Research Policy 47 (2018) 70–87

Contents lists available at ScienceDirect

Research Policy
journal homepage: www.elsevier.com/locate/respol

Innovation intermediaries and collaboration: Knowledge–based practices T

and internal value creation
⁎
Muthu De Silvaa,b, , Jeremy Howellsc, Martin Meyera
a
Kent Business School, University of Kent, Canterbury, Kent, CT2 7FS, UK
b
Birkbeck, University of London, Malet Street, London, WC1E 7HX, UK
c
Kellogg College, University of Oxford, 62 Banbury Rd, Oxford, OX2 6PN, UK

A R T I C L E I N F O A B S T R A C T

JEL classification: This paper analyses how knowledge-based practices adopted by innovation intermediaries enable them to
O300 generate value for themselves when collaborating with their clients. While the literature focuses on value
L10 creation for their client organisations, little is known about how innovation intermediaries create internal value
L14 even though this is essential for ensuring their long-term survival and sustaining their key facilitating role in the
M15
innovation system. This understudied issue is explored using empirical evidence from a sub-set of innovation
D830
intermediaries, Research and Technology Organisations (RTOs). The results indicate that by capitalising on
Keywords: existing knowledge vested in employees and collaborators as well as understanding and shaping the knowledge
Innovation intermediaries
base of the innovation ecosystem, innovation intermediaries generate internal value from their involvement in
Knowledge based practices
collaborative innovation, which range from different financial to non-financial types of value. Implications for
Value creation
European union intermediaries, their collaborators and for policymakers are then discussed.
Research and technology organisations
Knowledge capitalization

1. Introduction and supporting the partnership) engage in co-creative innovative ac-

tivity with collaborators, in a process of wider co-creation and co-de-
This paper investigates how knowledge-based practices adopted by velopment (Boon et al., 2011).
innovation intermediaries enable them to generate internal value for Whilst past research has extensively discussed the role of innovation
themselves when collaborating with their clients. Innovation has long intermediaries and how they generate value for their clients (Sawhney
been seen as a major factor in economic growth and development. In et al., 2003; Howells, 2006; Verona et al., 2006; Sapsed et al., 2007;
turn, innovation intermediaries have become recognised as key actors Nambisan and Sawhney, 2007; Arnold et al., 2010; Tran et al., 2011;
in the innovation landscape, improving firms’ innovative speed and Landry et al., 2013; Knockaert and Spithoven, 2014), there is a lack of
performance (Knockaert et al., 2014; Lin et al., 2016). Innovation in- understanding on what enables intermediaries to generate value for
termediaries, broadly defined as “organisations that provide a suppor- themselves in the context of their engagement in collaborative in-
tive role for collaboration between two or more parties during various novation. This is a non-trivial issue as business and innovation move
stages of the innovation process” (Howells, 2006, 721), are therefore towards more open and networked environments. ‘Who gets what’ in
seen to be central to creating and maintaining a successful innovation terms of value creation is a crucial issue in modern business dynamics
ecosystem (Sieg et al., 2010). The role of innovation intermediaries with actors operating in a system in which ‘we move from single-firm
extends from linking parties for collaboration, to setting up and med- revenue generation towards multi-firm control and interface issues’
iating relationships and bridging a wide array of knowledge, compe- (Ballon, 2007, 7). How to efficiently capture value for each actor within
tency and capability gaps (Smedlund, 2006; Klerkx and Leeuwis, 2008a; dispersed networks remains a daunting task, but is essential for the
Klerkx and Leeuwis, 2008b; Edler and Yeow, 2016). Among the varied long-term survival of not only businesses (Johnson et al., 2008), but
types of engagement by innovation intermediaries (Howells, 2006) also innovation intermediaries (del Águila-Obra et al., 2007; Lopez and
their interaction in collaborative projects represents one of their more Vanhaverbeke, 2009).
complex, enriched and involved roles as they (in addition to developing It is this aspect of the internal value creation by innovation

⁎
Corresponding author at: Kent Business School, University of Kent, Canterbury, Kent, CT2 7FS, UK.
E-mail addresses: muthumalie@yahoo.com, l.de-silva@kent.ac.uk (M. De Silva), jeremy.howells@kellogg.ox.ac.uk (J. Howells), M.S.Meyer@kent.ac.uk (M. Meyer).

http://dx.doi.org/10.1016/j.respol.2017.09.011
Received 16 November 2016; Received in revised form 23 September 2017; Accepted 25 September 2017
Available online 10 October 2017
0048-7333/ © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
M. De Silva et al. Research Policy 47 (2018) 70–87

intermediaries, which the literature has not yet been able to adequately knowledge-based practices depending on which type of value they in-
address (Krenz et al., 2014), that is the focus of this paper. The internal tend to generate; (2) indirectly, by allowing innovation intermediary
value could be multidimensional comprising both financial and non- clients’ to better identify and appreciate which aspects of their colla-
financial values. The immediate gain for innovation intermediaries boration intermediaries benefit from the most; and, (3) for policy-
from client engagement can be expected to be from financial benefits in makers to put in place better measures to support intermediaries as a
terms of: (a) revenues generated by services they provide for clients, or pathway for wider innovation ecosystem value generation.
(b) funding secured from research grants. Apart from these financial The remainder of this paper is organised as follows. The next section
gains, intermediaries also generate non-financial value that consists of outlines a set of hypotheses developed to indicate how different
knowledge, market and network based benefits. In collaborative projects knowledge-based practices may determine an innovation inter-
intermediaries with other partners may develop new knowledge mediaries’ ability to generate both financial and non-financial value.
(Mowery et al., 1996; Gulati, 1999; Kale et al., 2000), capabilities re- This will be followed by methodology section and the paper then
quired to deploy such knowledge (Hardy et al., 2003; Cepeda and Vera, concludes by discussing the results, limitations and future research di-
2007; Martín-de Castro, 2015), intellectual property (Martín-de Castro rections of the study.
2015), new forms of innovation (Earl, 2001) and useful networks
(Nagaoka and Kwon, 2006). Although this, taken to the extreme, might 2. Background literature
be seen as parasitic or exploitative of their clients, innovation inter-
mediaries need to generate sufficient gains for their long-term survival, 2.1. Internal value generation by innovation intermediaries through
without which the generation of value to their clients and their wider collaborative projects
systemic benefits would not be possible.
In the ability of innovation intermediaries to generate internal Innovation intermediaries have been characterised as organisations
value, the effective development and utilisation of knowledge is a cri- that generate value to other actors within a system of innovation
tical component (Wenger and Snyder, 2000; Teece, 2004; Alavi et al., (Sawhney et al., 2003; Verona et al., 2006; Nambisan and Sawhney,
2005; Knockaert et al., 2014), especially in the context of collaborative 2007; Arnold et al., 2010; Tran et al., 2011). The focus has therefore
innovation (Pisano and Teece 2007; Lichtenthaler and Lichtenthaler, been on studying how intermediaries enable their clients to leverage
2009). The knowledge-based view of the firm (Kogut and Zander, 1996; external technologies (Howells, 2006), existing design solutions
Spender, 1996) as well as the open innovation literature (Chesbrough, (Hargadon, 2002), the knowledge and experience of customers (Verona
2003; Chesbrough, 2006; Dahlander and Gann, 2010) highlight that et al., 2006), the expertise of external specialists (Tran et al., 2011) and
valuable knowledge exists not only within organisational boundaries, exchange platforms (Lichtenthaler and Ernst, 2008). Nevertheless, in
but also outside the firm. As such, firms’ ability to explore, acquire, order for innovation intermediaries to successfully perform these tasks
retain, integrate and exploit knowledge (Grant, 1996), is central to firm they need to generate internal value for themselves from such en-
value creation (Gold et al., 2001; Klerkx and Leeuwis, 2008c; gagement with clients. ‘Internal value’ is defined as the sum of both
Lichtenthaler and Lichtenthaler, 2009; Knockaert and Spithoven, financial and non-financial values generated from their clients by in-
2014). novation intermediaries. The internal value generated by inter-
Where do innovation intermediaries come into this process of mediaries during collaborative projects will be multi-dimensional,
knowledge management and coordination? The agency issue and role of comprising both (a) financial and (b) non-financial gains (Huizingh,
cooperation and coordination is complex and not easy to resolve 2011).
(Holmstrom, 1989), but a number of studies have started to explore the As noted earlier, the immediate gain for innovation intermediaries
role of intermediaries in the knowledge creation and innovation pro- from client engagement can be expected to be from financial benefits, in
cess. They cover knowledge search, problem solving and connecting terms of service revenue generation and funding secured from research
and coordinating knowledge between actors (Agogué et al., 2013a; grants, but there will also be non-financial value generation. Non-fi-
Hargadon and Sutton, 1997) through to commercialising new tech- nancial value, associated with the concept of ‘social capital’, highlights
nologies (Aldrich and von Glinow, 1992). Thus, the successful provision the central role played by the social network in which actors, such as
of innovation intermediaries’ services involves the handling of complex innovation intermediaries, engage. It is associated with cooperation,
knowledge (Klerkx and Leeuwis, 2008c; Lichtenthaler and reciprocity and information sharing in the value creation process (Tsai
Lichtenthaler, 2009). Innovation intermediaries therefore act as and Ghoshal, 1998; Landry et al., 2002; McElroy, 2002; de Felice,
knowledge repositories that introduce new combinations of knowledge 2015). For instance, in collaborative projects intermediaries with other
and also make knowledge-based contributions when providing solu- partners may develop new knowledge (Mowery et al., 1996; Gulati,
tions to their clients (Hargadon and Sutton, 1997; Howells, 2006). In- 1999; Kale et al., 2000) and technical and operational capabilities re-
novation intermediaries appear to be developing new methods and quired to deploy such knowledge (Hardy et al., 2003; Cepeda and Vera,
practices in more unknown knowledge environments where risk and 2007; Martín-de Castro, 2015). This new knowledge will be important
uncertainty are high and where sophisticated management principles for more effective running of the intermediation process, particularly
have to be developed (Agogué et al., 2013b). Thus, the successful de- when new combinations of knowledge are introduced. Innovation in-
livery of the intermediation process requires the effective adoption of termediaries’ engagement in collaborative projects would also enable
knowledge-based practices, through which internal value is generated. them to generate ‘closer’ to market benefits. This includes the devel-
Even though there is extensive research on investigating how to manage opment and improvement of products, processes (Earl, 2001), services,
knowledge effectively in collaborative arrangements (Kale et al., 2000; technologies, equipment and intellectual property (Martín-de Castro,
Grant and Baden-Fuller, 2004), the positioning of innovation inter- 2015) that provide them with competitive advantage (Gulati et al.,
mediaries in these studies has been as institutions that help other or- 2000b), particularly when bridging technology and market gaps be-
ganisations to manage knowledge and extract value (Lee et al., 2010; tween partners (Lichtenthaler and Ernst, 2008; Nell and Lichtenthaler,
Spithoven et al., 2010) rather than on the intermediaries themselves. 2011; Lichtenthaler, 2013; Melle and Russo-Spena, 2015). Lastly, in-
Using data from a survey of Research and Technology Organisations termediaries also develop new networks of contacts including new
(RTOs), an important form of innovation intermediary (Howells, 2006; markets and distribution channels (Gulati et al., 2000a; Nagaoka and
Oxford Economics, 2008; Arnold et al., 2010; Landry et al., 2013; Kwon, 2006; Lichtenthaler, 2007). These are of value when finding
Miller, 2014), the paper seeks to contribute in three ways to our partners for collaboration and providing support services, such as in-
knowledge of innovation intermediaries and their clients: (1) for in- troducing new distribution channels to clients and international market
novation intermediaries to more effectively assess and design access (Arnold et al., 2010).

