The social aspects of industrial symbiosis: The application of social network analysis to industrial symbiosis networks

Abstract

This paper explores the potential of the application of social network analysis and network theory to the field of Industrial Symbiosis (IS), both as a methodological stance and as a conceptual framework, as a way to approach an understanding of the complexity of IS networks. We argue that the network approach can provide a comprehensive framework for the understanding of the social aspects behind material and energy exchanges. Aspects such as the structure of networks and the exchange and social conditions under which networks are likely to emerge and thrive are examined in the light of this approach. The main concepts of social network analysis and network theory are introduced and its applicability to IS discussed. A methodology is proposed to be applied to research the social and institutional aspects of IS. Some conclusions from the application of the methodology, its potential and shortcomings are presented.

Full text

Progress in Industrial Ecology – An International Journal, Vol. X, No. Y, xxxx 1
Copyright © 200x Inderscience Enterprises Ltd.
The social aspects of industrial symbiosis:
the application of social network analysis to
industrial symbiosis networks
Teresa Doménech* and Michael Davies
The Bartlett School of Graduate Studies
University College of London
WC1E 6BT, UK
E-mail: t.domenech@ucl.ac.uk
E-mail: michael.davies@ucl.ac.uk
*Corresponding author
Abstract: This paper explores the potential of the application of social network
analysis and network theory to the field of Industrial Symbiosis (IS), both as a
methodological stance and as a conceptual framework, as a way to approach
an understanding of the complexity of IS networks. We argue that the network
approach can provide a comprehensive framework for the understanding
of the social aspects behind material and energy exchanges. Aspects such as
the structure of networks and the exchange and social conditions under
which networks are likely to emerge and thrive are examined in the light of
this approach. The main concepts of social network analysis and network
theory are introduced and its applicability to IS discussed. A methodology is
proposed to be applied to research the social and institutional aspects of IS.
Some conclusions from the application of the methodology, its potential and
shortcomings are presented.
Keywords: social network analysis; industrial symbiosis; network structure;
exchange conditions; social mechanisms of control; dyads; nodes.
Reference to this paper should be made as follows: Doménech, T. and
Davies, M. (xxxx) ‘The social aspects of industrial symbiosis: the application
of social network analysis to industrial symbiosis networks’, Progress in
Industrial Ecology – An International Journal, Vol. X, No. Y, pp.000–000.
Biographical notes: Teresa Doménech is a PhD candidate at the Bartlett
School of Graduate Studies, UK. Her PhD focuses on the analysis of the
provision of a modelling framework for the analysis of the social and
institutional factors of industrial symbiosis based on social network analysis,
grounded theory and discourse analysis. She has a background as a Researcher
and Consultant in environmental and sustainability issues. She has taken
part in an EU-funded project to research the applicability of industrial ecology
principles to the design of industrial estates and planning policies. She has also
worked as an Environmental Consultant providing assistance to industrial land
developing companies.
Prof. Michael Davies is the Manager of the Complex Built Environment
Systems (CBES) group at the University College of London (UCL), UK.
He has a background of monitoring and modelling work in the field of the
built environment – an example of a current project is ‘The Development
of a Local Urban Climate Model and its Application to the Intelligent
Development of Cities’. He is also currently managing the UCL components
of research projects being undertaken for the Department of Communities

2 T. Doménech and M. Davies
and Local Government (CLG) in the Sustainable Buildings Division. These
research projects contribute to the development of the Building Regulations
(England and Wales), in particular, Parts F and L, which deal with ventilation
and energy.
1 Introduction
The field of industrial ecology aims to transform industrial systems into cyclical systems
so that the “waste of one process can be used as resource for another process” (Frosch
and Gallopoulos, 1989). Within this field, Industrial Symbiosis (IS) has emerged as
a body of exchange structures to progress to a more eco-efficient industrial system,
by establishing a collaborative web of material and energy exchanges among different
organisational units. Even though the area has attracted a lot of academic attention
and has been reported to lead to economic and environmental benefits (Chertow and
Lombardi, 2005), most of the contributions focus on the engineering and technical
feasibility of the exchanges and there is still limited understanding of the social
mechanisms that allow these networks to emerge and thrive and the conditions under
which they can operate and lead to the reported outcomes. We argue that social network
analysis provides a framework for the understanding of the social aspects of IS, offering
insights to the process of formation and operation of IS networks as well as the
mechanisms that can foster their emergence. We assume that IS exchanges cannot be
considered in isolation of the social context in which they take place and that the
understanding of this context is crucial for the design of policy action to promote IS.
Behind the material flows occurring in the industrial system there is a complex
web of actors, with different and sometimes conflicting, interests, that are interacting
with each other and who are, in fact, defining the actual realisation of the physical
flows. Therefore, we consider that IS networks are ‘embedded’ (Uzzi, 1996) in
social systems and, as such, decision-taking processes are shaped by social relations.
Regulation systems, trust, beliefs and knowledge are in fact crucial aspects influencing
the direction and management of physical exchange and therefore, defining the scenario
for eco-industrial approach to be adopted. Understanding these aspects is a first step in
the design of policy frameworks for the promotion of IS.
The aims of this paper are: (a) to present a methodological approach based on social
network analysis and network theory to contribute to the understanding of the social
dimension of IS networks and (b) to discuss the issues that arise from the implementation
of the methodology proposed.
The paper has been structured as follows. Firstly, the main aims of the paper are
presented and some insights to the application of social network analysis to understand
the social aspects of IS are provided. Secondly, the integration of the social dimension
in the areas of industrial ecology and IS is briefly reviewed. The most relevant concepts
of social network analysis and network theory are presented in the next section and
their application to IS discussed. Then, the methodological approach to the study of the
social aspects of IS networks is proposed; and finally, the potential and shortcomings of
the methodology are discussed and some conclusions are drawn.

The social aspects of IS: the application of social network analysis to IS networks
3
2 Understanding the social aspects of industrial symbiosis
Traditionally it has been assumed that there is a tradeoff between economic
and environmental performance. The theory of ecological modernisation (Mol and
Sonnenfeld, 2000) challenges this idea, introducing a new perspective for understanding
environmental and economic policy in mutually reinforcing ways. Within the framework
of ecological modernisation, the emerging field of industrial ecology aims to improve
the resource and energy efficiency of the industrial system as a whole while reducing
pollution, through cycling the material and energy flows throughout the system.
In moving towards a closed-loop system, the approach of IS focuses on maximising
material and energy recovery within specific areas or regions by fostering the
creation of interfirm networks for energy and material exchange. It thus aims to bind
companies together into a coherent and innovative system of collaborative linkages and
interorganisational alliances aiming to reduce the environmental impact of industrial
activity in an economically rational way. This approach has been reported to provide
positive economic and environmental outcomes for the organisations involved as well as
for the community as a whole (Burnström and Korhonen, 2001).
Although the field of industrial ecology was at first mainly addressed by engineering
approaches (see, for example, Ayres and Ayres, 1998; 1999), considerable progress
has been undertaken in introducing economic, social and institutional aspects in the
understanding of the decision-making structures that govern material and energy flows.
The integration of a social perspective in the field of industrial ecology, has taken place at
two different levels: at the conceptual and at the methodological level. Binder (2007a–b)
provides a comprehensive review of attempts to couple social science modelling
approaches with material flow analysis. Most of these attempts stem from the field
of economics (Binder, 2007a). General or partial equilibrium models, input-output
approaches or microeconomic models have been proposed to analyse the relationship
between material flows and economic factors (Binder, 2007a). However, as noted by
Kytzia et al. (2004), “all these approaches lack a common system definition for both
physical and economic dimensions of industrial system”. The economic analysis is,
thus, often introduced as an ‘add-on’ dimension. To overcome the shortcomings and
inconsistencies identified in the previous attempts to integrate material and economic
dimensions, Kytzia et al. (2004) proposed the ‘Economically Extended Material Flow
Analysis (EE-MFA)’ as a ‘combined analysis of material and money/cash flows’, which
derives from the integration of Material Flow Analysis and Input-Output analysis. Even
though these approaches have brought a wider understanding of the transition process
towards a more cyclical approach to material and energy throughout the industrial
system, as Binder (2007a) points out, they fail to address other important aspects shaping
decision-making processes, such as ‘personal preferences and social norms or cultural
backgrounds’. These aspects are difficult to capture by quantitative economic models.
Therefore, in the transition towards a more sustainable organisation of the industrial
system (Korhonen, 2004b), the integration of material and economic systems might not
be sufficient for the ‘system understanding’. The combination of material flow and agent
analysis (Binder et al., 2004) provides a useful approach for a wider understanding of the
agent networks as well as the social and cultural norms that shape the definition of the
‘environmental’ problem. Building on this, Binder (2007b) provides a comprehensive
framework for the integration of social aspects in industrial ecology, based on Gidden’s

