’s main activities are aimed at developing direct market applications to capture the value of the outcome. As the project focuses only to a lesser extent on basic research, the preferred organisational form is collaboration with other companies. However, given the strategic importance of the project’s outputs, collaboration with companies in all of the project activities is highly undesirable. Accordingly, involvement of external resources was limited in all phases characterised by a low ease of industrialisation (i.e. which involved high tacit knowledge transfers) while it was extensively used when the project activity implied high novelty for the company but not for the market. The manager was less concerned about the appropriation issue when outcomes were easy to formalise as the ease of industrialisation enhances communication between the partners. Project 2 receives only a modest contribution from the university, but this did provide the opportunity to build internal capabilities. As the project manager observed, there is a trade-off between the cost of acquiring information externally and the time needed to develop it internally. The further you are from the final application, or from the value capture moment, the easier it is to develop some capabilities internally, not being pressed by time-to-market, while focusing on the value creation perspective. Given the broad range of applicability of such capabilities, internal development constituted a valuable long-term investment. Project 3 is the most basic R&D project of the three, and is hence predominantly focused on value creation. Basic research projects are frequently carried out in collaboration with the central R&D teams of other sites. ‘‘My group generally deals with projects oriented to basic research,’’ said the manager of Project 3. He stressed the need to maintain expertise in the basic scientific disciplines in order to identify and scan ideas from universities, from other companies and from other industries. The new ideas and technologies have come from the collaboration with research centres, which are based inside the ST site. To facilitate knowledge transfer, the company co-located teams in a facility that allowed them to work together closely. Technological collaboration often involves the transfer of tacit knowledge to solve complex problems. Yet such practices and the socialisation processes they invoke make regulation of the depth and scope of knowledge transfer difficult.

Research setting and methods The semiconductor industry provides an interesting setting for our study on how firms attempt to manage the trade-off between value creation and value capture. The semiconductor industry is highly cyclical as the industry is subject to rapid technological change and experiencing significant economic up and downturns over short periods. Roughly every four or five years, new technological innovations are introduced into the market (e.g., production of DRAM, SRAM, Flash, etc.), closely tied to the development of new markets and new applications (increasing demands on PC memory, DVD players and decoders, mobile telephones, PDAs and other handheld devices, etc.). The cyclical nature of the industry requires individual firms to adjust their R&D and innovation programmes rapidly, building on their strengths while incorporating external knowledge. In addition, because of the speed the industry moves, the tension between value creation and value capture is maximal as innovating firms need to assure a return on investment in innovation in this short period of time. Reinforcing the trend to look for external knowledge over the past decade, the world’s leading semiconductor manufacturers pursued the goal of developing System-on-Chip (SoC) technologies, which aim at integrating in only one chip different functions to reach the total integration of a system. The needs of the future e mobility, multimediality, security and mass memory e are linked with the capacity to integrate blocks of pre-developed circuits (blocks of Intellectual Property-IP) in a SoC. The SoC approach requires integrating technology of semiconductor manufacturers e the silicon knowhow e and the system knowhow, typical of the electronic devices producers. Therefore, the uncertainty due to market cyclicality and complementary technological capabilities needed naturally led to an innovation strategy based on co-opetition. Our analysis relies upon a quantitative case study based on data collected at STMicroelectronics (ST), the largest European firm in the microelectronic sector. STMicroelectronics was created in 1987 through the merger of SGS Microelettronica of Italy and Thomson Semiconducteurs of Long Range Planning, vol 42