71
M. De Silva et al. Research Policy 47 (2018) 70–87

2.2. The importance of adopting knowledge-based practices Capitalising on knowledge accumulated in trustworthy relationships
increases the reliability, breadth and depth of tacit and codified
Knowledge-based practices adopted by innovation intermediaries knowledge available to an organisation (Nielsen, 2005), improves an
can influence their ability to generate financial and non-financial value organisation’s innovative performance (Martín-de Castro, 2015) and
from collaborative projects (Wenger and Snyder, 2000; Gold et al., enhances their market advantage (Yli-Renko et al., 2001). Knowledge
2001; Alavi et al., 2005; Martín-de Castro, 2015). The knowledge-in- capitalisation also helps address common causes for failure in colla-
tensive role of innovation intermediaries (Aldrich and von Glinow, borative projects by reducing cost and uncertainty (Granovetter, 1985;
1992; Klerkx and Leeuwis, 2008a,b; Mount et al., 2015) heightens the Vlaar et al., 2007) as well as the ‘psychic’ distance between partners
need for them to systematically make use of external and internal (i.e., the degree of factors that prevent the flow of information between
sources of knowledge, such as employees, clients, collaborators and partners; see, for example, Bruneel et al., 2010; Bruneel et al., 2015),
business (Lin et al., 2012), by integrating these effectively within their whilst increasing trust (Paulraj et al., 2008) which is vital for inter-
organisational and innovation processes (Escribano et al., 2009). mediaries to generate internal value.
Nevertheless, managing knowledge effectively for sustainable value Nevertheless, this may not be a wholly positive process; there are a
generation is a complex process (Gold et al., 2001) since it requires the number of negative aspects and problems associated with this focus.
adoption of human capital associated knowledge-based practices, re- Capitalising on existing knowledge could increase in-breeding, reducing
cognised as crucial for collaborative innovation success (Gassmann the innovativeness of an organisation (Saviotti, 1996) and hence the
et al., 2010; Sieg et al., 2010; Boari and Riboldazzi, 2014). The re- long-term ability to generate value through collaboration (Sethi et al.,
cognition of the need for increased knowledge task specialization was 2002). A greater reliance on the knowledge of existing individuals
noted by Adam Smith (1776) who stressed that the most fundamental (whether they are employees or collaborators) also increases the risk of
aspect of the division of labour was the division of knowledge and the knowledge loss to intermediaries since these individuals could always
associated combinatorial ways of producing knowledge (Metcalfe, change their association or, even worse, join competitors (Davenport
2002). Invention, in particular, requires individuals (Rosenberg, 1965, et al., 2002; Oliver, 2004), diminishing the ability to generate internal
132–133) to draw upon the diverse fields of knowledge and therefore value through collaboration. Yet, when considering the role of in-
the know-how needed to perform complex tasks can be very divided novation intermediaries as ‘bridgers’ of knowledge between organisa-
(Nelson, 2003, 911). Successful knowledge management with regard to tions and those that introduce new combinations of knowledge – which
innovation must therefore be centred on the individual (Howells, 2012, necessitates them to make the most out of their existing trustworthy
1005) as meaning and knowledge is ultimately based within individual knowledge – (Hargadon and Sutton, 1997; Howells, 2006), it is possible
minds (Polanyi, 1961; Morton, 1997). to suggest that they would benefit from adopting practices for knowl-
edge capitalisation regardless of potential negativities:
3. Theoretical framework and hypotheses
H1. The adoption of practices for knowledge capitalisation positively
influences the generation of financial and non-financial value for
On the basis of the above discussion, this section proposes six hy-
innovation intermediaries in collaborative projects.
potheses to investigate how knowledge-based practices enable in-
novation intermediaries to generate internal value for themselves.
These six knowledge bases practices encompass mechanisms associated 3.2. Practices for knowledge advancement
with managing knowledge vested in ‘individuals’ within intermediaries
and their innovation ecosystem. They are: (1) ‘knowledge capitalisa- In addition to exploiting current knowledge, acquiring new knowl-
tion’ associated with leveraging and recombining existing knowledge, edge, absorbing and internalising it (Garud and Nayyar, 1994; Smith
both internally and externally between employees, partners and wider et al., 2005; Knockaert et al., 2014) is essential for advancing and de-
networks; (2) ‘knowledge advancement’ related to acquiring new veloping new business opportunities as well as helping to avoid lock-in
knowledge, absorbing and internalising it; (3) ‘knowledge spanning’ (Saviotti, 1996). This capability is termed here as ‘knowledge ad-
covering the ability of an innovation intermediary to access unrelated vancement’. By being able to identify knowledge gaps through mapping
knowledge that is non-adjacent to an intermediary’s knowledge base; current knowledge and having the capability and flexibility to integrate
(4) ‘knowledge worker empowerment’ associated with practices to new with existing knowledge are key prerequisites for knowledge ad-
empower employees engaged in collaboration as to how to use vancement (Lichtenthaler and Ernst, 2006). Examples of practices in
knowledge; (5) ‘innovation ecosystem knowledge access’ is related to knowledge advancement include adopting a structured approach to-
the ability of an intermediary to possess knowledge of its wider in- wards actively exploring and internalising new knowledge, hiring new
novation ecosystem; and, (6) ‘innovation ecosystem knowledge staff who possess the required new knowledge (Earl, 2001;
shaping’ is associated with practices and routines aimed at shaping the Lichtenthaler and Lichtenthaler, 2009) and better team-based learning
overarching collaborative framework and institutional form of the in- techniques both within and across organisational boundaries
novation system. Each of these hypotheses will now be explored in (Scardamalia, 2002). Advancing knowledge is considered critically
more detail (Fig. 1). important for sustainability in an era of short product life cycles, ad-
vanced technological developments and considerable economic un-
3.1. Practices for knowledge capitalisation certainty (Bowonder and Miyake, 2000).
However, since innovation intermediaries are mainly involved in
Leveraging and recombining existing knowledge, both internally and handling and integrating knowledge bases of innovation ecosystem
externally between employees, partners and wider networks, termed actors by exploiting current knowledge (i.e. their positioning as the
here as ‘knowledge capitalisation’, is found as essential if an organisa- ‘bridgers’ of basic and applied science; Hales, 2001), it is possible to
tion, such as an innovation intermediary, is to generate value from argue that knowledge advancement will not have significant impacts on
collaborative innovation (Davenport et al., 1998; Gold et al., 2001). their value creation. There are also problems in relation to the extent to
‘Knowledge capitalisation’ practices include: retaining experienced which an organisation could correctly identify knowledge gaps and
staff; having a dedicated team working on specific tasks; capitalising on implement an approach to fill them due to the organisation being in a
personal networks of staff and business-to-business relationships; and, constant state of flux. Still, since innovation intermediaries act as net-
engaging with partners with positive working experience (Hobday, work and knowledge integrators, they need to constantly update and
2000; Lichtenthaler and Lichtenthaler, 2009; Dahlander and Gann, advance their knowledge base (Klerkx and Leeuwis, 2008a,b,c;
2010). Lichtenthaler and Lichtenthaler, 2009; Mount et al., 2015), without

72
M. De Silva et al. Research Policy 47 (2018) 70–87

Fig. 1. Influence of knowledge-based practices on internal value

creation by intermediaries.

which the identification of opportunities for new knowledge re- of employee empowerment on how they develop and use knowledge
combination would be impossible. Based on this particular importance can lead to employee dissatisfaction and demotivation (Ford and
and the positive effects of this practice, the second hypothesis is: Fottler, 1995). Hence, empowering knowledge workers by adopting
practices that engender positively motivated staff (Gold et al., 2001;
H2. The adoption of practices for knowledge advancement positively
Lichtenthaler and Ernst, 2006) is seen to be an effective practice for
influences the generation of financial and non-financial value for
innovation success (Argyris, 1998; Collis and Moonen, 2008).
innovation intermediaries in collaborative projects.
Nevertheless, there is the danger that employee empowerment
could potentially dilute the achievement of organisational outcomes
(Jensen 2001). Since intermediaries work with several external orga-
3.3. Practices for knowledge spanning
nisations, empowering employees may result in coordination problems,
as staff adopt individualistic or temporary practices, which do not
The ability of an innovation intermediary to access new, unrelated
conform to more standard or transferable rules between individual staff
knowledge that is non-adjacent to an intermediary’s knowledge base
and their team members. This can often lead to a loss of control,
through ‘knowledge spanning’ is also an important competence for an
thereby inducing opportunism and increased uncertainty (Mils and
organisation (Grant and Baden-Fuller, 2004). External unrelated
Ungson, 2003). Hence, while the adoption of these types of practice is
knowledge is often required for a specific project (Earl, 2001), the
important for value creation in collaborative projects, empowerment
complexity of which necessitates multi-disciplinary knowledge bases of
could also result in the inability of an organisation to coherently
different parties to be combined (Curley and Salmelin, 2013; Perkmann
achieve intended objectives. This leads to the fourth hypothesis:
and Schildt, 2014). Intermediaries often access technical knowledge
from universities or other organisations that possess ‘unrelated’ H4. The adoption of practices to empower knowledge-based workers
knowledge bases to achieve common goals for the success of a specific negatively influences the generation of financial and non-financial
project. Since organisations engage in technical knowledge access are value for intermediaries in collaborative projects
not inclined to internalise partners’ knowledge, they remain specialists
in their own domains of knowledge (Grant and Baden-Fuller, 1995; Lui,
2009).
3.5. Practices for access to innovation ecosystem knowledge
Whilst projects that involve integrating ‘unrelated’ or multi-
disciplinary knowledge are found to generate several benefits (Rhoten
A successful intermediation role requires having a better awareness
and Pfirman, 2007), accessing technical knowledge is not been without
of the needs, resources and competencies of a wide array of actors of a
its challenges. This is due to difficulties faced by being unaware of the
system of innovation and their institutional framework (Howells and
knowledge bases of other collaborators, which in turn leads to high
Roberts, 2000; Muller and Zenker, 2001; Kauffeld-Monz and Fritsch,
transaction costs, uncertainty of outcomes and risk of failure (Zukin and
2013), which is termed here as ‘innovation ecosystem knowledge’. The
DiMaggio, 1990; Petruzzelli and Rotolo, 2015). Particularly when
practices adopted by innovation intermediaries to access innovation
considering the role of innovation intermediaries, it could be assumed
ecosystem knowledge include using digital platforms that have in-
that such challenges would be at a minimum since intermediaries are
formation on the interests, knowledge and competences of potential
inherently successful at working with a wide array of actor types. On
partners as well as actively developing new relationships with key
this basis, the third research hypothesis is:
players (Howells and Roberts, 2000) through which intermediaries
H3. The adoption of knowledge spanning practices positively influences could enhance their awareness of the ecosystem. These practices and
the generation of financial and non-financial value for innovation routines for innovation-ecosystem knowledge access differ from those
intermediaries in collaborative projects. for knowledge capitalisation (H1) and knowledge advancement (H2).
The former involves enhancing awareness of ecosystem through new
relationship building, whereas the latter involve in making use of
3.4. Knowledge worker empowerment knowledge vested in pre-existing relationships (H1) and in acquiring
‘new’ knowledge to advance and fulfil ‘knowledge gaps’ by recruiting
There is a danger of having an exclusive emphasis on organisational new staff (H2), respectively.
level factors and disregarding individual level determinants. Since in- Nevertheless, the benefit of accessing the innovation ecosystem
novation is a multi-level phenomenon (Gupta et al., 2007), the attitudes knowledge base using external sources and actively attempting to de-
of individuals/employees engaged in collaboration influence the overall velop new relationships may be questionable as new knowledge/re-
success of the organisation (Gavetti, 2005). Whilst top-down approache lationships might not be reliable or prove to be successful (Davenport
ensures the whole organisation to coherently achieving its vision, lack et al., 1998; Dyer and Nobeoka, 2000). Hence, the chances of benefiting