4 T. Doménech and M. Davies
structuration theory. A methodological approach, ‘Structural Agent Analysis’, is
proposed to analyse the impact of social structures on action as well as “the feedback
from action on social structures” (Binder, 2007b). According to this approach, social
structures are divided into rules and resources. The action of an agent is thus shaped by
the social structure and it has an impact on the environment. At the same time, changes in
the environment affect the agent, his awareness and perception of the problem, and,
diachronically, the social structure. A seven-step method is proposed (Binder, 2007b),
consisting of:
Step 1 identifying relevant agents
Step 2 defining the relevant structural factors affecting agents’ actions
Step 3 weighting of structural factors on agents’ actions
Step 4 drawing an ‘agent-structure diagram’
Step 5 identifying agents’ options and restrictions
Step 6 identifying potential interactions and interferences among agents
Step 7 estimating the potential effects of agents’ actions on structure.
All these previous studies have contributed to the progressive integration of the social
dimension in the analysis and understanding of the material and energy flows, in the
field of industrial ecology. Although many of these approaches offer guidance for
the understanding of the ‘soft components’ of IS, there is still a lack of integrative
approaches for the evaluation, analysis and understanding of the social and organisational
dimension in IS developments. IS initiatives have tended to focus on engineering and
technical solutions for the connection of processes and activities for the exchange of
materials and energy. Even though the engineering approach of IS sets the basis for
the functioning of these exchange networks, the authors argue that social aspects that
influence the decision-making process of single units constitute critical factors of the
development of IS. The growing attention that IS has recently received in academic
and policy spheres has not always been accompanied by well-defined strategies for its
implementation at the practical level and the use of modelling techniques is still limited
(Doménech and Davies, 2006). On the empirical level, IS networks are still a ‘rare’
phenomenon and, although new IS projects have been launched, the ‘success’ of many of
these initiatives has been questioned (see, for example, Goss et al., 2005), raising doubts
about the ability of planning initiatives and policies to promote IS (Desrochers, 2002).
Even though the importance of social aspects has been increasingly recognised in
the literature (see, for example, Eilering and Vermeulen, 2004; Korhonen et al., 2004),
as it has been pointed out before, most of the analyses fail to provide a comprehensive
framework for the evaluation of the social aspects of IS and the understanding of the
dynamics of collaboration and networking.
Examining the social factors of IS requires an understanding of the rationale of
the individual units in interaction with their context, as companies and networks are
influencing each other on a continual basis. This will lead to a subsequent question: why
may networking and cooperation provide better results in environmental and economic
terms compared to other forms of interaction or environmental management alternatives?
Indeed, it is necessary to emphasise that networking has implications that differ from

The social aspects of IS: the application of social network analysis to IS networks
5
other market-based waste management solutions. Understanding these differences
may also help to distinguish when IS initiatives are genuinely operating as networks
or when they only consist of market exchanges of material and energy by-products.
These two forms of governance have different implications and may lead to diverse
results. Therefore, the strategy to promote one or the other necessarily differs. Indeed,
IS literature has generally emphasised in its discourse the ‘cost-based’ approach as a
main driver for the emergence of IS networks (see, for example, Lowe and Evans, 1995).
However, when doing so, they are assimilating IS networks to atomistic market
conditions, overlooking the ‘structural’ conditions that might actually set the basis of IS
networks and that can better explain the emergence of IS networks. Binder et al. (2004)
provides an example, applied to the wood flows in a Swiss region, of how structural
conditions (‘rules’ and ‘resources’) can foster or hinder the progress to more
sustainable scenarios. In this paper, we will argue that IS networks are closer to
the theoretical construct of networks rather than market-based exchange conditions
and therefore, even though the potential for IS networks to reduce the costs associated
with environmental management is recognised, we contend that the central elements
in explaining the network exchanges challenges a narrow ‘costs-view’ (Larson, 1992),
falling closer to the ‘heuristics’ approach (Lewis and Weigert, 1985). According to the
latter, the decision-making process of companies interacting in a network would not
only be influenced by cost but other rather more ‘qualitative’ factors, such as trust
and information, which would have a primary influence (see also Begré and Hadorn,
2002; Hadorn et al., 2002).
3 Introductory concepts of social network analysis:
structure and organisation
Social Network research has steadily grown in recent years and has been applied to many
disciplines and social contexts. Social network analysis introduces a new perspective
in the examination of social environments and processes, by focusing the attention on
the relational information, that is, the relationships among interacting social units, rather
than in the units themselves. From this exploration, some patterns or regular traits can
be identified, giving rise to the ‘Structure’ (Wasserman and Faust, 1994). This approach
encompasses a multiplicity of theories, models and methods of analysis that put
the emphasis on the relational content and the linkage structure between different
organisational units. Indeed, as pointed out by Cook and Whitmeyer (1992), the social
network analysis approach results from a combination of a variety of sources:
• ‘empirical work in social anthropology’
• ‘sociometry’
• ‘mathematical models and theory such as biased net theory and graph theory’.
Notwithstanding this diversity of perspectives, Wasserman and Faust (1994) propose a
core of central conceptual principles underlying network analysis:

6 T. Doménech and M. Davies
• Actors are conceptualised as interdependent and interacting units, linking to
each other.
• The linkages between actors act as channels for the transfer of resources of
different nature.
• Network structure shapes the action of individual units/actors by providing
opportunities and constraints.
• The (social, economic, etc.) structure is configured by ‘lasting patterns of relations
among actors’.
On the basis of these principles, this analytical framework has offered significant
theoretical outcomes in a wide range of disciplines, from communication and
decision making (see, for example, Laumann et al., 1977; Laumann and Pappi, 1973),
to entrepreneurship (see, for example, Hoang and Antoncic, 2003), competition
(see, for example, Uzzi, 1997; Spencer, 2003), community, social relations and
friendship/support structures (see, for example, McPherson et al., 2001), social capital
(Walker et al., 1994), power (see, for example, Ouimet et al., 2004) or exchange
and innovation (see, for example, Valente, 1995; Powell et al., 1996). The approach
of social network analysis has also been applied, alone or in combination with other
methodological or conceptual frameworks, to the field of sustainability. Hermanns
(2004), in his analysis of the ‘diffuse pollution problem’ in the province of North
Holland, proposes the use of ‘Dynamic Actor Network Analysis’, an approach based on
social network analysis, as a useful tool to “gain an overview of the different actors
involved in the problem”, by providing insights to the understandings, interests, conflicts
and values of the different actors and the interaction among them. Also focused on the
Actor Network, the heuristic tool ‘options and restrictions’, developed by Begré and
Hadorn (2002) and Hadorn et al. (2002), provides a method for the analysis of the
structural processes that lead actors to act in sustainable ways, or, on the contrary, keep
them from sustainable options. Although it can be conceptualised as a hybrid method,
that combines elements and approaches from different social theories, the emphasis on
the interaction among the actors, as a determinant factor in shaping actor’s decisions and
choices, and the way this interaction supports or constraints other actors’ actions, clearly
points to the underlying framework of social network analysis. Binder (2007a–b) also
bases its ‘Agent Analysis’ on social network analysis. Even though all these contributions
point to the potential of social network analysis as a conceptual and methodological
framework for the analysis of the social aspects of industrial ecology, the application of
this approach to the study of the emerging field of IS still needs to be explored. In the
opinion of the authors, the implementation of social network analysis can offer some
useful insights to social and policy aspects of IS networks, an area, that although being
recognised as relevant (see, for example, Sterr and Ott, 2004; Haskins, 2006; Mirata and
Emtairah, 2005; Jacobsen, 2006), lacks an integrative analytical framework to date.
Table 1 briefly presents some of main concepts of social network analysis that can be
applied to the study of IS networks. A comprehensive review of the concepts, methods
and applications of social network analysis can be found in Wasserman and Faust (1994)
and Scott (2000).

The social aspects of IS: the application of social network analysis to IS networks
7
Table 1 Applicability of social network concepts to IS
Concept Definition Applicability to IS
Network Larson (1992) offers a definition of
networks as “distinct from market
and hierarchical arrangements in
their heavy reliance on reciprocity,
collaboration, complementary
interdependence, a reputation
and relationship basis for
communication, and an informal
climate oriented toward
mutual gain”.
Even though there is no unified
concept of a network, some general
characteristics of networks can be
inferred from the literature:
• interactions are based
on cooperation
• informal arrangements are
preferred to formal contracts
• exchange relations tend to be
long-term and recurrent
• there is frequent and reciprocal
communication between
network members
• shared rules and (tacit) agreed
norms govern the operation of
the networks
• reciprocity and mutual gain
is pursued
• interactions rely on close social
relationship and trust.
Most of these general characteristics apply
to IS networks. Yet, some peculiarities
should be taken into account. IS networks
focus on a single aspect of the company
performance, that is, environmental issues,
which have not traditionally formed the core
of the business strategy. Therefore, the
culture of the company will condition the
predisposition of the firm to engage in IS
networks. Moreover, the participation of the
firm in IS networks might influence the
overall firm’s strategy, if the collaborating
culture permeates other areas of the activity.
However, it should be noted that not all
IS exchanges are organised as networks.
IS exchanges based on market mechanisms
consist of one-off exchanges, formally
arranged by purchase contracts and
regulated by prices. These ‘arm’s length
ties’ imply that there is no recurrent
interaction between companies and
therefore no space for reciprocity. In this
scenario cost/price would be the main
coordination mechanism (Uzzi, 1997).
On the contrary, IS based on network
governance is characterised by recurrent
interaction between the companies, informal
arrangements (although this might also be
formalised by contracts), reciprocity and
trust. In this scenario, though cost is still an
important aspect, it might not act as the
main coordination mechanism. Cost is
evaluated in a global way, taking into
account not a single transaction but the
overall interaction.
Nodes, ties
and dyads Actors are represented by nodes that
are linked by lines “representing the
direct ties between a pair (dyad)”
(Yang and Knoke, 2001). The ties
between the nodes can be either
directed or asymmetric or either
undirected or symmetric. The
latter imply the reciprocity of the
interaction, while direct ties
represent one-way relations between
the actors.
In the case of IS, the organisations linked by
IS exchanges represent the ‘nodes’ and the
exchanges the ‘ties’ between them.