2009

223

France. Since its formation, ST pursued an aggressive growth strategy, investing heavily in R&D, forging strategic alliances with customers and academia, building up a presence in major economic regions and honing one of the world’s most efficient manufacturing operations. ST has consistently grown faster than the semiconductor industry as a whole and it has been one of the world’s top ten semiconductor suppliers since 1999. ST is active in numerous collaborative research projects worldwide. The effectiveness of partnerships ST has forged with many leading participants in the global semiconductor industry has been a key contributor to its success. These include strategic alliances with key customers, technology development alliances with both customers and competitors, development alliances with major equipment, materials and CAD suppliers and partnerships with multinational R&D organisations, universities and research institutes. During this time, ST was clearly best in class in combining internal and external knowledge sources. Therefore, we believe that ST with its ‘‘collaborate and compete’’ innovation strategy provides an interesting and relevant setting for our study. We obtain unusually detailed information on R&D projects by focusing on a single firm. The single firm focus helps us to better understand the relation between project knowledge attributes, partner type and governance form of an R&D project, while controlling for other firm-specific factors. Our empirical analysis focuses on an ST subsidiary in southern Europe which distinguishes itself for its research focus and the large amount of resources allocated to R&D. Our methodology combines qualitative and quantitative methods. Data collection and measures Data were collected, for more than a year, both from primary and secondary sources. After reviewing public documents concerning ST’s ‘‘co-operate and compete’’ innovation strategy, we interviewed three R&D division directors, a project manager and the head of the department for external R&D contracts to better understand ST’s internal innovation process. With the help of industry experts at ST, in 2003 we then collected fine-grained data at the project level from internal archival data, developing a database that contains 52 R&D projects representing all the projects started between 1998 and 2003 in one of the largest subsidiaries of ST. Our analysis required collecting data at the project level and we analysed the structure of each individual project as each was composed of a number of activities, which generally get started sequentially. For each project, we know the governance form e co-operation or not, and the type of partner e university or other firm in case of a co-operative agreement. We collected additional objective project measures such as total cost and length. Moreover, each project is evaluated and characterised with respect to its knowledge content by internal experts in accordance to explicit, specific internal procedures. These measures will serve as our knowledge attribute measures. First, on a 1-to-3 scale, the novelty of the knowledge developed in the project compared with the firm’s technological domain is evaluated. Second, on a 1-to-3 scale, the relevance of the knowledge developed to achieve product or process innovations that can enhance the competitiveness of the firm, i.e. the strategic importance of the project, is assessed. And third, on a 1-to-4 scale, the ease of industrialisation and of transferability to manufacturing of the outcomes of the project is evaluated. In all of the cases, a higher value on the scale implies a higher value of the construct. See Appendix 1 for further elaboration on research methods.18 For each activity, we know what the activity is technically directed to; the governance form of the particular activity e contract or not; and the partner in charge of the activity e university or firm in the case of a contract. Furthermore, we know whether the activity is devoted to the development and the acquisition of new knowledge, or rather involves the application and the achievement of concrete results from knowledge previously developed or acquired. From this latter measure we construct a project-level measure considering the percentage of the project activities that aim at developing new knowledge (as opposed to applying it). We use this measure as our indicator of the basicness of the project. Basicness may thus range from 0 to 1. 224

Organising R&D Projects to Profit From Innovation

Table 1. Descriptive data on R&D projects’ Organisation

Type of Project Organisationa Co-operation Partnership

Contract

Other organisation involved

Total

Conditional on having a Partnership

Total

Conditional on having a Contract

University Research Centre Firm

37% (19) 21% (11) 42% (22)

59% 34% 69%

40% (21) 31% (16) 8% (4)

70% 53% 13%

The table indicates the share of the projects included in the sample that are conducted in co-operation with or contracted to business firms, universities and research centres. The first column (‘‘Total’’) indicates the total share, with respect to the whole sample. The second column (‘‘Conditional on’’) indicates the share of projects that involves a specific partner (contractor) only for those projects that are conducted in co-operation (contracting). a

Partnerships and Contracts can involve different and multiple organisations simultaneously.

The sample The projects included in our sample have an average length of 31.5 months and require 8.2 manyears to be completed. Seven different strategic lines of innovation are explored: technological and design platforms; advanced applications, new devices and optoelectronic integrated circuits; memories and system on chip; nanotechnologies; new materials; bioelectronics, health; new computational models. About 90 per cent of the projects included in our sample span the boundaries of ST, having at least one partner in co-operation or at least one activity of the project contracted out; 29 per cent have both a partner and some activities contracted out. Table 1 reports the aggregate data regarding co-operative partnerships and contracts. The data indicate that co-operation and contracting are widely adopted. Still, while co-operative partnerships are conducted with business firms almost as likely as with universities (42 per cent versus 37 per cent respectively), one is significantly less likely to observe contracting of individual activities with firms compared with contracting with universities (8 per cent versus 40 per cent respectively). Research centres often supply special services and are more likely to be a contractual partner in a particular activity than a co-operative partner on the whole project.