73
M. De Silva et al. Research Policy 47 (2018) 70–87

from the time and effort made to access the knowledge of the innova- government funding which supply services to firms individually or
tion ecosystem might be low. Yet, developing knowledge about the collectively in support of scientific and technological innovation and
wider innovation ecosystem is vital to intermediaries in order to iden- which devote much of their capability to remaining integrated with the
tify new opportunities to combine knowledge, mitigate uncertainty science base”. Most RTOs, such as the Fraunhofer-Gesellschaft, were
about profitability of collaboration, assess the value of available in- created to facilitate the transfer of knowledge from the science base to
tellectual property and to bring suitable parties together for colla- firms and have an applied research focus (van Lente et al., 2003). The
boration (Kodama, 2008). On this basis, the fifth hypothesis is: majority of RTOs operate between a technical science base, on the one
hand, and manufacturing industries, on the other; what they define as
H5. The adoption of practices for innovation ecosystem knowledge
“hard intermediary” functions, such as technology testing. However,
access positively influences the generation of financial and non-
the functions of RTOs have broadened over time (Miller, 2014) to in-
financial value for innovation intermediaries in collaborative projects
clude more soft forms, such as business service activities around net-
working and consultancy (EARTO, 2017, 1). Research has highlighted
3.6. Practices for shaping the innovation ecosystem knowledge base that there is no single ‘ideal type’ RTO, and that each must be tailored
to its innovation environment (Miller, 2014). Thus, RTOs are a sig-
Collaboration is difficult when partners have diverse interests, goals nificant type of innovation intermediary and key actors within the
and motivations. One way of overcoming this is through shaping the wider European system of innovation. RTOs receive about 30% of their
interests of actors within an innovation system to increase the chances income from government, 30% from competitive public and private
of reaching a shared understanding and mutuality between the parti- grants and the rest from industry as contract income (Martínez-Vela,
cipating actors which is important for successful collaboration (West 2016). RTOs also coordinated a third of all EU research and innovation
and Gallagher, 2006; Wallin and von Krogh, 2010; Tjong et al., 2015). projects (Arnold et al., 2010) and led some 258 Horizon 2020 projects
Practices and routines aimed at sustaining influence over potential in 2015 (EARTO, 2017, 7). RTOs engage in a range of tasks in these EU
collaborators and influential decision makers are seen as the ‘political’ projects, ranging from bringing parties together for collaboration to
side of collaboration (Hardy and Phillips, 1998) During this process, providing applied research to the consortium. A database of RTOs
innovation intermediaries externalise relevant knowledge to influence compiled by the Big Innovation Centre has identified that there are 122
actions and interests of potential partners (Arnold et al., 2010). Thus, European RTOs in total distributed across the seven EU member states
innovation intermediaries, in collaboration with other actors in the with eight types of organisation (Andersen and Blanc, 2013) was the
innovation system, often engage in helping to shape the strategic policy empirical base for the survey.
direction, which results in convergence around the interests of actors A mixed method approach to improve the validity of the overall
within the region or nation (Bouwen, 2002). Some even argue that study (Tashakkori and Teddlie, 1998; Bisbe et al., 2007) was adopted
developing consensus is one of intermediaries’ key functions with three main stages: (1) an initial qualitative data gathering stage;
(Etzkowitz, 2008; Meyer et al., 2017). (2) an online, quantitative survey; and, (3) a follow-up, in-depth in-
Nevertheless, there is no direct evidence to suggest that shaping the terview phase. The data gathered through initial interviews were
ecosystem would help intermediaries to generate internal value checked against theoretical explanations to validate the conceptual
through collaborative projects (see Mazey and Richardson 2006) and framework and to develop the questionnaire survey, which was im-
indeed they may generate negative impacts by exposing intermediaries’ portant since the paper addresses an underexplored topic (Edmondson
knowledge too much to external parties, with intermediaries losing and Mcmanus, 2007; Autio et al., 2013). A total of 68 responses were
their competitive advantage (Dyer and Nobeoka 2000). Yet, con- received back, of which 59 were then usable after data cleaning, re-
sidering the specific role of intermediaries as those who bring together presenting a response rate of 48.3%. The non-response bias test re-
partners with different knowledge bases (Agogué et al., 2013a; vealed that there is no significant difference between respondent RTOs
Hargadon and Sutton, 1997), it could be argued that these practices and the full empirical base of 122 RTOs in terms of type of centre [X2(5,
aimed at shaping the knowledge base of the ecosystem increase the 179) = 2.217, p = 0.818 > 0.05], sector of operation of RTOs [X2(3,
chance of securing projects and finding partners with common interests, 179) = 1.546, p = 0.672 > 0.05], turnover of RTOs (i.e. categorised
thereby leading to value generation. This leads to the final hypothesis: as small medium and large) [X2(2, 179) = 3.308, p = 0.191 > 0.05],
and number of employees (i.e. categorised as small medium and large)
H6. The adoption of practices that shape the knowledge base of an
[X2(2, 179) = 1.272, p = 0.529 > 0.05].
innovation ecosystem positively influences financial and non-financial
The third phase interviews generated context-specific and in-depth
value generation by intermediaries in collaborative projects
understanding of the causality derived from the survey data
These six hypotheses were then used in the study to investigate how (Downward and Mearman, 2007). Finally, a validation event with
practices adopted by innovation intermediaries enable them to generate survey participants and a group of their collaborators, representing the
financial and non-financial value. Please note that, considering the ecosystem, was conducted to further verify the study findings. The
exploratory nature of this study, signs are merely preliminary ex- adoption of this sequential mixed method design improved the validity
pectations about the causality (Fig. 1). and reliability of the overall study.

4. Methodology 4.2. Variable construction

4.1. Methodological framework The main source of data was derived from the online survey and the
findings of in-depth interviews were used to validate the former (Kim
The empirical base is a study of EU based RTOs that investigated and Miner, 2007; Autio et al., 2013). In addition, several control vari-
how knowledge-based practices adopted by innovation intermediaries ables capturing the characteristics and innovation approaches of in-
influence the generation of financial and non-financial value during novation intermediaries were also included in the final model.s
their engagement in EU funded projects, undertaken in collaboration
with businesses, universities, public research organisations and RTOs 4.2.1. Dependent variables
(Andersen and Blanc, 2013). RTOs are a particular form of innovation The dependent variables are the four types of internal value in-
intermediaries with origins around public ownership and the need to novation intermediaries generate for themselves in collaborative pro-
support the transfer of knowledge between science and industry. Hales jects, namely, financial value and the three non-financial of knowledge,
(2001, p.4) has defined them as “organisations with significant core market and network based value. The variable on financial value was

74
M. De Silva et al. Research Policy 47 (2018) 70–87

Table 1 based practices is those adopted during the EU Framework Programme

Variable construction of non-financial value: questionnaire items. 7 (FP7) ran from 2007 to 2013. These practices were presented to the
respondents who were requested to state to what extent they would
Non-financial values Factor loading
adopt each practice on a 1–5 Likert scale. In relation to ‘innovation
Question: Please tick on a Likert Scale of 1–5 (1- Not at all, 2- Very low, 3- Low, 4- High, ecosystem knowledge influence’ it was evident during initial interviews
and 5- Very high) to indicate the extent to which your organisation generated following that different RTOs use different routes to shape the knowledge base of
benefits (i.e. internal value) when engaging in EU FP 7 projects during last three years. the European system of innovation, which may differently influence
Knowledge based internal value (Cronbach's Alpha = 0.731; CR = 0.748;
value creation. A total of 18 such sources were identified and re-
AVE = 0.743) spondents were requested to state the extent of use on a Likert scale of
Developed new and advanced knowledge 0.820 1–5. The reliability and validity of the measures were further confirmed
Developed other disciplinary knowledge 0.745 by empirical evidence. The first five variables of knowledge routines
Developed new research based capabilities required 0.734
and practices were derived by performing a Principal Component
to deploy new knowledge
Developed new management capabilities required to 0.673 Analysis (PCA) on the 13 practices. Five components with Eigen
deploy new knowledge value ≥ 1 were identified (Table 2). The Kaiser-Meyer-Olkin Measure is
Market based internal value (Cronbach's Alpha = 0.824; CR = 0.816; AVE = 0.737) 0.655 (> 0.5); indicating that the patterns of correlations are relatively
Developed new equipment and technologies 0.789 compact and factor analysis should yield distinct and reliable factors.
Developed products and services 0.772 The five components explain 70.25% of variance (i.e. C1–27.8%,
Produced intellectual property (e.g. patents) 0.759 C2–14.03%, C3–11.46%, C4–9.4%, C5–7.48%). The CR and AVE in-
Developed capabilities to engage in ‘closer’ to market 0.720
dices being more than 0.5 further confirms the reliability of the mea-
research
Developed capabilities to help industry to pursue 0.715 sures and AVE also confirms the convergent validity (Table 2).
their market-oriented goals A second PCA was conducted to construct variables for the sixth
Developed capabilities to help industry to gain 0.668 knowledge-based practice, ecosystem knowledge influence, which re-
commercial knowledge
sulted in identifying five components with high factor loading
Network based internal value (Cronbach's Alpha = 0.645; CR = 0.684; AVE = 0.722) (i.e. > 0.5). Kaiser-Meyer-Olkin Measure of Sampling Adequacy was
Gained access to global markets 0.756 0.763 and 78.307% variance was explained by the five components.
Developed networks of future value 0.793
Developed capabilities to help industry to access 0.617
The first component (Table 3) was identified as influencing through
international markets active involvements in EU activities and membership-based network
organisations (for example, trade associations), the second component
represents influence exerted from written media communication, the
derived by requesting respondents to state the average annual income third component indicates participating in expert advisory groups, the
generated by their organisation (i.e. with respect to a particular centre, fourth the influence made through collaboration with external bodies
but not the group of centres) through EU Framework programmes (FP) (such as EU liaison offices and universities) and the fifth component
during last three years. Since the data was collected in 2014, the period illustrates the influence felt by the use of media events. It was analysed
covered was 2011–13. Similarly, focusing on the same time span, how the level of use of these five types of routes determine inter-
variables for the three non-financial types of value were constructed by mediaries’ ability to generate financial and non-financial value.
requesting respondents to state the extent to which they have generated
specific benefits as a result of their engagement in EU projects on a 4.2.3. Control variables
Likert Scale of 1–5. These specific benefits representing each type of Several control variables capturing the characteristics of innovation
non-financial value were identified by checking the findings of initial intermediaries and their innovation approaches were used. The general
interviews against the theoretical framework. These responses clearly characteristics of innovation intermediaries may influence value gen-
supported the conceptual development of three non-financial values eration in EU projects. Turnover (i.e. average annual turnover during
(Section 2.1), thus used in the questionnaire survey to gather in- last three years; Hauser, 2010), age (as an indicator of experience;
formation on internal value creation. Bruneel et al., 2010), sector of operation, and number of employees
In terms of the reliability and validity of the measures internal engaged in EU activities (Fontana et al., 2006; Kirkels and Duysters
consistency and uni-dimensionality were tested to assess the possibility 2010), together with variables covering RTOs’ FP7 engagement, were
of generating a single score for each non-financial value by aggregating all controlled for. Thus, the extent to which innovation intermediaries
the ratings of different benefits. The factor analysis resulted in gen- collaborate with different actors in an ecosystem may determine their
erating only one factor with an Eigen value more than 1 for knowledge ability to generate value through collaboration (Lee et al., 2010). For
(2.22, 55.50%), market (3.270, 54.5%) and network (1.581, 52.71%) instance, collaborating with large firms provides innovation inter-
based value (Table 1). Cronbach's Alpha for knowledge, market and mediaries with more economic advantages compared with universities
network based values were 0.731, 0.824, and 0.645 respectively, in- and small firms (Wright et al., 2008; Huizingh, 2011; Howells et al.,
dicating an acceptable level of internal consistency among items (i.e. 2012) due to large firms’ capacity to develop long-term relationships
benefits) used to construct three variables. The Composite Reliability and the size of their resource base (Huizingh, 2011). Nevertheless, large
(CR) (Chin, 1988) and Average Variance Extracted (AVE) (Fornell and firms may have less need for intermediaries. By contrast, small and
Larcker, 1981) indices being more than 0.6 and 0.7 respectively further medium-sized enterprises (SMEs) and universities may have a higher
confirms the reliability of the measures and AVE also confirms the reliance on innovation intermediaries (Yusuf, 2008; Lee et al., 2010;
convergent validity. All these results indicate the suitability of variables Howells et al., 2012) due to their need to overcome organisational and
constructed to measure three types of non-financial value. cultural issues of dealing with increased external contacts (Van de
Vrande et al., 2009). Hence, intermediaries may have a greater chance
4.2.2. Independent variables to interact with SMEs and universities. Past studies on RTOs suggest
Independent variables to represent five of the six knowledge-based that RTOs differ in terms of their collaborators. Whilst some RTOs seem
practices were constructed using 13 practices identified from the lit- to collaborate with large firms to a greater extent, other RTOs seem to
erature (Sections 3.1–3.6), validated through initial interviews. During focus on developing a stronger network with local small businesses
initial interviews it was evident that the practices adopted by RTOs (Albors et al., 2014), which would indeed impact on the type of internal
would not drastically change at least during one EU framework pro- value they generate. Hence, the extent to which innovation inter-
gramme. Since the data was gathered in 2014, the data on knowledge- mediaries collaborate with different actors in an EU project was

75
M. De Silva et al. Research Policy 47 (2018) 70–87

Table 2
PCA on the use of knowledge-based practices: questionnaire items.

Statements Component

1 2 3 4 5

Knowledge capitalisation (Cronbach's Alpha = 0.794; CR = 0.856; AVE = 0.549)

1. Capitalise on the personal networks of staff when engaging in EU Framework programmes 0.889 −0.082 0.029 −0.018 −0.023
2. Have a dedicated team working on EU engagements 0.785 −0.008 −0.155 0.289 0.038
3. Make an effort to retain those employees with successful EU experience 0.725 0.334 0.142 0.036 0.056
4. In EU consortia, engage with partners with whom you have good relationships 0.691 0.105 0.171 −0.173 −0.115
5. Capitalise on relationships with local/national government 0.578 −0.060 0.125 0.469 0.259

Knowledge advancement (Cronbach's Alpha = 0.554; CR = 0.616; AVE = 0.802)

6. Engage in advancing core competences 0.249 0.843 −0.103 −0.095 −0.091
7. Recruit a portfolio of employees with different skills (e.g. basic research, applied research, management, etc.) as a strategy −0.136 0.760 0.308 0.205 0.253
to advance knowledge and skills

Knowledge spanning (Cronbach's Alpha = 0672; CR = 0.614; AVE = 0.796)

8. Accessing university knowledge to achieve project outcomes 0.231 −0.111 0.868 0.203 0.022
9. Engage with partners from different disciplines to make use of their knowledge to achieve project outcomes −0.016 0.339 0.723 −0.081 −0.259

Innovation ecosystem knowledge access (Cronbach's Alpha = 0.591; CR = 0.5870; AVE=0.711)

10.Use cross EC networks such as ERRIN and other digital platforms to find potential partners and to understand them better −0.134 −0.036 0.246 0.771 −0.032
11.Develop relationships with key business players and other potential partners in the market 0.283 0.150 −0.215 0.651 −0.252

Knowledge worker empowerment (Cronbach's Alpha = 0.549; CR = 0.553; AVE = 0.618)

12. Your organisation’s staff has freedom of deciding their EU engagements (e.g. what to engage in, with whom to engage in, −0.069 0.003 −0.218 −0.238 0.722
how to engage in etc.)
13.Provide freedom for your employees to work in your member organisations in the form of a placement 0.449 0.270 0.131 0.287 0.514

Extraction Method: Principal Component Analysis.