8 T. Doménech and M. Davies
Table 1 Applicability of social network concepts to IS (continued)
Concept Definition Applicability to IS
Path distance The ‘path’ can be described as the distance between
two nodes, represented by the set of nodes and lines
that connect them. The shortest path between two
nodes, that with the smallest length, is referred to
as the ‘geodesic’ path (Yang and Knoke, 2001).
The optimal connection “involves a combination of
the fewest intermediaries with the most intense
interactions” (Yang and Knoke, 2001).
The concept of path distance
can be fully applied to IS
networks, based on the
number of lines to connect
two organisations.
Centrality The simplest indicator of centrality is “the number
of ties that a network member has to other
members” (Walker et al., 1994). However, different
measures of centrality have been developed to try to
capture the importance of a node in a network.
This measure is an indicator
of the position of different
nodes in the network and
the relevance of the role
played by each of them.
Characterisation
of network
structure
A multiplicity of other factors might in fact also
influence network structure:
• openness/closeness of the network – This refers
to the entry of new members to the network in a
certain period of time. Open networks promote
the integration of new members while closed
networks tend to establish ‘barriers for entry’
that discourage the engagement of new members
in the network.
• weak versus strong bonds, depending on the
continuity, frequency and quality of the
interaction between the members
• trust and past history shared – This links with
the notion of stability, which refers to the
evolution of network’s pattern over time.
• geographic distance – local versus
extent networks.
All these aspects are of
relevance in the analysis of
the structural aspects of
IS networks.
Reciprocity The term ‘reciprocity’ in social network theory
has been used to refer to mutually beneficial
arrangements considered in a global and long-term
perspective (Walker et al., 1994). This implies
taking into account:
• direct gains involved in a transaction
• a generalised sense of reciprocity, in which the
linkage is not compensated in concrete terms or
as a result of a specific transaction but it may be
compensated in future transactions or with some
other kind of support
• network balancing, where the repayment may
come from the network as a whole as a reward
from a concrete network member contribution to
the network or to one of its members (Walker
et al., 1994).
This notion of reciprocity
fully applies to IS networks.
IS literature has emphasised
that companies in IS
networks are engaged in
mutually beneficial
exchanges. However, when
analysing concrete IS dyads,
cooperative links also seem
to occur when no direct
payback is attached to them,
analysed on a single
transaction basis. In this last
case, the rationale of the
behaviour has to be linked
to more subtle and
intertemporal framework as
the one presented above,
taking into account
generalised reciprocity and
network balancing.

The social aspects of IS: the application of social network analysis to IS networks
9
Table 1 Applicability of social network concepts to IS (continued)
Concept Definition Applicability to IS
Trust Trust lies at the core of the network form of
governance and stands as a major coordination
mechanism in network-based transactions.
According to Lewis and Weigert (1985) certain
conditions allow trust to emerge:
• all participants must believe that action is
aimed at common values
• these common values have to be translated
into common goals
• there must be a shared sense of solidarity in
accordance with participants expectations
• trust must be reinforced by empirical evidence
and past action.
Some contextual aspects such as the homogeneity
of the group and the level of connectedness
between members might actually positively affect
the levels of trust. Lewis and Weigert (1985)
also point to the size of the network as an aspect
negatively influencing the levels of trust.
Moreover, the process of building of trust seems
to be directly linked to the sharing of a common
past-history. Baker (1987) shows the influence of
pre-existing ties between companies and the
interpersonal ties between managers and
employees in subsequent cooperative
relationships in the network.
Lewis and Weigert’s (1985)
concept of trust brings
understanding to the
conditions that might foster
the process of building trust in
IS networks and its role in
network development.
Embeddedness Uzzi (1997) analyses the performance of
embedded networks and its implications for
efficiency and company performance and defines
embedded networks as characterised by three
main features:
1 trust
2 fine-grained information transfer
3 joint-problem solving.
These three features would allow companies
engaged in embedded ties to be more flexible and
adapt more quickly in environments characterised
by complexity and continuous change. As a result
of this, companies engaged in embedded networks
can gain advantages in comparison to other forms
of governance (Uzzi, 1997).
When characterising IS
networks, attention must
be paid to the level of
embeddedness of its ties,
as it might be a good indicator
of the performance of
the network.

10 T. Doménech and M. Davies
4 Network governance theory and the emergence of networks:
implications from theory
Social network analysis constitutes a collection of analytic tools and techniques focused
on the linkages among interacting units. However, as argued by Cook and Whitmeyer
(1992), social network analysis needs to be combined with other theories to provide
“theoretical underpinning for network conceptions of structure”, allowing the progress
from a descriptive to a prescriptive dimension, in an attempt to identify the direction of
the causalities and effects of structure. Some examples of the combination of Network
Analysis and other theoretical frameworks have been given above (Hermanns, 2004;
Begré and Hadorn, 2002; Hadorn et al., 2002). The combination of concepts and
methods from Network Analysis with Giddens’ structuration theory has an antecedent in
Haines (1988) and is also used, in the field of sustainability, by Binder (2007a–b) as a
baseline theoretical foundation. In an attempt to generate a ‘Network Theory’ (Cook and
Whitmeyer, 1992), that is, a comprehensive theoretical framework for the application of
social network analysis, Jones et al. (1997) proposed the General Network Theory.
This theory tries to identify the conditions under which ‘network governance’ tends to
‘emerge and thrive’. This aspect is of crucial interest to IS initiatives, as it may help to
understand under what structural conditions the exchange of by-products can succeed,
or, from a policy view, what conditions should be favoured to promote intercompany
cooperation in environmental issues.
According to this theory, and, assuming that “social mechanisms influence the cost
of transacting exchanges”, network governance is likely to emerge in environments
characterised by “the need for high adaptation, high coordination and high safeguarding”
(Jones et al., 1997). In the proposed theory, Jones et al. (1997) also show how
structural embeddedness “provides the foundation for social mechanisms to appear such
as restricted access, macrocultures, collective sanctions and reputations”.
These concepts need to be adapted and empirically contrasted in the case of IS
networks. By applying and adapting network governance theory to IS, we aim to define a
coherent framework that identifies the conditions under which IS networks are likely to
emerge. In the case of IS networks it is necessary to identify under which contexts and
conditions structural embeddedness appears, providing the ground for social mechanisms
to develop, reducing the transaction costs, and thus favouring the operation of the
network in a highly competitive and rapidly evolving environment.
4.1 Exchange conditions for network governance
Jones et al. (1997) distinguish between different kinds of interfirm governance:
• hierarchies
• market solutions
• network governance.
Network governance stands as a distinctive form of interfirm organisation characterised
by informal social structures to coordinate complex products or processes ‘based on
implicit and open-ended contracts’ to adapt to uncertain and competitive environments.
In their proposed theory, the authors point to three main exchange conditions which
would determine the form of governance adopted:

The social aspects of IS: the application of social network analysis to IS networks 11
1 uncertainty
2 asset specificity
3 frequency of contact.
Network governance thus would be more likely to emerge when (Jones et al., 1997) there
are “high adaptation needs, ... high coordination needs and ... high safeguarding needs”,
as a way of managing uncertainty.
These conditions, however, need to be adapted to the specific context of waste-flow
exchanges. From the analysis of the cases reported in the literature of IS and the
empirical research undertaken by the authors, some main conditions seem to characterise
the contexts where IS develop:
• Stringent and rapidly evolving regulatory frameworks with regard to waste
management and environmental performance that require high adaptability from
companies. In this context knowledge dissemination becomes a crucial aspect.
• Waste-flow exchanges require customised, nonstandard, applications or involve
an innovative component or approach, and, therefore, imply uncertainties with
regards to the outcomes and process. Customised solutions generally require the
development of routines, team works and protocols, enhancing cooperation and
the transfer of tacit knowledge between the parts. Having developed a cooperative
culture with one or more partners makes it easier to generate multiple projects in
what Jones et al. (1997) refer to as ‘relationship-specific capital’.
• As a result of the need for customised solutions, high coordination is required,
which implies frequent interaction between companies, favouring the transfer
of tacit knowledge, ‘learning by doing’, and the creation of a shared culture or
‘macroculture’ (Jones et al., 1997). Having developed ‘relationship-specific capital’,
frequent exchanges reduce the transaction costs between partners that have learnt
‘to speak each others’ language’.
The fact that companies are faced with the uncertainty of an increasingly stringent
regulatory framework and the need for coordination and integration that derives
from the customised solutions, pushes companies to find forms of governance in the
environmental area that foster cooperation while maintaining the independency of the
firm in other areas. Network governance emerges then as a suitable form of facing
environmental challenges in a cost-effective way.
4.2 Social control mechanisms
Structural embeddedness provides the basis for the development of social control
mechanisms that regulate transactions in network structures. Jones et al. (1997) mention
the following social mechanisms, which contribute to the emergence of the network form
of governance:
• restrict access to exchanges
• macroculture
• collective sanctions
• reputation.