Empirical analyses As we argued before, partner choice and governance form of the R&D project should balance the co-operative and competitive forces embedded in the knowledge attributes of the R&D project when the firm acquires knowledge externally. We, therefore, compare the knowledge attributes of projects carried out in co-operation (contract) with firms with those carried out in co-operation (contract) with a university or research centre. Table 2 reports the average value of different project knowledge attributes across different partner and governance form combinations and the significance of a t-test for differences in means in the value of knowledge attributes across projects that are carried out through alternative partner/governance form combinations.19 The top panel examines the co-operative governance form while the bottom panel focuses on the contractual governance form. We find that the more basic a project, i.e. the more it tends to develop new knowledge as opposed to apply earlier knowledge, the more likely that external capabilities are sought under a co-operative agreement. Projects conducted in co-operation with universities are significantly more basic (p < 0.01) than those that are not; the same happens for those conducted in co-operation with a firm (p < 0.05). Thus, in spite of the risk related to the uncertainty, more basic projects are Long Range Planning, vol 42

2009

225

Table 2. Projects’ knowledge attributes depending on projects’ form of governance: Means and t-test for difference in means

Dummy for co-operation in the project with a university

Basicness Novelty Strategic importance Ease of industrialisation

Dummy for co-operation in the project with a firm

1

0

1

0

0.92** 2.10 2.10y 2.73y

0.75 1.93 2.43 2.17

0.90* 1.95 2.00** 2.82*

0.76 2.03 2.50 2.10

Dummy for contracting in the project with a university

Basicness Novelty Strategic importance Ease of industrialisation N ¼ 52; yp < .10; )p < .05;

))

Dummy for contracting in the project with a firm

1

0

1

0

0.86 2.21*** 2.52** 2.55

0.76 1.74 2.00 2.22

0.90 2.25 2.75 2.00

0.81 1.98 2.25 2.44

p < .01; )))p < .001.

primarily carried out through co-operation to take advantage of external capabilities and seize learning opportunities irrespective of whether the partner is a university or a firm. At the same time, projects conducted in co-operation are characterised by knowledge that is significantly more easily transferable to manufacturing (p < 0.1 for universities and p < 0.05 for firms). This result is more surprising and we believe that this may relate to the lower cost of communication between partners of such a project. Yet, the search for external resources has to be balanced against competitive considerations: when project knowledge is characterised by a high strategic importance, co-operation at the project level does not seem to be the preferred option. On average, projects conducted in co-operation with firms and universities show a significantly lower strategic importance (respectively, p < 0.01 and p < 0.1).20 Tests for differences in means of knowledge attributes across projects conducted in co-operation with firms and universities confirm that there is no significant difference with one exception. Projects conducted in co-operation with universities display a higher novelty compared with those conducted in co-operation with business firms (p < 0.1). Not surprisingly, being at the science frontier, universities are specifically sought for projects that considerably depart from the current knowledge base of the firm. The bottom panel of Table 2 reports the average value of project knowledge attributes across projects in which at least part of the project is contracted out, to firms or universities, and projects in which there is no activity contracted out to these organisations. From these results, it appears that R&D contracting is chosen for strategically important projects and, contrarily to co-operation, not in more basic projects. Co-operative agreements are unlikely to sufficiently control the competitive forces in strategically important R&D projects. Governance through contracting seems therefore to provide a better ground for managing co-operative and competitive forces in strategic projects. On the contrary, co-operation allows the firm to tap specific resources that contracting does not allow. This difference is evident in more basic projects, where new knowledge has to be developed and in which relying on productive, specific resources e the value creation part e may be more important: in these projects co-operation is preferred to contracting. Co-operation across different activities of the R&D project allows not only the transfer of the agreed resources, but also the exchange of tacit knowledge and competencies, as well as the sharing of risks involved in basic projects.21 226