Rotation Method: Varimax with Kaiser Normalization.
Rotation converged in 9 iterations.

controlled for. both the approaches) each of which have different consequences on
Second, the innovation strategy adopted by innovation inter- value creation (Lichtenthaler and Ernst, 2008). Indeed, research by
mediaries may also influence value creation (Kirkels and Duysters Polzin et al. (2016, 41–42) has shown a key function now of innovation
2010). Whilst traditionally innovation intermediaries have been known intermediaries is in reducing uncertainty about future market oppor-
for adopting ‘technology push’ approach, there has been an increasing tunities for new technologies and in niche market development for
trend towards adopting ‘market pull’ approach (or a combination of them. Respondents stated the extent to which they adopt ‘technology

Table 3
PCA on the use of sources to influence ecosystem knowledge: questionnaire items.

Statements Component

1 2 3 4 5

Direct EU involvements and memberships (Cronbach's Alpha = 0.910; CR = 0.830; AVE = 0.699)
1. Meeting EU policy makers 0.750 0.086 0.413 0.102 0.261
2. European Technology Platforms and structures associated with Public Private Partnerships 0.621 0.160 0.385 0.174 0.327
3Participating in EARTO 0.603 −0.052 0.508 0.282 0.282
4. Participating in membership networks other than EARTO 0.512 0.121 0.604 0.111 0.354
5. Participating in EU investment plan development groups 0.784 0.232 0.160 0.176 0.025
6. Direct tenders from EC – Policy related 0.727 0.193 0.398 0.196 −0.021
7. Through trade associations 0.881 0.127 0.086 0.054 0.067

Written communication (Cronbach's Alpha = 0.883; CR = 0.859; AVE = 0.786)

8. Through joint publications −0.018 0.579 0.352 0.146 0.552
9. Through press releases 0.240 0.842 0.077 −0.025 0.328
10. Through newsletters 0.077 0.917 0.095 0.126 0.081
11. Social media (e.g. Twitter, Websites, Blogs, etc.) 0.257 0.805 0.056 0.139 0.211

Expert advisory groups (Cronbach's Alpha = 0.713; CR = 0.618; AVE = 0.805)

12. Participating in expert panels/workgroups 0.282 0.077 0.764 0.019 0.091
13. Participating in selecting/referee panels 0.279 0.148 0.845 0.033 −0.065

Through external bodies (Cronbach's Alpha = 0.801; CR = 0.712; AVE = 0.825)

14. Through your business and university network of contacts 0.233 0.126 0.030 0.772 0.148
15. Through connections with local or national governments 0.078 −0.052 −0.028 0.891 −0.073
16. Through EU liaison offices 0.129 0.216 0.183 0.812 0.081

Media events (Cronbach's Alpha = 0.761; CR = 0.620; AVE = 0.816)

17. Through conferences 0.070 0.320 0.212 0.132 0.796
18. Through industry fares 0.264 0.227 −0.102 −0.040 0.836

Extraction Method: Principal Component Analysis.

Rotation Method: Varimax with Kaiser Normalization.
Rotation converged in 7 iterations.

76
M. De Silva et al. Research Policy 47 (2018) 70–87

push’ (develop a technology to a marketable product or service), investigate the effect of knowledge-based practices on financial value
‘market pull’ (project being induced by the market need) and a ‘com- creation) F (11, 47) = 0.990, p = 0.469 > 0.05, indicating the suit-
bined approach’, when engaging in EU framework programmes using a ability of the model. The plot of residuals further confirms the sa-
1–5 Likert Scale. Another indication of the heterogeneity of their in- tisfaction of homogeneity assumption (p = 0.051 > 0.05).
novation profile is the technology readiness levels of projects, in which Measures were undertaken to check the problem of endogeneity
the higher levels of readiness indicate closer to the market orientation that could occur when independent variables might not be endogenous
of the organisation (Mankins, 1995). In order to control for this, re- (Shadish et al., 2002). Firstly, the sequential mixed method design
spondents were requested to state the highest Technology Readiness helped improve both internal and construct validity. Second, as pre-
Level (TRLs)1 they engaged in when undertaking EU FP7 programmes. sented in Section 4.1, a non-response bias test revealed that there is no
The data revealed the presence of three groups: with those reaching up significant difference between respondent RTOs and the full empirical
to TRL 4; those up to TRL 6; with the rest on TRL 7. None worked at TRL base of 122 RTOs. Third, satisfactorily meeting Cronbach's Alpha, CR,
8 or 9. A categorical variable with these three groups was then devel- AVE, Kaiser-Meyer-Olkin Measure of Sampling Adequacy and factor
oped to profile this. Third, the role of innovation intermediaries in loading criteria confirmed the reliability and validity of the chosen
collaborative projects could vary from acting as a networking agent and constructs, minimising measurement errors. Fourth, measures were
providing associated coordination services (Batterink et al., 2010) to taken to avoid reverse causality. Even though the impact of practices on
undertaking basic research and delivering associated training (Kodama, internal value was investigated, it could be argued that reverse caus-
2008) and conducting ‘closer’ to market research and providing related ality is possible, where organisations decide to introduce practices if
technology technological services and resources (Tran et al., 2011). As they are not generating enough internal value from their engagements.
innovation intermediaries may engage in a combination of these ac- Nevertheless, initial interviews revealed that since EU framework pro-
tivities to a different degree, respondents were requested to rate each grammes are long-term, practices were unlikely to be changed over the
types of contributions to indicate the extent to which they provided short term. Thus, the extent to which RTOs adopted a given list of
these in EU FP7 consortia. practices during the FP 7 programme was gauged with DV based on
value generation over the last three years of the programme (2011-3),
4.3. Estimation strategy increasing the likelihood that the practices adopted cause the values
rather than vice versa. This structure therefore provides a good time
The effect of knowledge-based practices adopted by innovation in- frame from which to evaluate the influence of specific practices. Fifth,
termediaries on value generation was then analysed. Pearson correla- measures were undertaken to address common method bias. Informa-
tion revealed a significant positive correlation between three dependent tion was cross-validated from secondary information derived on the
variable of non-financial value (for knowledge and marketing turnover of each organisation (Y) by multiplying it with the survey data
r = 0.510, n = 59, p = 0.000: knowledge and networking r = 0.604, on the percentage of turnover derived from EU projects (Z) in order to
n = 59, p = 0.000: marketing and networking r = 0.524, n = 59, reduce common method bias. Additional questions were asked that
p = 0.000). A Multivariate Analysis of Covariance (MANCOVA) was mixed up independent and dependent variable related questions so that
therefore performed that took into account interactions between de- the respondents were unable to know the antecedents and outcome
pendent variables in a context of multiple dependent and independent variable. The highest correlation among the constructs was also statis-
variables with both fixed variables and covariates (Meyers et al., 2006). tically checked (see Appendix A). The highest value was 0.640 (with a
There was no theoretical evidence to suggest using different sets of majority less than 0.3), which is low (Bagozzi et al., 1991). Common
explanatory variables for different dependent variables (i.e. non-fi- method variance, using Harman’s one-factor analysis (Podsakoff and
nancial value), which to justify the use of Seemingly Unrelated Re- Organ, 1986) showed that the 10 factors extracted with eigenvalues
gression (SUR) over MANCOVA model (Zellner, 1962). A separate above 1.0 (74.98% of the total variance) were explained by the first
univariate analysis was then run to investigate the influence of factor, which accounted for only 23.47%. Sixth, since our model com-
knowledge-based practices on the generation of financial value. bines factor analysis and regression analysis, Structural Equation
Modelling (SEM) was conducted for the univariate analysis to further
check the model fit (Miles, 2003). The results on the influence of
4.4. Robustness checks
practices on internal value were similar, except for the significant ne-
gative effect of two elements relating to innovation ecosystem influence
Variance Inflation Factors and correlation analysis revealed that
that were not significant in the univariate analysis yet reported a ne-
there is no evidence of multi-colinearity. Also in relation to MANCOVA,
gative influence. This similarity enabled us to use the model fit mea-
Box's Test of Equality of Covariance Matrices was found not to be sig-
sures of SEM to further support the findings of the univariate analysis.2
nificant (Value = 76.968, p = 0.271 > 0.1), indicating the ability to
Seventh, as many theoretically justified control variables as possible
pool variance-covariance matrices without any concerns and the
were included to avoid the omitting of a regressor (Rubin, 2008;
normal distribution of the sample. Similarly, Levene's Test of Equality
Shadish et al., 2002), which is recommended despite the cost of reduced
of Error Variances was also not significant (Knowledge F (11, 47)
efficiency (i.e., higher standard errors; Cameron and Trivedi, 2005).
= 1.297, p = 0.256 > 0.05, Market F (11, 47) = 1.869,
The results of SEM discriminant validity (i.e. measurement model of a
p = 0.069 > 0.05 and Network F (11, 47) = 1.969,
construct is free from redundant items) test found no items with high
p = 0.054 > 0.05), satisfying MANCOVA assumptions. The plot of
Modification Indices (MI), confirming that our control variables are not
residuals further confirms the satisfaction of homogeneity assumption
redundant (Jöreskog and Sörbom, 1984). This comprehensive approach
(Knowledge p = 0.200 > 0.05, Market p = 0.091 > 0.05 and
of robustness checks confirm that the model does not suffer from en-
Network p = 0.200 > 0.05). Levene's Test of Equality of Error
dogeneity and that the model is robust.
Variances was not significant for the Univariate analysis (i.e. to

1
TRL 1 – Concept proposed with scientific validation; TRL 2 – Application and validity
of concept validated or demonstrated: TRL 3 – Experimental proof of concept completed:
TRL 4 – Production validated in lab environment: TRL 5 – Basic capability demonstrated:
2
TRL 6 – Process optimised for production rate on production equipment: TRL 7 – GIF = 1.00(> 0.95); AGFI = 0.989 (> 0.95) (Tabachnick and Fidell, 2007);
Capability and rate confirmed: TRL 8 – Full production process qualified for full range of RMSEA = 0.000 (< 0.05), p = 0.892 (> 0.05) (Byrne, 1998); TLI = 1.548 ( > 0.95);
parts: TRL 9 – Full production process qualified for full range of parts and full metrics IFI = 1.001 ( > 0.95); CFI = 1.000 (> 0.95) (Tabachnick and Fidell, 2007); CMIN/
achieved. DF = 0.021 (p = 0.884) (< 2) (Carmines and McIver, 1981).

77
M. De Silva et al. Research Policy 47 (2018) 70–87

Table 4
Descriptive statistics I

Descriptive Statistics

Minimum Maximum Mean Std. Deviation

Dependent variables
Knowledge based value 1.75 4.75 3.83 0.58
Market based value 2.40 5.00 4.01 0.65
Network based value 2.67 5.00 3.85 0.59
Financial value 0.00 5600000000.00 795820359.32 1276972477.44

Characteristics of RTOs
Age of RTO 3.00 100.00 42.49 27.93
Number of employees in EU engagement 1.00 800.00 92.75 167.74
Turnover of RTO 400000.00 700000000.00 87896785.63 141815294.21

Nature of contribution by RTOs for EU activities

Applied research and technology service provision 1.00 4.75 3.36 0.69
Act as a networking agent 2.50 5.00 3.62 0.66
Engage in basic research and training 1.00 4.50 2.86 0.61