12 T. Doménech and M. Davies
These social mechanisms help reduce the costs associated with the coordination and
safeguarding of the exchanges. Exploring the presence of these mechanisms in operative
IS networks may help to understand the role played by them in favouring the emergence
and development of this kind of networks.
Restricted access occurs through ‘status maximisation strategy’, that is, avoiding
partners with lower status and through ‘relational contracting’. Both strategies emphasise
the importance of past interactions; by reducing the number of partners a company
interacts with, it reduces the costs associated to coordination and negotiation as well
as guaranteeing quality standards based on previous experience. Moreover, restricted
access and recurrent interaction reduces the chances of opportunistic behaviour among
partners (Tutzauer et al., 2006). However, very severe exchange restrictions would
reduce the opportunities and synergies, being an “intermediate level of restriction”
optimal (Huberman and Hogg, 1995; Jones et al., 1997).
The macroculture defines a common understanding that includes knowledge, values,
approaches and patterns of action. The higher the degree of structural embeddedness,
the more network members identify with a macroculture. The dissemination of a
macroculture enhances coordination and integration as a result of three mutually
reinforcing processes (Jones et al., 1997):
1 by creating ‘convergence of expectations’
2 by defining a common language to ‘summarise complex routines and information’
3 by generating ‘tacit rules’ and norms for the interaction of actors.
It is also important to notice that the content of the macroculture can be an element
fostering or hindering cooperation. Thus, the development of a culture that stresses
the value of intercompany interaction and the positive impact of cooperation may be a
crucial factor in the further integration among parties. In this sense, and with regard to IS,
the development of a broader ‘macroculture of cooperation’ between companies in other
areas of business can set the basis for the cooperation on environmental management
and exchange of waste-flows. Similarly, IS can be the first step into the definition of
cooperative patterns in other areas of the company.
Collective sanctions, whether tacit or explicit, condemn deviating behaviour and the
violation of norms and rules of network operation. The sanctions generally affect one’s
reputation, reducing the changes of further interaction and can even lead to detachment
from the network. Collective sanctions are generally supported by ‘meta-norms’ that
also punish “those who do not punish devian(cy)” (Jones et al., 1997), increasing the
social control of ‘offenders’. Connected to collective sanctions, reputation stands as an
important mechanism to reduce the uncertainty of exchanges. Reputation refers to an
overall picture of one’s skills, reliability and capacity to meet specifications, based on
past experiences. Reputation reduces the risk associated with the exchange by providing
information of one’s past capacity and performance. Therefore, opportunistic behaviour
or failure to meet specifications have a negative impact on reputation and involve social
sanctions with economic consequences.
As more of these social mechanisms are put in place, the likelihood of a network
form of governance to develop rises. These mechanisms “favour cooperation in the face
of collective or social dilemmas” (Jones et al., 1997). Companies when faced with the
decision to cooperate or to defect would have more incentives to cooperate, achieving a
better social (economic and/or environmental) outcome (Tutzauer et al., 2006).

The social aspects of IS: the application of social network analysis to IS networks 1
3
Social mechanisms are used in network governance as a way to deal efficiently with
highly uncertain scenarios, characterised by customised exchanges and complex tasks.
Examining the social mechanisms of control and structural embeddedness can thus
provide a basis for the identification of the conditions under which IS networks are
more likely to emerge and develop. A priori, these conditions seem to be present in
spontaneously developed IS networks, such as Kalundborg (see Section 5), where the
development of the cooperative links between the companies have been generated under
the pressure of strict regulatory frameworks, scarcity of raw materials and the need to
adopt ad hoc, nonstandardised solutions to manage waste flows.
5 Understanding the social aspects of industrial symbiosis networks:
methodological framework
A methodology has been designed, combining elements from social network analysis
and network theory (Uzzi, 1996; Uzzi, 1997; Jones et al., 1997) to contribute to the
understanding of the social dimension of IS, by examining the relationships occurring
between all the members of the network and the structures that are supporting or
hindering those relations (Binder, 2007b). The approach described here has been
structured as a four-step method consisting of (Schensul et al., 1999):
Step 1 mapping of the main participants and identification of the boundaries of
the network
Step 2 analysis of the core-periphery structure for the network; potential local bridges
are also identified
Step 3 evaluation of the structural characteristics of the network
Step 4 examination of the exchange conditions and degree of embeddedness;
discourses within the network are studied to define the tacit rules of operation
that govern the network, giving an insight of the content of the macroculture and
the social mechanisms of control that provide the basis for the realisation of the
IS exchanges.
Although the process is presented as sequential, in practice all the phases are interrelated
and in the process of implementation, phases can merge or change the proposed order,
as part of an iterative process. Indeed, Steps 1 and 2 are generally undertaken together as
one initial phase.
5.1 Research methods and data sources
The input information for the methodology presented here consists of a combination
of qualitative and quantitative analysis (Wasserman and Faust, 1994; Scott, 2000), though
qualitative methods have been giving priority in an attempt to identify the ‘soft’ elements
of IS networks, which quantitative data cannot capture. The methodology presented
has been applied to date to two case studies, although an additional case study is being
considered. The case study allows a deep immersion into a specific context, helping to
unravel the subtle and tacit processes (Yin, 1989) behind the outcomes and performance
of an IS network. This allows a better understanding of the direction of the causalities

14 T. Doménech and M. Davies
and the interaction among different elements and actors, not only by their explicit
strategies but also by the ‘hidden agendas’ (Hermanns, 2004). However, one of the
caveats of the case study approach is that, as it is context-dependant, the possibilities of
identifying general patterns are diminished (Miles and Huberman, 1994). Therefore, the
methodology selected here follows a multiple-case design, where cross-case comparisons
can help to shed some light on more general rules and conclusions about the social
aspects influencing IS, transcending the limit of a particular case. The two case studies
selected for the analysis of the socio-institutional aspects of IS networks are: Kalundborg
(Denmark) and National Industrial Symbiosis Programme (NISP) (UK). These cases
stand as ‘successful’ and operational examples of IS networks, although they differ
greatly in their structural characteristics. From a comparative analysis, some structural
features of successful networks can be identified, helping to determine the conditions
under which IS networks are likely to emerge and develop. For the selection of the case
studies we have followed an ‘information-oriented’ strategy (Flyvbjerg, 2004). It is not
the purpose of this paper to offer a full description of the cases or of the findings
derived from the application of the methodology; rather some references will be made as
examples to illustrate the application of the methodology. A detailed description of the
case studies and findings will be reported in a subsequent paper. Table 2 summarises the
selection criteria of the case studies.
Table 2 Selection criteria for the case studies
Case study Strategy Criteria
Kalundborg
(Denmark) Paradigmatic
case
• most referred case in the literature
• this network has performed especially well and have established
precedents and inspired models of operation of industry based on
collaboration between companies
• has emerged as model of eco-industrial parks and IS networks
• long-term development
• spontaneously company-driven process
• high levels of cooperation and integration between the companies
• exchange of by-products and waste streams in continuous and
profitable ways
• the collaboration of companies has extended to areas other than
the exchange of material and energy flows.
NISP (UK) Critical case
• has achieved a wide recognition within UK-based industrial
sector in a short period of time
• focuses on the diversion of materials from landfill and the
cutdown of CO2 emissions
• operates as a national programme but has a network of support
teams in each region of UK
• it is not geographically bound
• virtual eco-industrial park
• cross-sectoral and cross-regional exchanges are considered
• central coordination exercised by a third party

The social aspects of IS: the application of social network analysis to IS networks 1
5
For each case study, primary data has been collected mainly though semistructured
and in-depth interviews with representative actors of the network, carried out by
the authors between October 2006 and November 2007, as part of an ongoing project.
The information obtained in the interviews has been cross-checked and contrasted
to direct field observation. These primary data sources have been complemented
with information from other sources (policy statements, strategic programmes,
and quantitative data available) and observations from workshops, site-visits and
conference/meetings with network participants and coordinators. The data from
interviews has been coded and analysed, following the methodological guideline of
Grounded Theory (Glaser and Strauss, 1967) and discourse analysis (Foucault, 1972;
Hammersley, 2003), and supported by the qualitative software ATLAS.ti. The process of
analysis and theory building is based on Social Network Perspective and Discourse
Theory (Dryzeck, 1997). A more detailed account of the data sources and methods used
is summarised in Table 3.
Table 3 Data sources and methods of research
Phase Method Data sources Actors involved
Mapping of actors Social network analysis
(Wasserman and Faust,
1994; Scott, 2000)
Literature review
Expert interviews
Snow-balling
Informal and
expert meetings
Ethnographic
fieldwork
Entrepreneurial
organisations
Networks
coordinators
Institutional actors
Researchers
Definition of
core-periphery structure Social network analysis
(Wasserman and Faust,
1994), core-periphery
structure (Laumann and
Pappi, 1976; Laumann
et al., 1983)
Structured interviews
Questionnaires
Expert interviews
Ethnographic
fieldwork
Entrepreneurial
organisations
Networks
coordinators
Companies
Identification of the
structural characteristics
of the network
Social network analysis
(Tichy et al., 1979) In-depth interviews
Expert interviews
Questionnaires
Networks
coordinators
Companies
Analysis of exchange
conditions and structural
embeddedness
General Network Theory
(Jones et al., 1997)
Discourse analysis
(Dryzeck, 2000)
In-depth interviews
Expert interviews
Networks
coordinators
Companies
Institutional actors
Researchers
5.2 Step 1 – Mapping of actors and establishing the boundaries of the network
A crucial question in analysing IS networks is to define the boundaries of the network
by tracing a border between the companies and organisations belonging to the network
and those outside it. The mapping of the (direct and indirect) actors (Binder, 2007b)
that constitute the network is thus the prerequisite or pre-phase that allows the drawing