Organising R&D Projects to Profit From Innovation

Co-operation allows the firm to tap specific resources that contracting does not allow From our analysis on the relation between knowledge attributes and co-operation in Table 2 (top panel), it seems that universities and firms play very similar roles in co-operative projects except for the level of novelty of the project. By further disaggregating the level of analysis to the project activity level, we can gain insights on the different roles universities and firms may play in this coopetitive process. Therefore, we analyse which activities companies and universities were more likely to lead in projects in which they co-operate. As shown in Table 3, given an average project carried out in cooperation, two dimensions appear relevant in the decision on which external partner to choose for leading a particular activity: (1) whether an activity is aimed at developing new knowledge (as opposed to applying previously-held knowledge), i.e. the basicness of the individual activity; and, (2) the position of the activity within the sequence of activities constituting the project, i.e. initial development activities versus later stage application activities. The results presented in Table 3 highlight that universities are more likely to lead an activity in an R&D project when new knowledge has to be developed within this activity. Firms, on the contrary, are sought-after to lead activities in the later phases of a project. By co-operating with universities, companies can share the risk of an uncertain result, i.e. improve value creation, and at the same time reduce the risk of expropriation of value, i.e. reduce the threat to value capture. Moreover, most experts maintain that the success of a project is rooted in its early phases. These are phases in which the bases for the new technology are defined, where fundamental decisions are taken that will influence the whole course of the project. Firms may therefore be reluctant to expose themselves to the risk of opportunistic behaviour by potential e competitive e partners, fearing that competitive forces could prevail over co-operative ones. When the later activities of a project require applied abilities and the need to experiment with the knowledge previously developed, other firms may be the ideal partners. Not only does this reflect the nature of a particular industry technological practice we have discussed earlier, but also the fact that firms possess specific skills and may represent the potential customer of the final commercial application of the current project.22 Co-operative forces prevail when value creation considerations are at stake, while the competitive forces dominate when turning to value capture considerations. A profitable balance of these two opposing forces requires the selection of the partner and of the governance form that are carefully

Table 3. Organisation at the activity level: knowledge content and activity position

Firm leads the Activity

Basicness Activity Relative position Activity

University leads the Activity

1

0

1

0

0.92 0.54

0.94 0.49**

0.98*** 0.52

0.79 0.52

N ¼ 428

N ¼ 488

The number of observations N refers to the number of activities included in the R&D projects in which there is formal co-operation respectively with a firm and a university. Basic Activity is a dummy that takes the value of 1 if the activity is primarily aimed at creating new knowledge as opposed to applying it. Relative Position Activity is calculated dividing the position of the activity itself by the total number of activities comprising the project. For instance, the third activity of a project composed by 12 activities has a relative position of 3/12 ¼ 0.25. **p < .01; ***p < .001. Long Range Planning, vol 42

2009

227

aligned with the characteristics of the project. The complex relationship between these three elements e knowledge attributes, form of governance, partner profile e appears most evident when considering the effect of the project’s strategic importance for the organisation. For low levels of strategic importance, external partners are given full involvement as the importance of their complementary resources requires openness and full collaboration (co-operation paradigm). Actually, one might have expected the risk of opportunistic behaviour to make the firm more reluctant to engage in projects codeveloped with a firm as opposed to a university. This is not the case. Clearly, co-operative considerations related to the opportunities for value creation prevail. However, for high levels of strategic importance the hold-up risk dominates (competition paradigm) and co-operation fails. In this case, contracting with universities, widely used in strategic projects, seems to alleviate this problem. Both the governance form e a contract allowing tight control over the partner’s access to the project, and the partner’s different incentive structure e universities are less interested in capturing value, rather in creating and disseminating e mitigate the risk of opportunistic behaviour, and, consequently, the competitive forces. Our results therefore reinforce the idea that collaboration and competition can go hand-in-hand when working with a partner, but need to be carefully managed across different R&D projects based on their knowledge attributes requiring access to external knowledge. Management heuristic Our results provide some guidance about how to manage the trade-off between value creation and value capture forces when organising R&D projects. The knowledge attributes of the R&D project seem a natural starting point as R&D managers need to spell out the knowledge requirements for each project. Conditional on the knowledge requirements, the R&D manager can start considering the organisation of the project, i.e. given the knowledge attributes of the projects, which partners and which governance form are more suitable to make sure that the firm will profit from this innovation project. This will require balancing the value creation and value capture forces discussed before. Figure 2 shows the sequence of knowledge attributes to consider and how they can be converted into a heuristic for managing R&D projects. A high level for a given project knowledge attribute indicates that the project is above the mean on this attribute. The numbers in the figure show the percentage of projects with above (below) the mean knowledge attributes with that particular governance form and partner. Considering the level of basicness of the R&D project is the first step when thinking about the organisation of the project. A very basic project (Basicness High) is likely to be organised co-operatively with the university. In our sample, 52 per cent of projects of above average basicness are Basicness H: CoopU 0.52 Basicness L: CoopU 0.26

