Tilburg University
Density and Strength of Ties in Innovation Networks
Gilsing, V.A.; Nooteboom, B.
Publication date:
2005
Link to publication in Tilburg University Research Portal
Citation for published version (APA):
Gilsing, V. A., & Nooteboom, B. (2005).
Density and Strength of Ties in Innovation Networks: An Analysis of
Multi-Media and Biotechnology
. (CentER Discussion Paper; Vol. 2005-41). Organization.
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Download date: 09. aug.. 2022
No. 2005–41
DENSITY AND STRENGTH OF TIES IN INNOVATION
NETWORKS: AN ANALYSIS OF MULTI-MEDIA AND
BIOTECHNOLOGY
By Victor A. Gilsing, Bart Nooteboom
February 2005
ISSN 0924-7815
1
DenseweakEmpirical
Density and strength of ties in innovation networks: An analysis of
multi-media and biotechnology
Victor A. Gilsing
Eindhoven University of Technology
Bart Nooteboom
Tilburg University
Abstract
In this article we provide an empirical illustration of hypotheses, developed in the literature,
on the role of density and strength of ties in innovation networks. We study both exploration
and exploitation networks in the Dutch multimedia and pharmaceutical biotechnology
industry. We find support for most of our hypotheses but not all. These findings, in line with
the mixed results in the literature, seem to indicate that the distinction between exploration
versus exploitation, albeit useful, is still too general. There may be a stronger sectoral effect in
how exploration and exploitation settle in network structural properties than anticipated thus
far.
Key words: innovation, networks, density, strength of ties, governance, biotechnology, multi-
media
JEL classification: D23, D83, D85, L14, L63, L65
Introduction
An important issue, in the network literature, is whether in networks for innovation ties
should be sparse and weak, to allow for variety, flexibility and low cost of exploration, as
claimed in the thesis of the ‘strength of weak ties’. Or alternatively, that ties should be
‘cohesive’, to facilitate trust and collaboration. These questions connect with a debate in the
sociological literature on networks where there are opposing views concerning the ‘strength
of weak ties’. On the one hand is the view that dense and strong ties (‘cohesion’ or ‘network
closure’) allow for social control, and facilitate the build-up of reputation, and social capital,
in the form of trust and social norms, which facilitate collaboration (Coleman 1988). On the
other hand, Granovetter (1973), in his famous article on ‘the strength of weak ties’, proposed
that weak rather than strong ties are appropriate for access to new information. He associated
strong (weak) ties with a dense (sparse) structure. In frequent and intense interaction between
many actors, in a dense structure, much of the information circulating in the system is
redundant. An example Granovetter used was the discovery of new employment
opportunities, through acquaintances with which one has only sporadic contacts. Burt (1992)
made a clearer conceptual separation between the strength and the density of ties. It is
important to acknowledge this, since it is conceivable that sparse ties may be strong and that
dense ties may be weak (Reagans and McEvily, 2003). Indeed, the present article will give
illustrations of that. According to the thesis of the strength of weak ties, a dense structure
2
yields redundancy, when the aim is access to new knowledge. If A is connected to B, and B is
connected to C, then A does not need a direct connection to C because he can access
information from C through B. The cost of redundancy, in setting up and maintaining ties,
increases with the strength of ties. Thus, according to Burt, efficiency can be created in the
network by shedding redundant ties and selectively maintaining only a limited set of ties that
bridge ‘structural holes’. Then, time and energy are saved for developing new contacts to
unconnected nodes.
Hansen (1999) made a distinction between acquiring knowledge about and knowledge from
others, i.e. between the identification of the location and usefulness of knowledge, and the
transfer or sharing of knowledge. He, and earlier Uzzi (1997), argued that strong ties promote
the transfer of complex knowledge, while weak ties promote the transfer of simple
knowledge.
In this debate, the empirical evidence is mixed. McEvily and Zaheer (1999) found evidence
against redundancy in an advice network, for the acquisition of capabilities. Ahuja (2000)
found evidence against structural holes, for innovation in collaboration. Walker, Kogut and
Shan (1997) found evidence in favour of cohesion, for innovation in biotechnology. In view
of these apparently inconsistent findings, subsequent studies have taken a ‘contingency’
approach (Bae and Gargiulo, 2003), investigating environmental conditions that would favour
the one or the other view (Podolny and Baron 1997; Ahula, 2000, 2001; Rowley, Behrens and
Krackhardt, 2000; Ahuja, 2000; Podolny 2001; Hagedoorn and Duysters, 2002).
Nooteboom and Gilsing (2004) developed additional hypotheses concerning density and
strength of ties for networks for exploration and for exploitation. The aim of this paper is to
provide an empirical illustration of these hypotheses, through a study of the development of
networks in two Dutch knowledge intensive industries. The test is not rigorous in the sense of
being quantitative, but gives an account of network development on the basis of a variety of
qualitative data. We study two types of networks in the multi-media industry and in
biotechnology.