16 T. Doménech and M. Davies
of the boundaries of the network. Based on the literature review, a first map is designed
that is then complemented by the input data from: (a) expert interviews and (b) research
and expert meetings. Through ‘snow-balling’ (Clarke, 2002), other interacting agents
are identified. Although an agent map is drawn at the early stages of the process,
the number of actors and boundaries are modified as new actors are incorporated and new
linkages found.
5.3 Step 2 – Identification of the ‘core-periphery’ structure
The definition of the boundaries is not an unproblematic issue, as in many cases, the
boundaries of the network remain blurred and fuzzy, even for the actors involved. The
core-periphery model (Laumann and Pappi, 1976; Alba and Moore, 1978) supports this
process by distinguishing between core-periphery structures within a network. According
to this model, networks would have a core group formed by “densely connected actors in
contrast to a more loosely connected class of actors forming the periphery of the system”
(Garcia Muniz and Ramos Carvajal, 2006). Yet, this scheme still poses questions as
to what constitutes the core and the periphery of the network (Garcia Muniz and Ramos
Carvajal, 2006). This approach is similar to that proposed by Uzzi (1996; 1997),
which differentiates between ‘arm’s length’ and ‘embedded’ ties. The first one would
be characterised by nonrecurrent, nonreciprocal relationships governed by price, while
the latter would refer to ‘special relationships’ characterised by “trust, fine-grained
information transfer, and joint problem-solving arrangements” (Uzzi, 1997).
The definition of the boundaries is also an issue unsolved in the literature of IS
(Korhonen et al., 1994). The novelty of IS networks as objects of study and the
lack, in many cases, of formalised structures increase the difficulty in establishing
these boundaries. Indeed, the criteria to define the boundaries of the network and the
core-periphery structure have to be analysed on a case-by-case basis, as the same
structure of the network might alter the definition of the participants and the rules for
entering and exiting the network. Notwithstanding this, some main guidelines can be
given to identify the core-periphery structure in an IS network (see Figure 1). The core of
the network will be formed by organisations within which:
• there is very frequent informal and formal communication
• information flows and knowledge transfer is a common practice and has led to high
levels of interconnection, including tacit rules of knowledge transfer and
communication protocols
• companies and companies’ representatives have direct access to each other and can
influence each others’ business practices
• material and energy flows occur on a continual basis
• companies are linked by more than one flow or have undertaken more than one
project together
• benefits of cooperation are calculated globally.

The social aspects of IS: the application of social network analysis to IS networks 1
7
Figure 1 Core-periphery structure (see online version for colours)
The periphery of the network, on the other hand, will be made up by those nodes
within which:
• information and knowledge transfer only occur in certain periods of time or linked to
a concrete exchange or transaction
• communication does not occur on a regular basis and tacit rules have not
been established
• there is little scope for companies to influence each others’ business strategies
• material and energy flows occur only at certain periods and not necessarily as
long-term agreements
• joint problem solving is not a common practice and only occurs under specific
circumstances or under particular projects
• benefits of cooperation are calculated as project-by-project basis.
Expert interviews, questionnaires and in-depth interviews or, more generally, a
combination of the three, can be used to trace the ‘boundary’ between the core and
the periphery of the network as well as the analysis of the data on the volume, content
and frequency of exchange, if available. However, it is important to note, that core
participants are commonly easier to identify, while peripheral actors, which may also
play an important role (as local bridges for ideas and innovations), are, in some cases,
more difficult to locate (Binder, 2007b). Even the actors themselves might not be aware
of the important peripheral actors in the process. Expert interviews can be of help here.
Input from other phases, especially when the structure of the linkages is analysed, can
also report significant information on this.
In the case of Kalundborg the identification of the core of the network is quite a
straightforward process, due to a conjunction of factors:

18 T. Doménech and M. Davies
• small size of the network
• existence of an abundant literature about the case
• its spatial concentration.
However, the identification of the indirect actors, situated in the periphery, is not without
difficulty. Indeed most of the analyses in the literature overlook the role that peripheral
actors might play in the emergence and development of the network. We refer here,
basically, to institutional actors that, even though they might not be directly involved
in the operation of the network, have provided a favourable and flexible interpretation of
the regulatory framework for the development of the IS exchanges. It is also to be noted
that even though the physical exchanges have taken place among a limited number of
actors, the informational and knowledge flows have extended beyond the boundaries of
the material transactions.
In the case of NISP, the identification of the boundaries and, especially the
core-periphery structure, of the network is more problematic. The fact that it is organised
as a national project, made up of different regional programmes, and the large size of
the member organisations, with more than 8000 industrial member organisations, adds
complexity to the analysis. Along with this, as the project is in the first phases of its
development, many projects or exchanges proposed still need to be technically tested or
researched or its economic feasibility studied. In this model of network central and
regional coordination play an essential role in:
• providing a web-based system to facilitate informational flows
• interpreting and processing large amount of information about potential ‘matches’ or
possibilities of exchange
• offering knowledge and technological support by the coordination with the
‘Knowledge Transfer Network’, that act as technological partner of the NISP
programme, and specific collaboration agreements with research institutions
and universities.
As expected, in this case, most of the transactions take place through the coordination
actors. However, not all the exchanges take place through these central nodes, as linkages
can be found among peripheral actors, without the intervention of the coordination
bodies. NISP workshops act, in this sense, as an excellent platform for the connection of
actors. Another aspect that might add complexity to drawing the boundaries in a network
like NISP is the fact that some of the linkages might consist of a ‘one-off’ exchange,
closer to the conception of ‘arm’s length ties’ rather than to ‘network governance’
cooperation (Uzzi, 1997).
An issue that may arise in the process of identification of the boundaries of the
network is the debatable existence of an ‘optimal network size’. The larger a network
grows the more opportunities and synergies it can offer to its members, having a positive
impact on the adaptability and availability of information and resources. However,
as it grows bigger the costs of coordination and integration rapidly rise, reducing the
benefits of network governance and its ability to define a common goal and culture,
reducing the chances to develop social mechanisms of control and thus lowering the level
of structural embeddedness. Small and locally concentrated networks such as Kalundborg
facilitate the development of social mechanisms of control such as collective sanctions,

The social aspects of IS: the application of social network analysis to IS networks 1
9
reputation (trust) and a common macroculture. Moreover, transaction costs are
significantly diminished due to the reduced number of actors and spatial concentration.
However, the small size of the network can also have a negative impact on the number
and variety of potential synergies and could create problems of ‘lock-in’. In contrast,
large networks such as NISP, although having more potential synergies, have to deal with
the following negative aspects:
• greater transaction costs
• slower diffusion of information
• risk of interruption of flows, due to longer path distances between two nodes and the
existence of weak ties.
Therefore networks might face the tradeoff between size and cohesion. Empirical data
and case studies can shed some light when trying to define the optimal size of IS
networks (Uzzi, 1997).
5.4 Step 3 – Identification of the structural characteristics of the network;
typologies of networks
Once the actors have been identified and a tentative core-periphery structure has
been defined, the next step in the proposed methodology consists of the examination
of the structural characteristics of the network and the nature of the linkages. Following
the scheme provided by Tichy et al. (1979), networks can be analysed according to
three main areas:
1 Transactional Content
2 the nature of the links
3 its structural characteristics (see Figure 2).
The first area refers to what is exchanged or shared within the nodes of a network.
Tichy et al. (1979) identify four types of transactional content:
1 exchange of emotional support
2 exchange of influence or power
3 exchange of information
4 exchange of goods and services.
In the case of IS, networks will generally involve transactions of Types 2, 3 and 4.
Social networks can be redrawn for each of these content types, so that the same
network can be structured in distinctive ways depending on the content that is being
analysed. Therefore, an industrial network could have a structure when informational
and knowledge content is examined, and a rather different one when material flows
are observed. Different levels of analysis would then be needed to analyse complex
IS networks.

20 T. Doménech and M. Davies
Figure 2 Areas of analysis of IS networks (see online version for colours)
Secondly, the nature of the links occurring between the nodes that form the network
requires the study of different dimensions:
• the intensity of the relationship, which can be a measure of the frequency of contact
in a unit of time or a proxy of the content-value of the exchange
• the degree of reciprocity of the relationships; this dimension permits to distinguish
between symmetric (nondirective) relationships and asymmetric relationships
• degree of formality of the relationships, which can vary from informal to formally
institutionalised relationships
• the multiplexity (Tichy et al., 1979) or the diversity of roles that a pair of nodes
can represent.
Finally, the study of the structural characteristics of the network should be also
undertaken at different levels: (1) the external level, which focuses on the definition of
the boundaries of the network and the interaction of the network with other networks,
isolated nodes or the environment; At this level, local bridges can be identified.
The concept of ‘bridge’ is that of a ‘line in a network which provides the only path
between two points’. Bridges act as the ‘gate’ for the connection with other groups or
networks, and in most cases it can actually be the ‘only’ way to possibly connect two
nodes. Thus, even though networks characterised by strong and dense ties may be more
cohesive, in absence of weak ties acting as local bridges, this system could lead to overall
fragmentation (Granovetter, 1973) and (2) the internal network, which basically captures

The social aspects of IS: the application of social network analysis to IS networks 21
the patterns of the relationships and structure of the network in itself. Core-periphery
structure can be analysed at this point together with the patterns of communication and
exchange. At this level, the structure of the network can be characterised as the result
of the combination of different properties, such as: density or connectedness, openness,
stability and centrality.
IS initiatives vary greatly depending on the conditions on which they are founded:
from core-centred networks such as Kalundborg to more loose-coupled exchange
structures such as NISP in the UK. Thus, different typologies of IS networks can be
derived from this analysis. Table 4 offers, for illustrative purposes, an example of a
comparative analysis of the IS network in Kalundborg and NISP.
Table 4 Comparison of structural characteristics of Kalundborg/NISP IS networks
Areas of analysis Conditions Characteristics Kalundborg NISP
Structural
characteristics
External Size Small-size network (six major
companies + municipality)
Extent network (8000 company
members)
Internal Core-periphery High degree of coreness
Unstructured periphery*
Coordination nodes – high
degree of coreness
Unstructured periphery*
Density High Low
Openness/
closeness
Close network Open network
Centrality
Centred structure, characterised
by a well identified core of
central actors
Coordination nodes are
central actors
Stability
Stability in the composition of
the network, although there have
been changes in the linkages
among actors
High-speed growing network
Less stable in structural terms
Transactional
content
Exchange of
information and
knowledge
Mostly based on informal
contacts
Partly institutionalised through
coordination nodes and
technological partner
Exchange of
influence and
power
Tacit Not identified
Material and
energy exchanges
Formalised in commercial
agreements
Formalised in commercial
agreements
Nature of
linkages
Intensity Mostly based on informal
contacts
Partly institutionalised through
coordination nodes and
technological partner
Frequency Frequent/Very frequent
Varies according to the
characteristic of the dyad, level
of complexity, duration, etc.
Reciprocity
High and not linked to specific
dyads (general reciprocity within
network actors)
It is linked to specific dyads
and varies according to the
characteristic of the dyad, level
of complexity, duration, etc.
Note: * Nodes in the periphery generally depend from the core and do not have dyads
with other nodes located in the periphery.