H

Co-operation with university Novelty H: ID 0.07 Novelty L: ID 0.43

Basicness L

Internal development

L Novelty H H

Strategic importance

Strat importance H: ContrU 0.82 Strat importance L: ContrU 0.56

Contract with university L

L

Ease of industrialisation H

Co-operation with a firm

Ease of indus H: Co-opF 0.82 Ease of indus L: Co-opF 0.25

Figure 2. Knowledge attributes and form of governance 228

Organising R&D Projects to Profit From Innovation

organised in co-operation with a university compared with only 26 per cent of the projects with low basicness. If basicness is low, we move on to the next knowledge attribute in the sequence: novelty. For low levels of novelty of the R&D project, knowledge is likely internally available and an internal project is organised. From our sample we find that 43 per cent of projects with low novelty are organised internally. However, if novelty is high, the strategic importance of the project comes into play. For high novelty and high strategic importance, contracting specific activities with the university is the likely outcome as the contract allows the control of the competitive aspects of this knowledge and universities are much more focused on value creation as opposed to value capture, and, hence, less threatening. At the same time, these universities own the capabilities needed in highly novel projects. As indicated, 82 per cent of projects with high strategic importance contract with a university. For low strategic importance, we examine the ease of industrialisation of the project. Low strategic importance and high ease of industrialisation leads to co-operation with a firm and 82 per cent of these projects in our sample are organised as such. In this case, the R&D manager should be less concerned about the competitive forces while it is easy to formalise agreements due to the easy transferability of project results. Low ease of industrialisation and high novelty of the project, however, are more likely lead to contracting with the university.

Discussion and conclusion In this paper we provided a novel framework and original evidence to explain when and how a firm decides to establish inter-organisational agreements in R&D projects. To do so, we disentangled the intricate interactions of the elements that might explain the decision on the boundaries and the organisation of R&D activities by exploring the relationship between project knowledge attributes, form of governance and characteristics of potential partners. The nature of our sample can only provide preliminary and exploratory results, and although this study has the potential for application in other settings, caution must be exercised when generalising from only one case study based in a specific industry. Nonetheless, we believe that (a) the theoretical framework we developed, (b) the particular empirical setting of the study, and (c) the empirical results we observed, all present interesting implications and contributions. Adopting a co-opetition perspective, we conceptualised firms’ R&D activities as an organisational process in which firms are subject to, combine and synthesise e in a word, manage e both co-operative and competitive actions. The former type of actions has primarily to do with value creation, while the latter with value capture. This approach is in line with a fresh stream of literature that has revamped the interest in considering separately the dynamics that underlie the value creation and the value capture processes, and in appreciating the differences between the two. While value capture is a zero-sum game that might induce firms to let competitive forces prevail, co-operative forces may dominate in the value creation phase. Even in a non-co-operative game theoretic setting, it is argued that when firms are deciding their strategic investments in R&D, ‘‘the total market pie may get enlarged when firms are accommodating’’.23 This might be true even at the expense of a possible increase in transaction costs, and is particularly true in an increasingly knowledgebased economy. Delmas, for instance, argues that ‘‘the Transaction Cost Economics framework, originally formulated in an environment of mature physical capital-intensive industries, may find limited application in environments where know-how is the key asset, where building rather than protecting specific assets is the main issue.’’24 In this paper, we have shown how firms can accommodate the often divergent forces of knowledge creation and knowledge appropriation through the proper management of different inter-organisational linkages. To get a deeper understanding of the relationship between these divergent forces, future studies could explore empirically the arguments proposed by Argyres and Zenger.25 They argue that the theories that are more concerned with knowledge appropriation e such as transaction cost economics e and theories that conversely stress the importance of knowledge creation are not competing but rather complementing each other in a ‘‘longitudinal’’ perspective. In other words, if capability differences between organisations play a key role in determining the firm’s governance Long Range Planning, vol 42