Table 5 of the adoption of these practices on value creation. A univariate ana-

Descriptive statistics II. lysis for financial value and MANCOVA for knowledge, market and
network value creation was conducted (Table 6).
% of respondents responding to each value on the Likert scale – the higher the score the
greater the involvement
The results revealed that different practices influence the generation
of different types of value. The only exception here was for practices
1 Never 2 Rarely 3 Sometimes 4 Often 5 Always adopted for knowledge capitalisation, which significantly positively
influenced all four types of value creation (Practice 1, Table 6), sup-
The extent to which RTOs adopts different innovation approaches in EU projects
Technology push 0.0% 5.1% 20.3% 71.2% 3.4%
porting H1. Since innovation intermediaries extensively interact with
Market pull 3.4% 11.9% 20.3% 44.1% 20.3% several external parties, capitalising on existing trustworthy relation-
Combined 1.7% 1.7% 10.2% 54.2% 32.2% ships, minimise the risk of failure and increase their ability to generate
The extent to which RTOs collaborate with different actors in EU projects new knowledge by combining existing knowledge bases. Innovation
Universities 0.0% 1.7% 10.2% 79.7% 8.5% intermediary employees engaged in EU projects seem to develop in-
Large Firms 0.0% 6.8% 23.7% 55.9% 13.6% valuable experience which is tacit and difficult to transfer and RTOs
Medium Firms 1.7% 5.1% 33.9% 44.1% 15.3%
made special efforts to retain them. Indeed, of the six knowledge-based
Small Firms 5.1% 22% 28.8% 44.1% 0.0%
Micro Firms 13.6% 40.7% 30.5% 13.6% 1.7% practices, only knowledge capitalisation was found to have a significant
RTO 0.0% 11.9% 20.3% 37.3% 30.5% positive influence on market value creation. This highlights the need for
capitalising proven and established knowledge bases inside innovation
intermediaries and/or strong networks to produce marketable products
5. Results and discussion and services.
In relation to the H2, it is evident that practices adopted for
The role of RTOs in the projects ranged from engaging in basic re- knowledge advancement have a significant positive influence only on
search and associated training, acting as a networking agent, engaging knowledge and financial value creation (Practice 2, Table 6). In-depth
in applied research to technology service provision, with a majority interviews validated this result suggesting that while knowledge ad-
mentioning a greater involvement in latter two (Table 4). A majority of vancement improves RTOs’ ability to secure funding, and therefore fi-
RTOs (86.4%) also reported adopting a combination of both ‘tech- nancial value, there is a lag period for newly recruited staff to generate
nology push’ (i.e. develop a technology to a marketable product or market or network value. This means that only when new knowledge is
service) and ‘market pull’ (project being induced by the market need) appropriately integrated into the existing knowledge base is market and
approaches (Table 5), although ‘market pull’ activities came a strong network value generated, which is captured in the H1 as a form of
second above that of ‘technology push’ factors (see Polzin et al., 2016). capitalising on existing knowledge.
Of the RTOs, a third (33.9%) had engaged in projects up to TRL4, 28.8 Interestingly, practices associated with the knowledge-spanning role
percent up to TRL 6 and 37.3 percent up to TRL 7, representing dif- had no significant positive influence on any type of value, except for the
ferent levels of market readiness levels of project outcomes. The in- significant negative influence on market value (Practice 3, Table 6),
novation intermediaries were more likely to collaborate with uni- rejecting H3. In-depth interviews revealed that RTOs mainly ‘access’
versities, large and medium-sized firms and other RTOs, rather than unrelated knowledge that is non-adjacent to their knowledge base,
small or micro firms (Table 5). which they do not intend to acquire and internalise, but use only to
Moving on to focus on internal value generation by intermediaries achieve project outcomes. Hence, these interactions may not generate
from client interaction, the descriptive statistics also indicated that in additional value other than achieving direct project outcomes. How-
addition to financial value generation, innovation intermediaries also ever, it may also be a lag effect in that there are considerable ‘start-up’
generated knowledge (M– 3.83; SD– 0.58), market (M– 4.01; SD– 0.65), costs to create sunk capital in acquiring new knowledge which generate
and network (M– 3.85; SD– 0.59)3 based value (Table 4). The in- negative impacts over the short term, but may lead to longer-term value
dependent variables constructed using factor scoring indicates that the generation. RTOs have also highlighted the difficulties they face when
respondents to a varying degree adopted the six types of knowledge- interacting with collaborators from different disciplines, whose
based practices, further supporting the need to investigate the influence knowledge is not familiar to RTOs, and thus, unless these players could
be of future value, interaction would be limited to achieving specific
project outcomes, which may explain the results of the regression
3
The value range from 0 to 5 and the higher the mean, the greater the non-financial analysis. On this basis, therefore, if RTOs perceive that these players
value generated.

78
M. De Silva et al. Research Policy 47 (2018) 70–87

Table 6
MANCOVA and univariate analysis.

Independent variables Dependent Variable

MANCOVA Univariate analysis

Knowledge Value Market Value Network Value Financial Value

2
R 0.904 0.886 0.858 0.930
Adjusted R2 0.746 0.700 0.627 0.816
Correct model
Intercept 6.541** (.953) 4.532** (1.167) 6.664** (1.173) 2831804952 (1762828643.2)
Practices adopted
1. Knowledge capitalisation 0.202** (.078) 0.278** (.095) 0.328** (.096) 327355109.1** (146998770.4)
2. Knowledge advancement 0.214** (.078) −0.079 (.095) 0.091 (.096) 355049550.3** (153226466.4)
3. Knowledge spanning −0.016 (.071) −0.236** (.087) 0.008 (.087) 138094579.6 (132681604.5)
4. Knowledge worker empowerment −0.227** (.076) −0.229** (.093) −0.168* (.093) −110381422.8 (139300095.5)
5. Innovation ecosystem knowledge access 0.186** (0.067) 0.041 (0.082) 0.357** (0.083) 283902599.8** (128149691.9)
6. Innovation ecosystem knowledge shaping
Direct EU involvements and memberships 0.200** (0.079) −0.345** (0.097) −0.026 (0.098) −271298400.1 (159128850.1)
Written communication 0.255** (0.111) −0.457** (0.136) 0.278* (0.137) −277295774.0 (206814528.8)
Expert advisory groups 0.198** (0.082) −0.045 (.100) 0.196* (0.101) −357856699.4** (153641960.4)
Through external bodies 0.319** (0.077) −0.168* (0.094) 0.208** (0.095) −180017052.8 (145053505.2)
Media events −0.194** (0.074) −0.176* (0.090) .001 (0.091) −96286262.4 (134138673.4)

Results after controlling for the characteristics of RTO, and other activities defining their role in EU projects such as the nature of contribution by RTOs, innovation approaches adopted by
RTOs, the extent of engagement with different types of actors, sector of operation of RTOs and the extent of commercialisation engagement by RTOs (see Appendix B for full model
including control variables).
** p < 0.05.
* p < 0.1.

from other disciplines would be of future value, they make an effort to generation. Influencing innovation system ‘architecture’ (Tjong et al.,
strengthen the relationship, the use of the knowledge of whom then 2015), including EU project calls, enabled RTOs to improve the chances
becomes capitalising on knowledge vested in close collaborators, cap- of developing knowledge of value during projects. Such influences have
tured in the H1. also been made in collaboration with other types of innovation inter-
Practices adopted to empower employees, such as giving freedom mediaries and like-minded organisations, which in turn become project
for staff to decide what EU projects to work on, with whom and what collaborators leading to strengthening the network of future value.
practices should be involved in such engagements, had significant ne- However, any source used to influence innovation ecosystem knowl-
gative influence on all the non-financial value creation and non-sig- edge does not seem to positively influence financial or market based
nificant, but still negative, influence on financial value creation value creation.
(Practice 4, Table 6), supporting H4. Interviews revealed that since
innovation intermediaries are by their nature outward-looking, having 6. Conclusions
some control over employee engagement is essential. Therefore, there
appears to be a dynamic balance between employee empowerment and 6.1. Key findings and implications
centralised control by RTOs that enable them to generate value. This is
further justified as RTOs engage in EU projects not only as a source of There has been a growing body of knowledge and research sur-
income but also as a source of long-term value creation in knowledge, rounding the role and impact that innovation intermediaries have on
network and market capability development. their client organisations, but little is known about the internal value
Practices adopted to access innovation ecosystem knowledge (H5) creation of innovation intermediaries from their interaction with their
significantly positively influence all the types of value creation apart clients. This is a non-trivial issue because if innovation intermediaries
from market-based value (Practice 5, Table 6). This was mainly asso- do have an important impact on innovation system and network
ciated with the role of RTOs as network builders, in which knowledge of ‘health’ and on the specific performance and growth of client en-
the ecosystem actors enable RTOs to bring together key players for EU terprises and organisations, it is important to understand what helps to
projects. This increases the chances of being successful both in terms of sustain and develop their own long-term growth and development. As
securing funding as well as delivering output. Having the opportunity Knockaert and Spithoven (2014, 1400) have suggested in their analysis
to work with these players enable RTOs to develop new knowledge and of innovation intermediaries and the role of absorptive capacity one
network of future value. Not having a positive influence on market needs to understand the capacities of both client firms and technology
value creation might be because of other compatibilities or precondi- intermediaries to understand the whole value generation process.
tions needed to be fulfilled (as discussed in relation to H1) to generate Our results show that it is the knowledge attributes and routines of
marketable products and services. intermediaries’ staff and their attitudes (Gavetti, 2005; see also
Innovation-ecosystem knowledge shaping (H6), interestingly, had a Tortoriello et al., 2012, 2014) as well as the support given to them in
significant positive influence on knowledge and network value creation, terms of knowledge capitalisation (but not practices allowing staff more
but not on market or financial value creation (Practice 6, Table 6). All empowerment) that have the most significant impact on value creation
the sources used to influence innovation ecosystem knowledge, except within innovation intermediaries. This finding is perhaps not surprising
for media events, positively influenced knowledge value creation. given that ultimately knowledge is held at an individual level. How an
Written communication, expert advisory groups and influence made innovation intermediary harnesses local personal knowledge and per-
through external bodies have a positive impact on network value sonal knowledge networks (Huber, 2012) is therefore essential for its

79
M. De Silva et al. Research Policy 47 (2018) 70–87

long term growth and development. This is also true in terms of the effects, such as the RTOs knowledge-spanning role are not observed in
impact of intermediaries on their client enterprises and organisations the data because the time period being considered is too short. In ad-
where the underlying absorptive capacity of both sets of organisations dition, the different value types and their impact on RTOs are also af-
also inherently lies. This is why perhaps knowledge advancement is fected by time and periodicity. Thus, as has been seen, shorter term
more limited in its impact, centred on financial and knowledge value financial value ‘first order’ effects are more prominent but taking a
creation, as the absorption and internalisation of new knowledge and longer term view ‘second order’ effects associated with knowledge,
knowledge practices take time. Knowledge spanning activities appear to markets and networks could become much more significant. The second
be of little (or indeed negative in the case of market value) importance set of issues is how the generation of internal value and its variable
in terms of value generation to intermediaries. This indicates that it is ‘geometry’ influences the future behaviour and routines of RTOs as a
the spanning of market knowledge with their existing technical com- subset of innovation intermediaries. There will be positive and negative
petence that is important for the intermediaries themselves. Technical feedback loops on RTO behaviours and practices which are not ex-
knowledge is therefore a condition for their other activities but not an plored here, again in part because of temporal issues. Both these lim-
immediate driver for innovation intermediaries. itations indicate the need for research in this topic. Qualitative data
By contrast, the benefit that innovation intermediaries generate as from the RTO survey is available and this may point to further rounds of
network integrators and shapers in terms of the architecture of the data collection, especially on a longitudinal basis. More fundamentally,
knowledge and innovation system is one of the most striking results. the research points to a lack of the interaction between value genera-
This is reflected in other studies of intermediaries that are emerging tion for both the clients of intermediaries and intermediaries themselves
with, for example, the study Tjong et al. (2015) on the role of in- and how this shapes the trajectory of the market or the sector and the
novation intermediaries on the development of bicycle technology ‘loosely coupled’ feedback loops between the two groups. This is im-
within the Netherlands and how they helped change the institutional portant in policy terms if we are to understand how innovation inter-
architecture of the national sectoral system. The role of innovation mediaries may be used to leverage benefits for the wider national or
intermediaries as network builders, bridgers (Caloffi et al., 2015) and regional innovation systems as well as just for their clients.
innovation system architects (Klerkx and Leeuwis, 2008b), with no
apparent impact on their financial bottom line, but one which enhances
their long term capabilities and attraction, is their most significant Acknowledgement
policy impact. This, together with their direct impact on enhancing the
performance of their client enterprises and organisations, is where in- Authors would like to acknowledge the financial assistance pro-
novation intermediaries continue to build their presence within the vided by the InnovateUK for this study. We would like to thank the
innovation system of nations, regions and sectors. editor, Prof Ben Martin, and three anonymous reviewers for immensely
valuable feedback and advice in the preparation of this article. Our
6.2. Limitations and further research gratitude also goes to the Big Innovation Centre for supporting data
collection.
The study has had to confront a number of issues, which must be Also, we express our sincere thank to Dr. Zaheer Khan and Dr.
acknowledged when reviewing the research, of which two are most George Chryssochoidis for providing us with extremely valuable gui-
significant here. The first set revolve around lag effects and this can be dance and insights to improve the paper. We are also thankful for the
seen on a number of levels. As noted above (Section 5.1) a number of support of Santander Bank and Kellogg College, University of Oxford.