22 T. Doménech and M. Davies
5.5 Step 4 – Identification of exchange conditions and structural embeddedness
A next step would be to identify, following the theoretical framework of Jones et al.
(1997), the conditions that have favoured the emergence of the network. It would be
necessary to examine both the exchange conditions as well as the social mechanisms in
place. Exchange conditions that are likely to be found in IS networks are:
• stringent and rapidly evolving regulatory frameworks
• shortage of raw materials
• the need for customised, ad hoc solutions1
• frequency of contact.
These conditions increase the need for coordination and flexibility, favouring the
emergence of network-related governance forms. At the same time, social mechanisms
develop to support the operation of the network. Among the social control mechanisms,
the existence of a macroculture and tacit collective sanctions foster collaborative
behaviour. An analysis of the culture of cooperation and the tacit rules that govern
the network is needed for a better understanding of its dynamics, potentials and threats
(see Figure 3).
Figure 3 Exchange conditions and social mechanisms of control (see online version for colours)
In the case of Kalundborg (Table 5) the exchange conditions have been defined by
a dense concentration of companies with a high environmental impact in the area,
the scarcity of a basic input, such as fresh water (Ehrenfeld and Gertler, 1997), and the
existence of a growing stringent environmental regulatory framework, defining the basis
for the development of structural embeddedness, which have given rise to the progressive
building of trust and general reciprocity. Social mechanisms of control developed along
the process of creation of the network, favoured by the social connection of the managers
of the main companies located in the area and the existence of a culture of collaboration,
have further allowed the development of embedded links. Frequent interactions have
led to a progressive unification of discourses between actors, creating a coherent
macroculture, which has contributed to the progressive introduction of environmental

The social aspects of IS: the application of social network analysis to IS networks 2
3
issues and collaboration at the core of the business strategy, defining the implicit “license
to operate” (quoted from interviews). “Joint problem solving, exchange of fine-grained
information and building of trust” (Uzzi, 1996) have contributed in this case to significant
competitive advantages (Jacobsen, 2006).
Table 5 Summary of exchange conditions and social mechanisms of control in Kalundborg
and NISP
Conditions General IS
network conditions Kalundborg NISP
Stringent and
rapidly evolving
regulatory
frameworks
• Stringent regulatory
framework combined
to a more flexible
approach to regulation
• Application of
Integrated Pollution
Prevention and
Control (IPPC)
• Stringent regulatory
framework on
environmental issues as a
consequence of the new role
of UK as ‘pace-setter’ in the
environmental policy of EU
• Increase in the landfill taxes
• Climate Change Levy
Shortage of
raw materials Fresh water scarcity Landfill space scarcity
The need for
customised,
ad hoc solutions
Very important due to
peculiarities of the
‘industry-mix’ and
competitive context
Combination of standard and
customised solutions
Exchange
conditions
Frequent
interaction
Face-to-face interaction.
Actors linked not only
by professional but
also by emotional
(friendship) links
• Limited interaction due to
large size of the network
• Opportunity to develop
cliques, or subgroups of
actors, with more frequent
interaction
Collective
sanctions Very important Underdeveloped due to
large-size network, although it
may develop in cliques
Reputation Very important Underdeveloped due to
large-size network, although it
may develop in cliques
Social
mechanisms
Macroculture of
cooperation Very important Potential to be developed
throughout the network,
though restricted by the
underdevelopment of the other
two mechanisms: collective
sanctions and reputations.
Moreover, uniformity of
discourses difficult to achieve
In the case of NISP, and due to the policy-driven nature of the network, an examination
of the exchange conditions in place may provide an indication of the chances of
the initiative to succeed in the medium/long term as well as the possibility of social
mechanisms of control to develop in the case of nonspontaneous networks. Some
exchange conditions seem to have favoured in this case the emergence and success of the

24 T. Doménech and M. Davies
programme. In the current context, companies in the UK are faced with two major
challenges in the environmental area. On the one hand, there is growingly stringent
environmental regulation in accordance with the current UK position as an ‘active
pace-setter’ in the context of EU environmental policy (Börzel, 2002). On the other hand,
the scarcity of landfill, has led to the steady increase of landfill taxes, fostering the
attractiveness of IS alternatives. However, as a result of the size of the network and
the character of the initiative, social mechanisms of control as collective sanctions and
macroculture might not be developed, reducing the chances of adopting innovative
collective solutions or transferring ‘fine-grained’ information, which imply higher levels
of risk and trust, and which would therefore require high coordination and safeguarding.
In fact, this can be a hindering aspect for the further development of NISP.
6 Discussion: methodological issues and potential shortcomings
of the approach
The application of social network analysis to the field of IS offers a new perspective
for the examination of the structural characteristics of these networks and the critical
social aspects that might foster or hinder their emergence and development. Here a
methodological protocol is proposed for the evaluation of the structural characteristics,
exchange conditions and social control mechanisms of IS networks, as a way to
understand the processes that are on the basis of the network operation. However, the
application of the method proposed is not free from difficulties. Some methodological
issues that have arisen during the process of implementation have been the following:
• One important barrier in the undertaking of an analysis based on Network
perspective is the absence of data on the qualitative aspects of IS. The fact that IS
develops in the edges of the policy agenda accounts for the lack of systematic,
‘hard’, contrasted data on the outcomes of IS networks and, specially, of the social
aspects influencing IS. ‘The creation of data’ is thus a necessary step for the
understanding of the social aspects surrounding IS development. As already noted,
the ‘data’ has been generated mainly through in-depth interviewing, as a way to
understand the processes, structures and discourses that lie behind the concrete
outcomes of IS and the subtle aspects influencing decision making such as
organisational culture, tacit rules and social context. The data available (literature,
company reports, institutional documents, etc.) have been contrasted and
complemented with the information obtained in the interviews, contributing to
support, clarify or contrast some of the statements inferred from the interviews.
• Another problem is that generated by the possible ‘bias’ in sampling procedures
and interviewing process. A qualitative research design like the one proposed here
differs greatly from a method based on the quantitative perspectives. We recognise
here a double bias, common to most qualitative research designs (Rapley, 2004):
(a) the ‘bias’ in the selection of the case studies, not based on ‘random’ selection but
on ‘information-oriented strategy’ (Flyvbjerg, 2004) and (b) the possible bias in the
selection of the actors to be interviewed and the ‘data’ derived from the interviews.
Taking into account the limitations derived from these potential biases, and given
the centrality of interviews in the methodology proposed to approach, the selection

The social aspects of IS: the application of social network analysis to IS networks 2
5
and recruitment of interviewees has been considered very carefully.
The interviewees’ recruitment process has to be guided by four key principles
(Rubin and Rubin, 2005):
1 getting a wide range of actors that holds different perspectives of the process
under analysis
2 within the organisations involved, find knowledgeable and
representative informants
3 testing emerging themes with new interviews
4 find new interviewees to extend and contrast the results (‘snowball or
chain’ sampling).
According to these principles, the strategy on recruitment relies largely on the
mapping of actors for each of the case studies. As for the process of in-depth
interviewing, it has been closer to the approach of ‘interview-data-as-topic’ rather
than ‘interview-data-as-resource’ (Seale, 1998). The ‘interview-data-as-topic’ approach
challenges the concept of ‘interview data’ as an accurate reflection of the reality outside
and independent from the interview. On the contrary, interviewer and interviewee are
seen as engaged in a process where they are given and represent different roles and
positions, collaboratively constructing and defining the topic and the course of the
interview. The interview is then built as an interaction event, which is ‘locally’ and
dialectically constructed (Rapley, 2004). This does not mean, however, that interviews
are intentionally led or biased by the vision of the interviewer but that the researcher
takes into account the cognitive limitations of the interviews and their role as constructed
dialogues. The interviewer’s strategy has combined neutral and collaborative approaches
depending on the necessity to build rapport between interviewer and interviewee and
allowing the contrast of the thesis, hypothesis and ideas with the interviewee, so that
agreed meanings could be created and discussed along the interview. This strategy has
been proven helpful in the course of the research to help to uncover the ‘hidden voices’,
truth and stories and to obtain a more reflexive understanding of the situation, perceptions
and practices behind a more standardised account of the facts.
• Divergences in the discourse and the concepts of ‘cooperation’ and belonging to
the network. During the process of implementation of the methodology, different
conceptions of cooperation and IS perspective have arisen. While in the case of
Kalundborg, interviews with actors showed a more elaborated discourse, based
largely on a ‘common narrative’ that accounts for the outcomes and causes of IS
cooperation and a clearly delimitation of the boundaries of the network, together
with a strong sense of ‘belonging’, in the case of NISP, the interviews seem to point
to a more unstructured discourse and a diverse understanding of the concept of
cooperation and belonging to the network. This may pose some challenges in the
process of interpretation and analysis of the data. As has been noted before, it has
been useful to understand these divergences of meaning to explore and agree
common understandings between interviewer and interviewee.
Beyond the more concrete methodological issues generated by the application of the
methodology, some caveats and shortcomings of the approach should be accounted for in
further applications and for method refinement. Here, the method has been presented as a