2009

229

choices, it is likely that transaction costs lie somewhere at the roots of these differences. That is, the distribution of specialised capabilities across firms at a particular point in time reflects a series of past decisions by these firms either to develop or not to develop capabilities internally, and these decisions, the authors argue, were likely driven by comparative governance or transaction cost considerations. However, an in-depth study on 52 R&D projects of STMicroelectronics allowed us to dig deeper into the mechanism through which co-operative and competitive forces can be reconciled. In particular we examined how different forms of governance (internal development, co-operation, and contract), different partners/contractors (universities or firms), and the different project knowledge attributes interact at the R&D project and R&D activity levels. While there have been a number of studies exploring the factors that drive the establishment of inter-organisational linkages in R&D, the great majority of these were conducted at the firm level; that is, they were aimed at identifying the firm-level variables that could help discriminating between ‘‘open’’ and ‘‘closed’’ firms.26 However, even the most open firms clearly do not involve other organisations in all their R&D activities. This study, by adopting the project level of analysis, allowed exploring in which specific activities within a firm’s innovation process other organisations are sought. By the same token, previous empirical studies have generally considered only one specific inter-organisational agreement (i.e., co-operation or contracting) as opposed to internal development, and did not distinguish between different types of organisations involved in these agreements. Conversely, in this study we simultaneously consider contracting as well as co-operation, and were able to distinguish between firms and universities. As we saw, some relevant differences across decisions related to organisational forms and partners do emerge. In particular, basic projects are generally developed through formal co-operative agreements. Such projects also tend to be strategically less important. For strategically more important projects, and when the knowledge to be developed is particularly novel, the firm resorts to formal contracting with a university for some specific activities of the R&D project, usually early on in the project. Moreover, thanks to our specific set-up, we were able see how managers should condition the organisation of R&D projects on the features of the knowledge they involve. While knowledge may seem an indefinite and abstract concept, we provide four characterisations e basicness, novelty, ease of industrialisation and strategic importance e that can guide the optimal organisation of R&D projects: when to partner, with whom to partner and how to govern the partnership. In the end, if R&D is about using, recombining and developing knowledge, one should not be surprised that knowledge attributes actually influence the organisation of R&D projects in an important way. The capability to organise R&D projects which requires considering and matching project knowledge attributes, project governance form and the specific features of potential partners is hard to develop overnight. STMicroelectronics has spent years learning how to manage these co-opetition relationships making it an integral part of its innovation strategy. This capability helps explain the success of its innovation process. ST has shown tolerance for risk, a culture that understands and fosters collaboration and a corporate strategy that emphasises the exploration of new opportunities. This capability is hard to imitate without the required investment of time and resources resulting in a ‘‘dynamic’’ capability sustaining competitive advantage in innovation over time. Furthermore, our results also present some relevant implications for theory. Most importantly, a more refined understanding of when and how other organisations are sought in R&D projects can elucidate the process underlying the benefits that opening the boundaries of R&D may bring about. For instance, links with science have been argued to be beneficial to firm performance, but little is known about the actual process underlying the economic consequences associated with such links.27 In addition, our theoretical developments and results about when and how to ally in R&D can enlighten future empirical studies aimed at investigating the performance outcomes of an open innovation strategy. By allowing a better estimation of the ‘‘selection’’ model, i.e. the decision to ally in R&D and, thus taking the endogeneity problem that these R&D decisions might entail into account, a better understanding of what affects the decision to ally as opposed to what directly affects R&D performance can be gained. 230

Organising R&D Projects to Profit From Innovation

Table A1. Results of multinomial logita

Co-op with a firm Constant Project cost Basicness Novelty Strategic importance Ease of Industrialisation N of observations c2 y

p < .10; )p < .05; a

))

p < .01;

2.68 (2.81) 0.12 (0.11) 5.85* (3.08) 0.70 (1.36) 3.57** (1.25) 0.34 (0.52) 52 51.29** )))

Co-op with university 1.31 0.01 4.55y 0.81 2.91** 0.08

(2.53) (0.10) (2.72) (1.10) (1.17) (0.44)

Co-op with both 9.2 0.08 24.25** 2.66 3.34** 0.43

(6.70) (0.12) (7.83) (1.75) (1.29) (0.56)

p < .001.

Standard errors in parentheses.

In sum, questions about the organisation of firms’ R&D activities are now too important to ignore. We hope that this study will help to reinvigorate their exploration.

Appendix Our sample comprises all the R&D projects for which ST has asked for some form of external financing in the time-frame of our analysis (1998-2003). These projects are traced in the ST database and archival data is available. Each project for which national or international external funding was asked was evaluated and characterised with respect to its knowledge content by the applicants Table A1. Measuring intangible variables is the main difficulty in this type of empirical research. While our measures may contain a degree of subjectivity, we believe they are valid measures of our constructs. First, the application for external public financing for industrial R&D projects is a complex process. All our interviewees e whether belonging to ST or public research institutes eacknowledged this fact. The process requires a deep technical knowledge as well as a sound understanding of the regulatory context. All the applications must contain some basic project characteristics, some are quantitative, some other qualitative. All the applications are reviewed by independent experts nominated by the funding organisation. This is a competitive process in which several applicants are denied financing. Misrepresentation of the characteristics of a project is easily detected by the reviewers, with the consequence of drastically decreasing the chances of being financed and harming the firm’s reputation. Second, financing applications e and thus knowledge attributes evaluations e have been redacted by the same team of experts over the five years of our sample of R&D projects. This team is not directly involved in the decision regarding the organisation of the project. Measures are thus compiled by highly qualified observers and made for entirely different purposes than those of our study and increase their validity.28 Last, we stress two important additional issues. First, ST generally asks external financing for the great majority of its research projects at our site, ruling out e or at least reducing significantly e one possible source of selection bias by only looking at R&D projects that request external financing. Second, the request for external financial support does not tend to alter the organisation of the projects. Our interviews highlight that the optimal organisational form is ex ante. And only then, the most appropriate financing programme is targeted, depending on project characteristics and organisation. What really matters is which funding programme to apply to, conditional on the project’s features. Therefore, we expect little systematic bias to affect our measures.