80
 Appendix A. Correlation matrices between each IV and DV.

Annual tor_Number_C Staff_Size_EU Age TRL_Combinations_C. Tech_Push Market_Pull Combined_Inno Partners_Universities Partners_Large
M. De Silva et al.

turnover firms

Annual turnover 1
Sector_Number_C 0.301 1
Staff_Size_EU 0.640 0.289 1
Age 0.432 0.285 0.327 1
TRL_Combinations_C 0.001 0.260 0.240 0.420 1
Tech_Push 0.147 0.175 0.169 0.236 0.317 1
Market_Pull −0.022 0.223 0.049 0.298 0.402 0.286 1
Combined_Inno 0.164 0.310 0.229 0.250 0.400 0.289 0.553 1
Partners_Universities −0.134 0.059 0.047 −0.069 −0.076 0.066 −0.033 −0.111 1
Partners_Large firms 0.076 0.179 0.051 0.386 0.432 0.555 0.369 0.416 0.013 1
Partners_Medium firms 0.016 0.150 0.138 0.425 0.533 0.280 0.291 0.368 0.118 0.678
Partners_Small firms −0.111 −0.019 −0.162 0.339 0.365 0.027 0.255 0.234 −0.024 0.304
Partners_Micro firms −0.063 −0.121 −0.020 0.186 0.234 0.085 −0.003 −0.089 0.017 0.114
Partners_RTO 0.166 0.288 0.097 0.492 0.415 0.195 0.288 0.198 0.124 0.520
Contri_ Ap. Research 0.185 −0.093 0.154 0.398 0.187 0.113 0.204 0.005 0.303 0.213
Contribution_networking 0.050 0.101 0.107 0.262 0.436 0.052 0.398 0.286 −0.144 0.369
Contri_Basic research −0.173 −0.125 −0.004 −0.194 −0.083 −0.182 −0.152 −0.231 0.114 −0.103
EU_involvements 0.289 0.214 0.177 0.382 0.297 0.102 0.071 0.075 −0.141 0.366
Written com 0.058 −0.067 0.090 0.335 0.263 0.124 0.392 0.269 −0.247 0.363

81
Expert_advisory −0.197 0.160 −0.112 −0.058 0.251 0.325 0.161 0.350 −0.137 0.502
External bodies 0.233 0.036 0.259 0.068 0.003 −0.128 −0.211 0.083 0.024 0.058
Influence Media−events 0.157 0.045 0.265 −0.035 0.043 0.047 −0.031 −0.030 0.044 −0.012
Know_Capitalisation 0.090 0.198 0.166 0.328 0.390 0.207 0.323 0.473 0.044 0.485
Know_Advancement 0.164 0.118 0.035 0.132 0.123 0.199 0.302 0.360 0.069 0.164
Tech_Know_Access −0.119 0.008 −0.014 −0.222 −0.045 −0.394 0.054 −0.025 0.115 −0.387
Eco_Know_Access 0.180 0.132 0.100 0.056 0.066 0.311 0.010 0.084 −0.096 −0.055
Know_Comm_Dev −0.028 0.085 0.068 −0.150 0.015 −0.096 −0.144 0.130 −0.035 −0.127
Know_Value 0.031 0.138 0.205 0.331 0.494 0.096 0.393 0.510 −0.045 0.244
Market_Value 0.030 0.178 0.034 0.341 0.515 0.316 0.464 0.488 −0.062 0.543
Network_Value 0.007 0.234 0.094 0.174 0.458 0.319 0.301 0.447 −0.045 0.364
Financila_Value N/A 0.176 0.566 0.402 0.051 0.156 0.099 0.263 −0.157 0.151

Partners_Medium Partners_Small Partners_Micro Partners_RTO Contri_ Ap. Contribution_ Contri_Basic EU_involvements Written Expert_ External
firms firms firms Research networking research com advisory bodies

Annual turnover
Sector_Number_C
Staff_Size_EU
Age
TRL_Combinations_C
Tech_Push
Market_Pull
Combined_Inno
Research Policy 47 (2018) 70–87

Partners_Universities
 Partners_Large firms
Partners_Medium firms 1
Partners_Small firms 0.545 1
Partners_Micro firms 0.353 0.633 1
M. De Silva et al.

Partners_RTO 0.550 0.392 0.236 1

Contri_ Ap. Research 0.341 0.303 0.289 0.266 1
Contribution_networking 0.550 0.394 0.337 0.494 0.000 1
Contri_Basic research −0.010 −0.094 0.108 −0.132 0.000 0.000 1
EU_involvements 0.315 0.068 −0.022 0.252 −0.090 0.362 −0.028 1
Written com 0.429 0.411 0.470 0.441 0.246 0.527 −0.218 0.000 1
Expert_advisory 0.177 0.132 −0.008 0.211 −0.055 0.088 0.104 0.000 0.000 1
External bodies 0.107 −0.106 0.090 0.166 −0.013 0.126 0.158 0.000 0.000 0.000 1
Influence Media−events 0.043 −0.033 0.069 0.103 0.344 0.181 0.190 0.000 0.000 0.000 0.000
Know_Capitalisation 0.531 0.378 0.328 0.574 0.136 0.571 0.065 0.284 0.359 0.328 0.361
Know_Advancement 0.202 0.148 −0.086 0.002 0.365 0.075 −0.148 −0.103 0.037 0.160 −0.085
Tech_Know_Access −0.306 0.004 −0.084 −0.091 0.071 0.030 0.132 −0.337 −0.224 0.053 −0.029
Eco_Know_Access −0.076 −0.158 0.090 0.109 0.088 −0.092 −0.057 −0.119 0.033 0.084 −0.007
Know_Comm_Dev −0.132 −0.264 −0.280 −0.234 −0.340 0.022 0.234 0.214 −0.111 0.018 0.023
Know_Value 0.338 0.183 0.040 0.274 0.079 0.381 −0.164 0.103 0.288 0.275 0.207
Market_Value 0.579 0.300 0.083 0.455 0.311 0.477 −0.119 0.043 0.252 0.323 0.031
Network_Value 0.296 0.232 0.059 0.360 −0.017 0.401 −0.213 0.073 0.244 0.361 0.146
Financila_Value 0.055 −0.061 −0.029 0.162 0.196 0.194 −0.126 0.291 0.094 −0.167 0.171

Influence Know_ Know_ Tech_Know_ Eco_Know_ Know_ Know_Value Know_Value Know_Value Know_Value

82
Media-events Capitalisation Advancement Access Access Comm_Dev

Annual turnover
Sector_Number_C
Staff_Size_EU
Age
TRL_Combinations_C
Tech_Push
Market_Pull
Combined_Inno
Partners_Universities
Partners_Large firms
Partners_Medium firms
Partners_Small firms
Partners_Micro firms
Partners_RTO
Contri_ Ap. Research
Contribution_networking
Contri_Basic research
EU_involvements
Written com
Expert_advisory
External bodies
Influence Media−events 1
Know_Capitalisation 0.065 1
Research Policy 47 (2018) 70–87
 Know_Advancement 0.112 0.000 1
Tech_Know_Access 0.128 0.000 0.000 1
Eco_Know_Access 0.290 0.000 0.000 0.000 1
Know_Comm_Dev −0.163 0.000 0.000 0.000 0.000 1
M. De Silva et al.

Know_Value −0.048 0.528 0.203 0.089 0.160 0.083 1

Market_Value 0.143 0.570 0.270 −0.038 0.079 −0.259 N/A 1
Network_Value 0.073 0.538 0.012 0.078 0.383 0.024 N/A N/A 1
Financila_Value 0.186 0.229 0.178 −0.022 0.180 −0.028 N/A N/A N/A 1

Appendix B. MANCOVA and univariate analysis: Including control variables

Independent variables Dependent Variable

MANCOVA Univariate analysis

Knowledge Value Market Value Network Value Financial Value

R2 0.904 0.886 0.858 0.930

Adjusted R2 0.746 0.700 0.627 0.816
Correct model
Intercept 6.541** (.953) 4.532** (1.167) 6.664** (1.173) 2831804952 (1762828643.2)
Practices adopted Independent variables
1. Knowledge capitalisation .202** (.078) .278** (.095) .328** (.096) 327355109.1** (146998770.4)
2. Knowledge advancement .214** (.078) .091 (.096) 355049550.3** (153226466.4)

83
−0.079 (.095)
3. Knowledge spanning −0.016 (.071) −0.236** (.087) .008 (.087) 138094579.6 (132681604.5)
4. Knowledge worker empowerment −0.227** (.076) −0.229** (.093) −0.168* (.093) −110381422.8 (139300095.5)
5. Innovation ecosystem knowledge access .186** (.067) .041 (.082) .357** (.083) 283902599.8** (128149691.9)
6. Innovation ecosystem knowledge shaping
Direct EU involvements and memberships .200** (.079) −0.345** (.097) −0.026 (.098) −271298400.1 (159128850.1)
Written communication .255** (.111) −0.457** (.136) .278* (.137) −277295774.0 (206814528.8)
Expert advisory groups .198** (.082) −0.045 (.100) .196* (.101) −357856699.4** (153641960.4)
Through external bodies .319** (.077) −0.168* (.094) .208** (.095) −180017052.8 (145053505.2)
Media events −0.194** (.074) −0.176* (.090) .001 (.091) −96286262.4 (134138673.4)
Control variables
Characteristics of RTO
Number of employees in EU engagement .001* (.000) .000 (.001) .000 (.001) 941915.6 (923396.5)
Age of RTO .005* (.003) .000 (.003) −0.005 (.003) 32960554.4** (15601447.6)
Turnover of RTO −1.743E-9** (6.831E-10) −8.110E-10 (8.368E-10) −5.861E-11 (8.415E-10) 7.199** (1.283)
Innovation Approach
Technology push −0.201* (.114) −0.101 (.140) .104 (.141) −381554969.5* (221358273.9)
Market Pull .073 (.075) −0.042 (.092) −0.004 (.092) −135241334.6 (142252788.3)
Combined −0.085 (.135) .327* (.165) −0.349** (.166) −13720343.1 (255866857.3)
Nature of contribution by RTOs for EU activities
Applied Research and Technology Service Provision −0.049 (.089) .340** (.108) −0.063 (.109) −75016565.7 (172519490.4)
Act as a Networking Agent .080 (.101) .417** (.123) .110 (.124) 311947824.3 (189418055.8)
Engage in Basic Research and Training −0.040 (.060) −0.004 (.074) −0.037 (.074) 124438046.9 (112843119.1)
The extent to which RTOs collaborate with different actors in EU projects
Universities .418** (.142) −0.423 (.174) .306* (.175) −328883621.8 (267090234.8)
Research Policy 47 (2018) 70–87
 Large Firms −0.291* (.155) .139 (.190) .037 (.191) 809398271.4** (290564911.7)
Medium Firms −0.148 (.131) .037 (.161) −0.283* (.162) −455080845.7* (247161315.4)
Small Firms .107 (.093) −0.158 (.114) .370** (.115) −100109312.0 (174080212.0)
Micro Firms −0.278** (.095) −0.096 (.117) −0.446** (.117) 5954792.4 (176315652.6)
M. De Silva et al.

RTO −0.317** (.089) .092 (.110) −0.370** (.110) −229561409.7 (173509405.1)

Sector of operation of RTOs
Information and Communication Technologies (ICT) −0.490* (.256) .124 (.314) −0.097 (.316) −200079845.7 (484160590.7)
Bioscience (Agri, Chemical and Env) −0.013 (.266) .212 (.326) .179 (.327) −435737278.0 (499194810.5)
Engineering and Energy −0.410** (.190) .157 (.233) −0.114 (.234) 23648252.9 (350848269.9)
Multi-sector 0 0 0 0
TRL up to 4 −1.093* (.554) .148 (.678) −1.271* (.682) −1253040670.5 (1060295024.9)
TRL up to 6 −0.464 (.277) .540 (.339) −0.512 (.341) −1221896009.7 (530608710.1)
TRL up to 7 0 0 0 0
ICT* TRL up to 4 1.355** (.546) −0.410 (.669) .597 m(.673) 1779898832.6 (1028946333.2)
ICT * TRL up to 6 .822* (.428) .365 (.524) −0.291 (.527) 1131956188.3** (800514957.4)
ICT * TRL up to 7 0 0 0 0
Bio* TRL up to 4 −0.513 (.609) −0.668 (.746) −0.097 (.750) 1432962951.4* (1154226347.7)
Bio* TRL up to 6 4.931E-5 (.392) −0.791 (.480) −1.284** (.483) 798795159.6 (730543338.8)
Bio* TRL up to 7 0 0 0 0
Eng * TRL up to 4 .779 (.561) −0.340 (.687) .885 (.691) 892814888.9 (1107527395.7)
Eng * TRL up to 6 1.047** (.293) −0.523 (.359) .380 (.361) 847229529.9 (558033311.6)
Eng * TRL up to 7 0 0 0 0
Multi * TRL up to 4 0 0 0 0
Multi * TRL up to 6 0 0 0 0
Multi* TRL up to 7 0 0 0 0