26 T. Doménech and M. Davies
‘stand alone’ perspective for analysing IS initiatives. However, its policy relevance is
subjected to its combination with other analytical tools, such as Material Flow Analysis
(Ayres and Ayres, 2002) or Life Cycle Assessment (Anastas and Lankey, 2000) that
provide the ‘material’ conditions for IS. Another caveat of the approach derives from
the difficulty to capture the ever-changing structural conditions of IS networks and the
changes in the composition, leadership, position and power of the actors involved.
Dynamic analysis and scenario building shall be considered as complementary strategies
to deal with this in the refinement of method. Finally, it is important to account for the
limitation of social network analysis on its own to provide prescriptive guidance for
is development. Social network analysis is a method of analysis but a prescriptive
dimension can only be attained when complemented with other theories. Here we have
opted for the General Network Theory as it offers a suitable framework for the analysis
of the social conditions influencing the operation of networks, but there is also potential
for its combination with discourse theory (see, for example, Dryzeck, 1997) or the
institutional theory (see, for example, Zucker, 1987). It has to be noted however, that
these issues do not diminish the ability of the approach to offer useful insights into the
process of the examination of the social aspects of IS. The aim of the analysis is to build
on existing methods and provide a comprehensive framework rather than to deny or
substitute previous analysis and tools.
7 Conclusions and research agenda
Institutional factors and social and cultural aspects are shaping the decision-making
process of companies participating in IS networks. Aspects such as the generation of
trust, the perception of risk and the cooperative culture, impact on the potential of IS
networks to develop. Therefore, understanding the dynamics of IS networks necessarily
requires an understanding of the social factors behind the material and energy exchanges
and conditions under which companies are more likely to cooperate. With this, we are not
dismissing the ‘material’ side of IS. Indeed we believe that there are certain critical
‘material’ conditions for the development of IS but our focus is on the provision of a
framework for the identification of the social and institutional factors that might foster
the emergence of an IS network. The aim of this paper has been to briefly introduce
Social network analysis in the field of IS, from a methodological point of view, offering
an overview of the opportunities and the potential it offers for exploring the ‘soft’ aspects
of IS networks and, more concretely, the structural characteristics of IS networks. As has
been suggested, the main contribution of network theory to the analysis of the field of IS
lies in the provision of:
• a comprehensive framework to analyse the structural features of IS networks, as a
complement to other approaches
• prescriptive outcomes on the basis of the exploration of the contextual and social
factors that might contribute to emergence of the network form of governance.
This is of crucial importance for a field with a lack of prescriptive frameworks to
orientate policy action. Moreover, this perspective helps to shed some light on the
process of decision making of companies when evaluating their integration into IS
networks, overcoming the understanding provided by the consideration of industrial units
as rational actors in a narrow economic sense.

The social aspects of IS: the application of social network analysis to IS networks 2
7
Further empirical research however, is needed to progress to a better definition of
the exchange conditions and social control mechanisms that may foster the emergence of
IS networks and the structural dynamics of successful IS networks. Aspects such as the
optimal size of the network or the optimal structure need to be addressed as well as the
policy measures that can speed up the process for social mechanisms of control to
develop, providing the ground for exchange flows to happen. A contestable issue is the
ability of policy mechanisms to create a macroculture of cooperation and the role
played by public actors in anchoring IS projects. All these questions need to be addressed
in depth in light of the findings of empirical research based on social network analysis.
Acknowledgements
The authors would like to thank two anonymous reviewers and Professor Yvonne Rydin,
whose comments helped to improve the manuscript significantly.
References
Alba, R. and Moore, G. (1978) ‘Elite social circles’, Sociological Methods and Research, Vol. 7,
No. 2, pp.167–188.
Anastas, P. and Lankey, R. (2000) ‘Life cycle assessment and green chemistry: the yin and yang of
industrial ecology’, Green Chemistry, Vol. 2, pp.289–295.
Ayres, R. and Ayres, L. (Eds.) (2002) A Handbook of Industrial Ecology, Cheltenham, UK:
Edward Edger Publishing.
Ayres, R.U. and Ayres, L.W. (1998) Accounting for Resources: 1. Economy-wide Applications of
Mass-Balance Principles to Materials and Waste, Cheltenham, UK: Edwar Elgar.
Ayres, R.U. and Ayres, L.W. (1999) Accounting for Resources: 2. The Life Cycle of Materials,
Cheltenham, UK: Edwar Elgar.
Baker, W. (1987) ‘What is money?’, in M.S. Mizruchi and M. Schwartz (Eds.) Intercorporate
Relations, Cambridge: Cambridge University Press.
Begré, M.S. and Hadorn, H.G. (2002) ‘Options & restrictions: a heuristic tool in transdiciplinary
research for an effective implementation of sustainable practices’, in F. Biermann, S. Campe
and K. Jacob (Eds.) Proceedings of the 2002 Berlin Conference on the Human Dimensions of
Global Environmental Change ‘Knowledge for the Sustainability Transition. The Challenge
for Social Science’, Global Governance Project, Amsterdam/Berlin/Potsdam (Germany)/
Oldenburg (Germany), pp.322–336.
Binder, C.R. (2007a) ‘From material flow analysis to material flow management, Part I:
social science approaches coupled to material flow analysis’, Special issue, Journal of Cleaner
Production, Vol. 15, pp.1596–1604.
Binder, C.R. (2007b) From material flow analysis to material flow management, Part II:
the role of structural agent analysis’, Special issue, Journal of Cleaner Production, Vol. 15,
pp.1605–1617.
Binder, C.R., Hofer, C., Wiek, A. and Scholz, R.W. (2004) ‘Transition towards improved regional
wood flow by integrating material flux analysis with agent analysis: the case of Appenzell
Ausserrhoden, Switzerland’, Ecological Economics, Vol. 49, No. 1, pp.1–17.
Börzel, T. (2002) ‘Pace-setting, foot-dragging, and fence-sitting: member state responses to
Europeanization’, Journal of Common Market Studies, Vol. 40, No. 2, pp.193–214.
Burnström, F. and Korhonen, J. (2001) ‘Municipalities and industrial ecology: reconsidering
municipal environmental management’, Sustainable Development, Vol. 9, No. 1, pp.36–46.
Please check name
of publisher.

28 T. Doménech and M. Davies
Chertow, M. and Lombardi, R. (2005) ‘Quantifying economic and environmental benefits of
co-located firms’, Environmental Science and Technology, Vol. 39, No. 17, pp.6535–6541.
Clarke, S. (2002) ‘Learning from experience: psycho-social research methods in the social
sciences’, Qualitative Research, Vol. 2, No. 2, pp.173–194.
Cook, K. and Whitmeyer, J. (1992) ‘Two approaches to social structure: exchange theory and
network analysis’, Annu. Rev. Sociol., Vol. 18, pp.109–127.
Desrochers, P. (2002) ‘Industrial ecology and the rediscovery of inter-firm recycling linkages:
historical evidence and policy implications’, Industrial and Corporate Change, Vol. 11, No. 5,
pp.1031–1051.
Doménech, T. and Davies, M. (2006) ‘Modelling industrial ecology: integrated modelling of
physical, economic and social system’, in J. Pollard (Ed.) European Modelling Symposium
Proceedings, pp.32–36.
Dryzeck, J. (1997) The Politics of the Earth: Environmental Discourses, Oxford: Oxford
University Press.
Dryzeck, J. (2000) Deliberative Democracy and Beyond: Liberals, Critics and Contestations,
Oxford: Oxford University Press.
Ehrenfeld, J. and Gertler, N. (1997) ‘Industrial ecology in practice: the evolution of
interdependence at Kalundborg’, Journal of Industrial Ecology, Vol. 1, No. 1, pp.67–79.
Eilering, J.A.M. and Vermeulen, W.J.V. (2004) ‘Eco-industrial parks: towards industrial symbiosis
and utility sharing in practice’, Progress in Industrial Ecology, Vol. 1, Nos. 1–3, pp.245–270.
Flyvbjerg, B. (2004) ‘Five misunderstandings about case-study research’, in C. Seale, et al. (Eds.)
Qualitative Research Practice, London: Sage.
Foucault, M. (1972) The Archaeology of Knowledge, London: Routledge.
Frosch, R. and Gallopoulos, N. (1989) ‘Strategies for manufacturing’, Scientific American,
Vol. 261, No. 3, pp.144–152.
Garcia Muniz, A. and Ramos Carvajal, C. (2006) ‘Core/periphery structure models: an alternative
methodological proposal’, Social Networks, Vol. 28, pp.442–448.
Glaser, B. and Strauss, A. (1967) The Discovery of Grounded Theory: Strategies for Qualitative
Research, New York: Aldine de Gruyter.
Goss, J., et al. (2005) ‘The eco-park: green nirvana or white elephant?’, International Sustainable
Development Research Conference, Finlandia Hall, Helsinki, 6–8 June.
Granovetter, M. (1973) ‘The strength of weak ties’, American Journal of Sociology, Vol. 78, No. 6,
pp.1360–1380.
Hadorn, H.G., et al. (2002) ‘Restrictions and options: a heuristic tool to integrate knowledge
for strategies towards a sustainable development’, International Journal of Sustainable
Development and World Ecology, Vol. 9, pp.193–207.
Haines, V. (1988) ‘Social network analysis, structuration theory and the holism-individualism
debate’, Social Networks, Vol. 10, pp.157–182.
Hammersley, M. (2003) ‘Conversation discourse analysis: methods or paradigms?’, Discourse and
Society, Vol. 14, pp.751–781.
Haskins, C. (2006) ‘Multidisciplinary investigations of eco-industrial parks’, Systems Engineering,
Vol. 9, No. 4, pp.313–330.
Hermanns, L.M. (2005) ‘Dynamic actor network analysis for diffuse pollution in the province of
North-Holland’, Water Science and Technology, Vol. 49, No. 2, pp.205–212.
Hoang, H. and Antoncic, B. (2003) ‘Network-based research in entrepreneurship: a critical review’,
Journal of Business Venturing, Vol. 18, No. 2, pp.165–187.
Huberman, B. and Hogg, T. (1995) ‘Communities of practice: performance and evolution’,
Computational and Mathematical Organization Theory, Vol. 1, No. 1, pp.73–92.