References 1. J. Barsh, M. Capozzi and L. Mendonca, How Companies Approach Innovation: A McKinsey Global Survey, McKinsey Survey on Innovation (2007). Long Range Planning, vol 42

2009

231

2. C. Clarke-Hill, H. Li and B. Davies, The paradox of co-operation and competition in strategic alliances: towards, Management Research News 26(1), 1e20 (2003). 3. A. Brandenburger and B. Nalebuff, Co-opetition, Currency Doubleday, New York (1996); Giovanni B. Dagnino and Giovanna Padula, On the nature and drivers of coopetition, International Studies in Management and Organization, Special Issue (2006). 4. Work has been done on the relationship between co-operation and competition in fields such as strategic management. E. L. Barbee and T. Rubel, Co-opetition in action, The Journal of Business Strategy 18(5) (1997); relationship marketing A. Palmer, Co-operation and competition: a Darwinian synthesis of relationship marketing, European Journal of Marketing 34(5/6), 687e704 (2000); networks T. Khanna, R. Gulati and N. Nohria, The dynamics of learning alliances: competition, cooperation, and relative scope, Strategic Management Journal 19(3) (1998); and supply chain management F. L. Martijn Rademakers and P. J. McKnight, Concentration and inter-firm co-operation within the Dutch potato supply chain, Supply Chain Management 3(4) (1998); as well as the professions D. Gilbert Jr., Co-opetition, Business and Society 37(4), 468e470 (1998). 5. B. Cassiman and R. Veugelers, In search of complementarity in the innovation strategy: internal R&D & external knowledge acquisition, Management Science 52(1), 68e82 (2006). 6. In previous literature, these two forces have been mostly analysed by the Resource-Based View perspective, as far as the search for external productive resources is concernedK. A. Bates and E. J. Flynn, Innovation history and competitive advantage: A resource-based view analysis of manufacturing technology innovations, Academy of Management Journal 235 (1995); J. H. Dyer and K. Nobeoka, Creating and managing a high-performance knowledge-sharing network: The Toyota case, Strategic Management Journal 21(3), 345 (2000); and by the Transaction Costs Economics literature, with respect to the risk of opportunistic behaviour. G. Pisano, The R&D boundaries of the firm: An empirical analysis, Administrative Science Quarterly 35, 153e176 (1990); O. E. Williamson, Markets and hierarchies: Analysis and Antitrust Implications, The Free Press, New York (1975). 7. J. Birkinshwaw, R. Nobel and J. Ridderstrale, Knowledge as a contingency variable: Do the characteristics of knowledge predict organization structure? Organization Science 13, 274e289 (2002); U. Zander and B. Kogut, Knowledge and the speed of the transfer and imitation of organizational capabilities: An empirical test, Organization Science 6, 76e92 (1995). 8. G. B. Dagnino and G. Padula, On the Nature and Drivers of Coopetition, International Studies in Management and Organization, Special Issue (2006). 9. For a discussion of the theoretical background of the knowledge attributes that are particularly relevant and influence the organisation of R&D projects, see B. Cassiman, M. C. Di Guardo and G. Valentini, Building Competitive Advantage through Links with Science: A Project Level Approach, IESE Working Paper, mimeo (2006). 10. J. Howells, A. James and K. Malik, The sourcing of technological knowledge: distributed innovation processes and dynamic change, R&D Management 33(4), 395e409 (2003). 11. B. H. Hall, A. N. Link and J. T. Scott, University as research partner, Review of Economics & Statistics 85, 485e491 (2003). 12. D. J. Bower, Successful Joint Ventures in Science Parks, Long Range Planning 26(6), 114e120 (1993); M. D. Santoro and S. Gopalakrishnan, Relationship Dynamics between University Research Centers and Industrial Firm: Their Impact on Technology Transfer Activities, Journal of Technology Transfer 26(1e2), 163e171 (2001). 13. P. Dasgupta and P. A. David, Toward a new economics of science, Research Policy 23, 487e521 (1994). 14. K. Arrow, Economic welfare and the allocation of resources for invention, The Rate and Direction of Inventive Activity, Princeton University Press, Princeton, 609e625 (1962). 15. L. Miotti and F. Sachwald, Cooperative R&D: why and with whom? An integrated framework of analysis, Research Policy 32, 1481e1499 (2003). 16. P. Ghemawat, Commitment. The Dynamics of Strategy, The Free Press, New York (1991). 17. T. S. Robertson and H. Gatignon, Technology development mode: a transaction cost conceptualization, Strategic Management Journal 19, 515e531 (1998); O. E. Williamson, Economic Institutions of Capitalism, The Free Press, New York (1985). 18. The use of different scales for novelty, originality and strategic value (1e3 scale) on the one hand and transferability (1e4 scale), on the other hand, is linked to the standardised internal ST procedures for classifying R&D projects. 232