84
**p < 0.05.
*p < 0.1.
Research Policy 47 (2018) 70–87
M. De Silva et al. Research Policy 47 (2018) 70–87

economic research: reorienting economics into social science. Camb. J. Econ. 31,
References 77–99.
Dyer, J., Nobeoka, K., 2000. Creating and managing a high performance knowledge
sharing network: the toyota case. Strateg. Manage. J. 21, 345–367.
Agogué, M., Yström, A., Le Masson, P., 2013a. Rethinking the role of intermediaries as an EARTO, 2017. How Joint Undertakings Boost RTOs-industry Collaboration and Leverage
architect of collective exploration and creation of knowledge in open innovation. Int. Private RD & I Investments in Europe. European Association of Research and
J. Innovation Manage. 17 (02), 1–24. Technology Organisations (EARTO), Brussels (Working Paper, 10 March 2017).
Agogué, M., Berthet, E., Fredberg, T., Le Masson, P., Segrestin, B., Stoetzel, M., Wiener, Earl, M., 2001. Knowledge management strategies: toward a taxonomy. J. Manage. Inf.
M., Yström, A., 2013b. A contingency approach of open innovation intermediaries- Syst. 18, 215–233.
the management principles of the intermediary of the unknown. In: 13th Annual Edler, J., Yeow, J., 2016. Connecting demand and supply: The role of intermediation in
Conference of the European Academy of Management. Istanbul, EURAM 26–29 June public procurement of innovation. Res. Policy 45 (2), 414–426.
2013. pp. 26–29. Edmondson, A.C., Mcmanus, S.E., 2007. Methodological fit in management field research.
Alavi, M., Kayworth, T.R., Leidner, D.E., 2005. An empirical examination of the influence Acad. Manage. Rev. 32, 1155–1179.
of organizational culture on knowledge management practices. J. Manage. Inf. Syst. Escribano, A., Fosfuri, A., Tribó, J.A., 2009. Managing external knowledge flows: the
22, 191–224. moderating role of absorptive capacity. Res. Policy 38, 96–105.
Albors, G.J., Rincón Díaz, C.A., Igartua, L., 2014. Research technology organisations as Etzkowitz, H., 2008. The triple helix: university-industry-government. Innovation in
leaders of R & D collaboration with SMEs: role, barriers and facilitators. Technol. Action. Routledge, London.
Anal. Strategic Manage. 26 (1), 37–53. Fontana, R., Geuna, A., Matt, M., 2006. Factors affecting university-industry R & D pro-
Aldrich, H.E., von Glinow, M.A., 1992. Business start-ups: the HRM imperative. In: Birley, jects: the importance of searching, screening and signalling. Res. Policy 35, 309–323.
S., MacMillan, I.C. (Eds.), International Perspectives on Entrepreneurial Research, pp. Ford, R.C., Fottler, M.D., 1995. Empowerment: a matter of degree. Acad. Manage.
233–253 North-Holland, New York. Perspect. 9, 21–29.
Andersen, B., Blanc, E.L., 2013. Catapult to Success: Be Ambitious, Bold and Enterprising Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unobservable
Learning from European Technology and Innovation Centres. Big Innovation Centre, variables and measurement error. J. Marketing Res. 18 (1), 39–50.
London. Garud, R., Nayyar, P.R., 1994. Transformative capacity: continual structuring by inter-
Argyris, C., 1998. Empowerment: the emperor’s new clothes. Harv. Bus. Rev. 76, 98–105. temporal technology transfer. Strateg. Manage. J. 15, 365–385.
Arnold, E., Clark, J., Javorka, Z., 2010. Impacts of european RTOs: a study of social and Gassmann, O., Enkel, E., Chesbrough, H., 2010. The future of open innovation. R & D
economic impacts of research and technology organisations. Report to European Manage. 40, 213–221.
Association of Research and Technology Organisations (EARTO). Technopolis Ltd., Gavetti, G., 2005. Cognition and hierarchy: rethinking the micro foundations of cap-
Brighton. abilities’ development. Organ. Sci. 16, 599–617.
Autio, E., Dahlander, L., Frederiksen, L., 2013. Information exposure, opportunity eva- Gold, A.H., Malhotra, A., Segars, A.H., 2001. Knowledge management: an organizational
luation, and entrepreneurial action: an investigation of an online user community. capabilities perspective. J. Manage. Inf. Syst. 18, 185–214.
Acad. Manage. J. 56, 1348–1371. Granovetter, M.S., 1985. Economic action and social structure: the problem of embedd-
Bagozzi, R.P., Yi, Y., Phillips, L.W., 1991. Assessing construct validity in organizational edness. Am. J. Sociol. 91, 481–510.
research. Adm. Sci. Q. 36 (3), 421–458. Grant, R.M., Baden-Fuller, C., 1995. A knowledge-based theory of inter-firm collabora-
Ballon, P., 2007. Business modelling revisited: the configuration of control and value. tion. Acad. Manage. Best Papers Proc. 8, 17–21.
Info: J. Policy Regul. Strategy Telecommun. Inf. Media 9 (5), 6–19. Grant, R.M., Baden-Fuller, C., 2004. A knowledge accessing theory of strategic alliances.
Batterink, M.H., Wubben, E.F.M., Klerkx, L., Omta, S.W.F., 2010. Orchestrating innova- J. Manage. Stud. 41, 61–84.
tion networks: the case of innovation brokers in the agri-food sector. Grant, R.M., 1996. Toward a knowledge-based theory of the firm. Strateg. Manage. J. 17,
Entrepreneurship Regional Dev. 22, 47–76. 109–122.
Bisbe, J., Batista-Foguet, J.M., Chenhall, R., 2007. Defining management accounting Gulati, R., Nohria, N., Akbar, Z., 2000a. Guest editors’ introduction to the special issue:
constructs: a methodological note on the risks of conceptual misspecification. strategic networks. Strateg. Manage. J. 21, 203–215.
Accounting. Organ. Soc. 32, 789–820. Gulati, R., Nohria, N., Akbar, Z., 2000b. Strategic networks. Strateg. Manage. J. 21,
Boari, C., Riboldazzi, F., 2014. How knowledge brokers emerge and evolve: the role of 199–201.
actors’ behaviour. Res. Policy 43, 683–695. Gulati, R., 1999. Network location and learning: the influence of network resources and
Boon, W.P., Moors, E.H., Kuhlmann, S., Smits, R.E., 2011. Demand articulation in firm capabilities on alliance formation. Strateg. Manage. J. 20, 397–420.
emerging technologies: intermediary user organisations as co-producers? Res. Policy Gupta, A.K., Tesluk, P.E., Taylor, M.S., 2007. Innovation at and across multiple levels of
40, 242–252. analysis. Organ. Sci. 18, 885–897.
Bouwen, P., 2002. Corporate lobbying in the European Union: the logic of access. J. Eur. Hales, M., 2001. Birds were dinosaurs once: the diversity and evolution of research and
Public Policy 9, 365–390. technology organizations. RISE Final Report. CENTRIM, University of Brighton,
Bowonder, B., Miyake, T., 2000. Technology management: a knowledge ecology per- Brighton.
spective. Int. J. Technol. Manage. 19, 7–8. Hardy, C., Phillips, N., 1998. Strategies of engagement: lessons from the critical ex-
Bruneel, J., D’Este, P., Salter, A., 2010. Investigating the factors that diminish the barriers amination of collaboration and conflict in an interorganizational domain. Organ. Sci.
to university-industry collaboration. Res. Policy 39, 858–868. 9 (2), 217–230.
Bruneel, J., D’Este, P., Salter, A., 2015. The impact of financial slack on explorative and Hardy, C., Phillips, N., Lawrence, T.B., 2003. Resources, knowledge and influence: the
exploitative knowledge sourcing from universities: evidence from the UK. Ind. organizational effects of interorganizational collaboration. J. Manage. Stud. 40,
Corporate Change 25, 689–706. 321–347.
Byrne, B.M., 1998. Structural Equation Modeling with LISREL, PRELIS and SIMPLIS: Basic Hargadon, A.B., Sutton, R.I., 1997. Technology brokering and innovation in a product
Concepts, Applications and Programming Mahwah. Lawrence Erlbaum Associates, development firm. Adm. Sci. Q. 42, 716–749.
New Jersey. Hargadon, A.B., 2002. Brokering knowledge: linking learning and innovation research.
Caloffi, A., Rossi, F., Russo, M., 2015. The emergence of intermediary organizations: a Organ. Behav. 24, 41–85.
network-based approach to the design of innovation policies. In: Geyer, R., Cairney, Hauser, H., 2010. The Current and Future Role of Technology and Innovation Centres in
P. (Eds.), Handbook on Complexity and Public Policy. Edward Elgar, Cheltenham, pp. the UK. Department for Business, Innovation and Skills, HMSO, London.
314–330. Hobday, M., 2000. The project-based organisation: an ideal form for managing complex
Cameron, A.C., Trivedi, P.K., 2005. Microeconometrics: Methods and Applications. products and systems? Res. Policy 29, 871–893.
Cambridge University Press, New York. Holmstrom, B., 1989. Agency costs and innovation. J. Econ. Behav. Organ. 12 (3),
Carmines, E.G., McIver, J.P., 1981. Analyzing models with unobserved variables. In: 305–327.
Bohrnstedt, G.W., Borgatta, E.F. (Eds.), Social Measurement: Current Issues. Sage Howells, J., Roberts, J., 2000. From innovation systems to knowledge systems.
Publications, Beverly Hills. Prometheus 18, 17–31.
Cepeda, G., Vera, D., 2007. Dynamic capabilities and operational capabilities: aknow- Howells, J., Ramlogan, R., Cheng, S.L., 2012. Innovation and university collaboration:
ledge management perspective. J. Bus. Res. 60, 426–437. paradox and complexity within the knowledge economy. Camb. J. Econ. 36,
Chesbrough, H.W., 2003. The era of open innovation. MIT Sloan Manage. Rev. 44 (3), 703–721.
35–41. Howells, J., 2006. Intermediation and the role of intermediaries in innovation. Res. Policy
Chesbrough, H.W., 2006. Open Innovation: The New Imperative for Creating and 35, 715–728.
Profiting from Technology. Harvard Business Press, Cambridge, MA. Howells, J., 2012. The geography of knowledge: never so close but never so far apart. J.
Chin, W.W., 1988. The partial least squares approach to structural equation modeling. In: Econ. Geogr. 12 (5), 1003–1020.
Marcoulides, G.A. (Ed.), Modern Methods for Business Research. Lawrence Erlbaum, Huber, F., 2012. On the role and interrelationship of spatial, social and cognitive proxi-
Mahwah, NJ. mity: personal knowledge relationships of R & D workers in the Cambridge in-
Collis, B., Moonen, J., 2008. Web 2.0 tools and processes in higher education: quality formation technology cluster. Reg. Stud. 46, 1169–1182.
perspectives. Educ. Media Int. 45, 93–106. Huizingh, E.K., 2011. Open innovation: state of the art and future perspectives.
Curley, M., Salmelin, B., 2013. Open Innovation 2.0. A New Paradigm. White Paper. Technovation 31, 2–9.
Dahlander, L., Gann, D.M., 2010. How open is innovation? Res. Policy 39, 699–709. Jöreskog, K.G., Sörbom, D., 1984. LISREL-VI User’s Guide, 3rd ed. Scientific Software,
Davenport, T.H., De Long, D.W., Beers, M.C., 1998. Successful knowledge management Mooresville.
projects. Sloan Manage. Rev. 39, 43–57. Jensen, M., 2001. The Foundations of Organizational Strategy. Harvard University Press,
Davenport, T.H., Thomas, R.J., Cantrell, S., 2002. The mysterious art and science of Boston.
knowledge-worker performance. MIT Sloan Manage. Rev. 44 (1), 23–30. Johnson, M.W., Christensen, C.M., Kagermann, H., 2008. Reinventing your business
Downward, P., Mearman, A., 2007. Retroduction as mixed-methods triangulation in model. Harv. Bus. Rev. 86 (12), 50–59.