This paper proceeds as follows. First, it summarizes the hypotheses. Second, it briefly
sketches our methodology of collecting qualitative empirical evidence that illustrates the
claims by Nooteboom and Gilsing (2004). Third, it provides an analysis of the development
of different types of innovation networks in the Dutch multimedia industry over the period
from the early 1990’s towards the early years of the new millennium. Next, over the same
period, it studies the development of innovation networks in the Dutch biotechnology
industry. Finally, it compares the empirical findings with the hypotheses, on the basis of
which we draw a number of conclusions.
1. The perspective and hypotheses
In their analysis of density and strength of ties for networks for exploitation and for
exploration, Nooteboom and Gilsing (2004) combined a perspective of ‘competence’, for the
acquisition of capabilities and knowledge, and the construction of new knowledge, with a
perspective of ‘governance’, for the management of ‘relational risks’ (Williamson 1999). The
literature on competence building has neglected the governance of relational risk, and
transaction cost theory has neglected issues of learning and innovation. A combination of the
two perspectives should yield a more complete understanding of inter-organisational relations.
So, there is a need to look at both potential rewards and risks, as is customary practice.
On the governance side, there are risks of dependence and of spillover, i.e. the risk that in
cooperation others adopt one’s knowledge to compete. For risks of dependence,
3
notwithstanding justified, fundamental criticism of transaction cost economics (TCE), the
notion of hold-up risk as a result of specific investments remains relevant, and indeed gains
new relevance in innovation networks. On the competence side, there are implications of
radical uncertainty, particularly in exploratory innovation, concerning the location and
identity of sources of information, and the relevance, absorbability, quality and reliability of
information. These yielded further support for existing arguments, and new arguments, for
density and strength of ties. The analysis considered six dimensions of tie strength. The
central hypothesis is that, for reasons of both competence and governance, in exploration ties
need to be dense and strong in most of the six dimensions, while in exploitation ties need to
be more sparse, and strong in other dimensions.
Nooteboom and Gilsing (2004) distinguish between networks for exploration and for
exploitation (March, 1991). Exploitation entails improvements with respect to established
practice, while exploration entails the development of new practices. In the literature on
innovation networks, the distinction between exploration and exploration was used earlier by
Rothermael and Deeds (2004). Exploration and exploitation build on each other: exploration
develops into exploitation, and exploration emerges from exploitation, in ways that go beyond
the present article. The point here is that networks for exploration and for exploitation may
require different structure and strength of ties.
In exploration, there is uncertainty about which technical standards will later yield the
‘dominant design’, there is much volatility of prototyping, the emphasis in competition lies on
technical feasibility and a ‘race to the market’, there is a great deal of trial and error, and
knowledge is often highly tacit. In exploitation, technical development has consolidated in a
dominant design, uncertainty in supply and demand has subsided, knowledge becomes more
codified and diffused, new players and consumers enter into the emerging market,
competition shifts to efficient production and distribution, and the emphasis shifts to a new
dominant design in organisation. These differences are summarised in Table 1. Next,
Nooteboom and Gilsing (2004) considered dimensions of tie strength. According to
Granovetter (1973: 1361), the strength of (personal) ties entails a combination of ‘amount of
time, emotional intensity, intimacy (mutual confiding) and the reciprocal services’ that
characterise the tie. In their earlier theoretical analysis Nooteboom and Gilsing (2004) arrived
at an adapted and extended set of dimensions of tie strength, based on combined perspective
of competence and governance. This is summarized in table 2.
The hypotheses for a network for exploration, proposed by Nooteboom and Gilsing (2004),
were as follows:
- Network structure needs to be sufficiently dense for three reasons. First, to hedge bets
on the future existence and relevance of sources of information. Second, to utilize
third parties to aid judgement of the meaning and value of knowledge (triangulation),
and to aid in its absorption. Third, to yield reputation mechanisms, needed in view of
the limited feasibility of contractual control. Cost of redundant relations is both
limited, in view of limited size of relation-specific investments, and of limited
relevance, since in exploration competition is less on price than on feasibility and fast
prototyping. Structure should not be too stable, allowing for sufficient entry and exit,
to enable variety of knowledge and flexibility of configuration.
- Ties are weak in control, but strong in terms of scope, frequency of interaction, trust
and mutual openness, and investment in mutual understanding. Trust is needed, next
to reputation mechanisms, due to the limited feasibility of contractual control (and
lack of hierarchy). Specific investments in mutual understanding require sufficient
duration to make them feasible and worthwhile. However, duration need
not be very