Organising R&D Projects to Profit From Innovation

19. Non-parametric tests for differences in means provide equivalent results. 20. As a robustness check, we estimated a multinomial logit model considering four possible outcomes: no co-operation, co-operation with a firm, co-operation with universities and co-operation with both a firm and university. The results are presented in the appendix. The baseline case / comparison group is ‘no co-operation’. The model shows virtually equal results to the simple differences in means: basicness spurs co-operation, but the strategic importance of the project diminishes the propensity to co-operate. 21. C. Chung-Jen, The effect of knowledge attribute, alliance characteristics, and absorptive capacity on knowledge transfer performance, R&D Management 34(3), 311e322 (2004). 22. Here, we assume that the paradoxical aspect of competition and co-operation can be juxtaposed in order to understand the strategic phenomena characterising actors’ interactions. Actors co-operate in some activities and compete on other ones. 23. H. Smit and L. Trigeorgis, Strategic options and games in analysing dynamic technology investments, Long Range Planning 40(1), (2007). 24. M. A. Delmas, Exposing strategic assets to create new competencies: the case of technological acquisition in the waste management industry in Europe and North America, Industrial and Corporate Change 8, 635e652 (1999). 25. N. Argyres and T. Zenger, Capabilities, Transaction Costs, and Firm Boundaries: A Dynamic Perspective and Integration, Working paper (2007). 26. G. P. Pisano, The R&D Boundaries Of The Firm: An Empirical Analysis, Administrative Science Quarterly 35(1), 153 (1990); P. Mohnen and C. Hoareau, What type of enterprise forges close links with universities and government labs? Evidence from CIS 2, Managerial and Decision Economics 24(2,3), 133e145 (2003); K. Laursen and A. Salter, Open for innovation: the role of openness in explaining innovation performance among UK manufacturing firms, Strategic Management Journal 27(2), 131e150 (2006). 27. Hall, Link and Scott (2003). 28. G. King, R. Keohane and S. Verba, Designing Social Inquiry, Princeton University Press, Princeton (1994).

Biographies Bruno Cassiman is a Professor of Strategic Management at IESE Business School and a visiting professor at the University of Leuven and a fellow of CEPR and the SPSP research center at IESE Business School. He studies how innovation strategy and innovation management lead firms to develop a sustainable competitive advantage. He has written a number of articles on innovation for journals including The American Economic Review, Management Science, The European Economic Review, The International Journal of Industrial Organization, Managerial and Decision Economics and Research Policy. He has been a consultant for the European Commission on matters of innovation policy. [email protected] Maria Chiara Di Guardo is an Assistant Professor of Organisation at the University of Cagliari (Italy). Her research interests include the organisation of the innovation process and high-tech clusters dynamics. She has published in Small Business Economics and Management Research, among other outlets. [email protected] Giovanni Valentini is an Assistant Professor of Strategy and a fellow of EntER at Bocconi University. His research explores how to efficiently organise the innovation process as well as the relationship between firms’ corporate strategy, business strategy and innovation strategy. His articles have been published in Strategic Organization, Management Research, Small Business Economics, The International Journal of Production Economics and the Academy of Management Conference Best Papers Proceedings. [email protected]

Long Range Planning, vol 42

2009

233