85
M. De Silva et al. Research Policy 47 (2018) 70–87

Kale, P., Singh, H., Perlmutter, H., 2000. Learning and protection of proprietary assets in Oxford.
strategic alliances: building relational capital. Strateg. Manage. J. 21, 217–237. Mount, M., Milewski, S., Fernandes, K., 2015. Exploring the knowledge complexities of
Kauffeld-Monz, M., Fritsch, M., 2013. Who are the knowledge brokers in regional systems innovation intermediaries: the case of nanotechnology in the UK. Int. J. Technol.
of innovation? A multi-actor network analysis. Reg. Stud. 47, 669–685. Manage. 69, 20–37.
Kim, J.-Y., Miner, A.S., 2007. Vicarious learning from the failures and near-failures of Mowery, D.C., Oxley, J.E., Silverman, B.S., 1996. Strategic alliances and interfirm
others: evidence from the U.S: commercial banking industry. Acad. Manage. J. 50, knowledge transfer. Strateg. Manage. J. 17, 77–91.
687–714. Muller, E., Zenker, A., 2001. Business services as actors of knowledge transformation: the
Kirkels, Y., Duysters, G., 2010. Brokerage in SME networks. Res. Policy 39 (3), 375–385. role of KIBS in regional and national innovation systems. Res. Policy 30, 1501–1516.
Klerkx, L., Leeuwis, C., 2008a. Matching demand and supply in the agricultural knowl- Nagaoka, S., Kwon, H.U., 2006. The incidence of cross-licensing: a theory and new evi-
edge infrastructure: experiences with innovation intermediaries. Food Policy 33, dence on the firm and contract level determinants. Res. Policy 35, 1347–1361.
260–276. Nambisan, S., Sawhney, M., 2007. A buyer's guide to the innovation bazaar. Harv. Bus.
Klerkx, L., Leeuwis, C., 2008b. Balancing multiple interests: embedding innovation in- Rev. 85 (6), 109–118.
termediation in the agricultural knowledge infrastructure. Technovation 28, Nell, P.S.V., Lichtenthaler, U., 2011. The role of innovation intermediaries in the markets
364–378. for technology. Int. J. Technol. Intell. Plann. 7, 128–139.
Klerkx, L., Leeuwis, C., 2008c. Grassroots innovations in community energy: the role of Nelson, R.R., 2003. On the uneven evolution of human know-how. Res. Policy 32,
intermediaries in niche development. Global Environ. Change 23, 868–880. 909–922.
Knockaert, M., Spithoven, A., 2014. Under which conditions do technology inter- Nielsen, B.B., 2005. The role of knowledge embeddedness in the creation of synergies in
mediaries enhance firms' innovation speed? The case of Belgium's Collective Research strategic alliances. J. Bus. Res. 58, 1194–1204.
Centres. Reg. Stud. 48, 1391–1403. Oliver, A.L., 2004. On the duality of competition and collaboration: network-based
Knockaert, M., Spithoven, A., Clarysse, B., 2014. The impact of technology intermediaries knowledge relations in the biotechnology industry. Scand. J. Manage. 20, 151–171.
on firm cognitive capacity additionality. Technol. Forecasting Social Change 81, Oxford Economics, 2008. Study of the Impact of the Intermediate Research and
376–387. Technology Sector on the UK Economy. Oxford, UK.
Kodama, T., 2008. The role of intermediation and absorptive capacity in facilitating Paulraj, A., Lado, A.A., Chen, I.J., 2008. Inter-organizational communication as a rela-
university–industry linkages: an empirical study of TAMA in Japan. Res. Policy 37, tional competency: antecedents and performance outcomes in collaborative buyer–-
1224–1240. supplier relationships. J. Oper. Manage. 26, 45–64.
Kogut, B., Zander, U., 1996. What firms do? coordination, identity, and learning. Organ. Perkmann, M., Schildt, H., 2014. Open data partnerships between firms and universities:
Sci. 7, 502–518. the role of boundary organizations. Res. Policy 44, 1133–1143.
Krenz, P., Basmer, S., Buxbaum-Conradi, S., Redlich, T., Wulfsberg, J.P., 2014. Knowledge Petruzzelli, A.M., Rotolo, D., 2015. Institutional diversity, internal search behaviour, and
management in value creation networks: establishing a new business model through joint-innovations. Manage. Decision 53, 2088–2106.
the role of a knowledge-intermediary. Procedia CIRP 16, 38–43. Pisano, G.P., Teece, D.J., 2007. How to capture value from innovation: shaping in-
Landry, R., Amara, N., Lamari, M., 2002. Does social capital determine innovation? To tellectual property and industry architecture. Calif. Manage. Rev. 50, 278–296.
what extent? Technol. Forecasting Social Change 69, 681–701. Podsakoff, P.M., Organ, D.W., 1986. Self-reports in organizational research: problems and
Landry, R., Amara, N., Cloutier, J.S., Halilem, N., 2013. Technology transfer organiza- prospects. J. Manage. 12 (4), 531–544.
tions: services and business models. Technovation 33, 431–449. Polanyi, M., 1961. Knowing and being. Mind (N.S.) 70, 458–470.
Lee, S., Park, G., Yoon, B., Park, J., 2010. Open innovation in SMEs-An intermediated Polzin, F., von Flotow, P., Klerkx, L., 2016. Addressing barriers to eco-innovation: ex-
network model. Res. Policy 39, 290–300. ploring the finance mobilisation functions of insitituional innovation intermediaries.
Lichtenthaler, U., Ernst, H., 2006. Attitudes to Externally Organising Knowledge Technol. Forecastinf Social Change 103, 34–46.
Management Tasks: A review, reconsideration and extension of the NIH syndrome. Rhoten, D., Pfirman, S., 2007. Women in interdisciplinary science: exploring preferences
R & D Manage. 36 (4), 367–386. and consequences. Res. Policy 36, 56–75.
Lichtenthaler, U., Ernst, H., 2008. Intermediary services in the markets for technology: Rosenberg, N., 1965. Adam Smith in the Division of Labour: two views or one? Econ.
organizational antecedents and performance consequences. Organ. Stud. 29, (N.S.) 32, 127–139.
1003–1035. Rubin, D.B., 2008. For objective causal inference, design trumps analysis. Ann. Appl. Stat.
Lichtenthaler, U., Lichtenthaler, E., 2009. A capability-based framework for open in- 2 (3), 808–840.
novation: complementing absorptive capacity. J. Manage. Stud. 46, 1315–1338. Sapsed, J., Grantham, A., DeFillippi, R., 2007. A bridge over troubled waters: bridging
Lichtenthaler, U., 2007. The drivers of technology licensing: an industry comparison. organisations and entrepreneurial opportunities in emerging sectors. Res. Policy 36,
Calif. Manage. Rev. 49, 67–89. 1314–1334.
Lichtenthaler, U., 2013. The collaboration of innovation intermediaries and manu- Saviotti, P.P., 1996. Technological Evolution, Variety, and the Economy. Edward Elgar,
facturing firms in the markets for technology. J. Product Innovation Manage. 30, Cheltenham, UK.
142–158. Sawhney, M., Prandelli, E., Verona, G., 2003. The power of innomediation. MIT Sloan
Lin, R.J., Che, R.H., Ting, C.Y., 2012. Turning knowledge management into innovation in Manage. Rev. 44, 77–82.
the high-tech industry. Ind. Manage. Data Syst. 112, 42–63. Scardamalia, M., 2002. Collective cognitive responsibility for the advancement of
Lin, H., Zeng, S., Liu, H., Li, C., 2016. How do intermediaries drive corporate innovation? knowledge. Liberal Education in a Knowledge Society 97. pp. 67–98.
A moderated examination. J. Business Res. 69 (11), 4831–4836. Sethi, R., Smith, D.C., Park, C.W., 2002. How to kill a team’s creativity. Harv. Bus. Rev.
Lopez, H., Vanhaverbeke, W., 2009. How Innovation Intermediaries Are Shaping the 80, 16–17.
Technology Market? An Analysis of Their Business Model. University of Muenchen, Shadish, W.R., Cook, T.D., Campbell, D.T., 2002. Experimental and Quasi-experimental
Muenchen (MPRA Paper No. 20458). http://mpra.ub.uni-muenchen. de/20458/. Designs for Generalized Causal Inference. Houghton Mifflin, Boston.
Lui, S.S., 2009. The roles of competence trust, formal contract, and time horizon in in- Sieg, J.H., Wallin, M.W., von Krogh, G., 2010. Managerial challenges in open innovation:
terorganizational learning. Organ. Stud. 30, 333–353. a study of innovation intermediation in the chemical industry. R & D Manage. 40,
Mankins, J.C., 1995. Technology Readiness Levels [White Paper]. (Retrieved September 281–291.
7, 2016). https://www.hq.nasa.gov/office/codeq/trl/trl.pdf. Smedlund, A., 2006. The roles of intermediaries in regional knowledge systems. J.
Martín-de Castro, G., 2015. Knowledge management and innovation in knowledge-based Intellectual Capital 7, 204–222.
and high-tech industrial markets: the role of openness and absorptive capacity. Ind. Smith, K.G., Collins, C.J., Clark, K.D., 2005. Existing knowledge creation capability, and
Marketing Manage. 47, 143–146. the rate of new product introduction in high-technology firms. Acad. Manage. J. 48,
Martínez-Vela, C., 2016. Benchmarking Research and Technology Organizations (RTO): A 346–357.
Comparative Analysis. MIT Industrial Performance Centre, Cambridge MA (MIT-IPC Smith, A., 1776. An Inquiry into the Nature and Causes of the Wealth of Nations, 1863 ed.
Working Paper 16-005). Adam and Charles Black, Edinburgh.
Mazey, S., Richardson, J., 2006. Interest groups and EU policy making: organizational Spender, J.C., 1996. Making knowledge the basis of a dynamic theory of the firm. Strateg.
logic and venue shopping. In: Richardson, J.J. (Ed.), European Union: Power and Manage. J. 17, 45–62.
Policy-Making. Routledge, Oxon, pp. 247–268. Spithoven, A., Clarysse, B., Knockaert, M., 2010. Building absorptive capacity to organise
McElroy, M.W., 2002. Social innovation capital. J. Intellectual Capital 3, 30–39. inbound open innovation in traditional industries. Technovation 30, 130–141.
Melle, C., Russo-Spena, T., 2015. Innomediary agency and practices in shaping market Tabachnick, B.G., Fidell, L.S., 2007. Using Multivariate Statistics, 5th ed. Allyn and
innovation. Ind. Marketing Manage. 44, 42–53. Bacon, New York.
Metcalfe, J.S., 2002. Knowledge of growth and the growth of knowledge. J. Evol. Econ. Tashakkori, A., Teddlie, C., 1998. Mixed Methodology: Combining Qualitative and
12, 3–15. Quantitative Approaches. Sage, London.
Meyer, M., Kuusisto, J., Grant, K., De Silva, M., Flowers, S., 2017. Towards New Triple Teece, D.J., 2004. Knowledge and competence as strategic assets. In: Holsapple, C.W.
Helix Organisations? A Comparative Study of Competence Centres as Knowledge, (Ed.), Handbook on Knowledge Management. Spring Verlag, Berlin, pp. 129–152.
Consensus and Innovation Spaces. Manuscript, submitted to R & D Management. Tjong, T., Tai, S., Veraart, F., Davids, M., 2015. How the Netherlands became a bicycle
Meyers, L.S., Gamst, G., Guarino, A.J., 2006. Applied Multivariate Research: Design and nation: users, firms and intermediaries, 1860–1940. Bus. History 57, 257–289.
Interpretation. Sage, Thousand Oaks, CA. Tortoriello, M., Reagans, R., McEvily, B., 2012. Bridging the knowledge gap: the influence
Miles, J., 2003. A framework for power analysis using a structural equation modelling of strong ties, network cohesion, and network range on the transfer of knowledge
procedure. BMC Med. Res. Methodol. 3 (27), 1–11. between organizational units. Organ. Sci. 23, 1024–1039.
Miller, S., 2014. The strathclyde technology and innovation centre (TIC) in scotland’s Tortoriello, M., McEvily, B., Krackhardt, D., 2014. Being a catalyst of innovation: the role
innovation system. regional studies. Reg. Sci. 1, 145–151. of knowledge diversity and network closure. Organ. Sci. 26, 423–438.
Mils, P.K., Ungson, G.R., 2003. Reassessing the limits of structural empowerment: orga- Tran, Y., Hsuan, J., Mahnke, V., 2011. How do innovation intermediaries add value?
nizational constitution and trust as controls. Acad. Manage. Rev. 28 (1), 143–153. Insight from new product development in fashion markets. R & D Manage. 41, 80–91.
Morton, A., 1997. A Guide Through the Theory of Knowledge, second edition. Blackwell, Tsai, W., Ghoshal, S., 1998. Social capital and value creation: the role of intrafirm

86
M. De Silva et al. Research Policy 47 (2018) 70–87

networks. Acad. Manage. J. 41, 464–476. 1205–1223.

Van de Vrande, V., De Jong, P.J., Vanhaverbeke, W., De Rochemont, M., 2009. Open Yli-Renko, H., Autio, E., Sapienza, H.J., 2001. Social capital, knowledge acquisition, and
innovation in SMEs: trends, motives and management challenges. Technovation 29, knowledge exploitation in young technology-based firms. Strateg. Manage. J. 22,
423–437. 587–613.
Verona, G., Prandelli, E., Sawhney, M., 2006. Innovation and virtual environments: to- Yusuf, S., 2008. Intermediating knowledge exchange between universities and businesses.
wards virtual knowledge brokers. Organ. Stud. 27, 755–788. Res. Policy 37, 1167–1174.
Vlaar, P.W.L., Van Den Bosch, F.A.J., Volberda, H.W., 2007. Towards a dialectic per- Zellner, A., 1962. An efficient method of estimating seemingly unrelated regressions and
spective on formalization in interorganizational relationships: how alliance managers tests for aggregation bias. J. Am. Stat. Assoc. 57, 348–368.
capitalize on the duality inherent in contracts, rules and procedures. Organ. Stud. 28, Zukin, S., DiMaggio, P., 1990. Introduction to Structures of Capital. Cambridge University
437–466. Press, Cambridge.
Wallin, M.W., von Krogh, G., 2010. Organizing for open innovation: focus on the in- de Felice, A., 2015. Knowledge, Social Capabilities and Innovation in the Modern
tegration of knowledge. Organ. Dyn. 39, 145–154. Industrial District. McGraw Hill, Milano.
Wenger, E., Snyder, W.M., 2000. Communities of practice: the organizational frontier. del Águila-Obra, A.R., Padilla-Meléndez, A., Serarols-Tarrés, C., 2007. Value creation and
Harv. Bus. Rev. 110, 139–145. new intermediaries on Internet: an exploratory analysis of the online news industry
West, J., Gallagher, S., 2006. Challenges of open innovation: the paradox of firm in- and the web content aggregators. Int. J. Inf. Manage. 27 (3), 187–199.
vestment in open-source software. R & D Manage. 36, 319–331. van Lente, H., Hekkert, M., Smits, R., van Waveren, B., 2003. Roles of systemic inter-
Wright, M., Clarysse, B., Lockett, A., Knockaert, M., 2008. Mid-range universities’ lin- mediaries in transition processes. Int. J. Innovation Manage. 7 (03), 247–279.
kages with industry: knowledge types and the role of intermediaries. Res. Policy 37,

87