The social aspects of IS: the application of social network analysis to IS networks 2
9
Jacobsen, N.B. (2006) ‘Industrial symbiosis in Kalundborg, Denmark: a quantitative assessment
of economic and environmental aspects’, Journal of Industrial Ecology, Vol. 10, Nos. 1–2,
pp.239–255.
Jones, C., et al. (1997) ‘A general theory of network governance: exchange conditions and social
mechanisms’, The Academy of Management Review, Vol. 22, No. 4, pp.911–945.
Korhonen, J. (2004b) ‘Theory of industrial ecology’, Progress in Industrial Ecology, Vol. 1,
Nos. 1–3, pp.61–68.
Korhonen, J., et al. (2001) ‘An anchor tenant approach to network management – considering
regional material and energy flow networks’, Int. J. Environmental Technology and
Management, Vol. 1, No. 4, pp.444–463.
Korhonen, J., et al. (2004) ‘Management and policy aspects of industrial ecology: an emerging
research agenda’, Business Strategy and the Environment, Vol. 13, pp.289–305.
Kytzia, S., et al. (2004) ‘Economically extended-MFA: a material flow approach for a
better understanding of food production chain’, Journal of Cleaner Production, Vol. 12,
pp.877–889.
Larson, A. (1992) ‘Network dyads in entrepreneurial settings: a study of the governance of
exchange relationships’, Administrative Science Quarterly, Vol. 37, No. 1, pp.76–104.
Laumann, E., et al. (1977) ‘Community-elite influence structures: extension of a network
approach’, American Journal of Sociology, Vol. 83, pp.594–631.
Laumann, E., et al. (1983) ‘The boundary specification problem in network analysis’, in R. Burt
and M. Minor (Eds.) Applied Network Analysis, London: Sage.
Laumann, E., Galaskiewicz, J. and Marsden, P. (1978) ‘Community structures as
interorganizational linkages’, in R. Turner, J. Coleman and R. Fox (Eds.) Annual Review of
Sociology, Vol. 4, Palo Alto: Annual Reviews.
Laumann, E. and Pappi, F. (1973) ‘New directions in the study of elites’, American Sociological
Review, Vol. 38, pp.212–230.
Laumann, E. and Pappi, F. (1976) Networks of Collective Action: A Perspective on Community
Influence Systems, New York: Academic Press.
Lewis, D. and Weigert, A. (1985) ‘Trust as social reality’, Social Forces, Vol. 63, No. 4,
pp.967–985.
Lowe, E. and Evans, L. (1995) ‘Industrial ecology and industrial ecosystems’, Journal of Cleaner
Production, Vol. 3, Nos. 1–2, pp.47–53.
McPherson, M., et al. (2001) ‘Birds of a feather: homophily in social networks’, Ann. Rev. Sociol.,
Vol. 21, pp.414–444.
Miles, M. and Huberman, A. (1994) Qualitative Data Analysis, 2nd ed., Thousand Oaks: Sage.
Mirata, M. and Emtairah, T. (2005) ‘Industrial symbiosis networks and the contribution to
environmental innovation: the case of Landskrona industrial symbiosis programme’, Journal
of Cleaner Production, Vol. 13, pp.993–1002.
Mol, A.P.J. and Sonnenfeld, D. (2000) ‘Ecological modernization around the world: an
introduction’, Environmental Politics, Vol. 9, No. 1, pp.3–14.
Ouimet, M., et al. (2004) ‘Network positions and radical innovation: a social network analysis of
the Quebec optics and photonics cluster’, Paper presented at the DRUID Summer Conference
2004 on Industrial Dynamics, Innovation and Development, Elsinore (Helsingør), Denmark,
14–16 June.
Powell, W., et al. (1996) ‘Technological change and the locus of innovation: networks of learning
in biotechnology’, Administrative Science Quarterly, Vol. 41, pp.116–145.
Rapley, T. (2004) ‘Interviews’ in C. Seale, et al. (Eds.) Qualitative Research Practice,
London: Sage.
Rubin, H. and Rubin, I. (2005) Qualitative Interviewing: The Art of Hearing Data, 2nd ed.,
Thousand Oaks, London: Sage.

30 T. Doménech and M. Davies
Schensul, J., et al. (1999) Mapping Social Networks, Spatial Data and Hidden Populations,
Walnut Creek, California.
Scott, J. (2000) Social Network Analysis: A Handbook, UK: Sage Publications.
Seale, C. (1998) ‘Qualitative interviewing’, in C. Seale (Ed.) Researching Society and Culture,
London: Sage.
Spencer, J.W. (2003) ‘Global gatekeeping, representation, and network structure: a longitudinal
analysis of regional and global knowledge-diffusion networks’, Journal of International
Business Studies, Vol. 34, pp.428–442.
Sterr, T. and Ott, T. (2004) ‘The industrial region as a promising unit for eco-industrial
development – reflections, practical experience and establishment of innovation instruments to
support industrial ecology’, Journal of Cleaner Production, Vol. 12, Nos. 8–10, pp.947–965.
Tichy, N., et al. (1979) ‘Social network analysis for organisations’, The Academy of Management
Review, Vol. 4, No. 4, pp.507–519.
Tutzauer, F., et al. (2006) ‘Network structure, strategy evolution and the game of chicken’, Social
Networks, Vol. 28, No. 4, pp.377–396.
Uzzi, B. (1996) ‘The sources and consequences of embeddedness for the economic
performance of organizations: the network effect’, American Sociological Review, Vol. 61,
No. 4, pp.674–698.
Uzzi, B. (1997) ‘Social structure and competition in interfirm networks: the paradox of
embeddedness’, Administrative Science Quarterly, Vol. 42, No. 1, pp.35–67.
Valente, T. (1995) Network Models of Diffusion of Innovations, Cresskill, NJ: Hampton Press.
Walker, J., et al. (1994) ‘Interfirm cooperation and startup innovation in the biotechnology
industry’, Strategic Management Journal, Vol. 15, pp.387–394.
Wasserman, S. and Faust, K. (1994) Social Network Analysis: Methods and Applications,
Cambridge, UK: Cambridge University Press.
Yang, S. and Knoke, D. (2001) ‘Optimal connections: strength and distance in valued graphs’,
Social Networks, Vol. 23, pp.285–295.
Yin, R. (1989) Case Study Research, Revised edition, Beverly Hills: Sage.
Zucker, L. (1987) ‘Institutional theories of organisation’, Ann. Rev. Sociol., Vol. 13, pp.443–464.
Bibliography
Aghion, P. and Howitt, P. (1998) Endogenous Growth Theory, Cambridge, Massachusetts/London:
MIT Press.
Albino, V., et al. (2003) ‘Analysing material and energy flows in an industrial district using
enterprise input-output model’, Economic Systems Research, Vol. 15, No. 4, pp.457–480.
Alverson, M. and Kerreman, D. (2000) ‘Varieties of discourse: on the study of organisations
through discourse analysis’, Human Relations, Vol. 53, pp.1125–1149.
Andersen, M.M. (2002) ‘Organising interfirm learning – as the market begins to turn green’,
in T.J.N.M. de Bruijn and A. Tukker (Eds.) Partnership and Leadership – Building Alliances
for a Sustainable Future, Dordrecht: Kluwer Academic Publishers, pp.103–119.
Axelsson, B. and Easton, G. (1992) Industrial Networks: A New View of Reality,
London: Routledge.
Borgatti, S. and Foster, P. (2003) ‘The network paradigm in organisational research’, Journal of
Management, Vol. 29, No. 6, pp.991–1013.
Chertow, M. (Ed.) (2002) ‘Developing industrial ecosystems: approaches, cases and tools’, Bulletin
Series Yale School of Forestry and Environmental Studies, Series No. 106.
Chertow, M.R. (1999) ‘The eco-industrial park model reconsidered’, Journal of Industrial Ecology,
Vol. 2, No. 3, pp.8–10.

The social aspects of IS: the application of social network analysis to IS networks 31
Christensen, J. (1999) Proceedings of the Industry and Environmental Workshop, Indian Institute
of Management, Ahmedabad, India.
Conway, S., et al. (2001) ‘Realising the potential of the network perspective’, in O. Jones, et al.
(Eds.) Social Interaction and Organisational Change: The Aston Perspectives on Innovation
Networks, London: Imperial College Press.
Gibbs, D. (2003) ‘Trust and networking in inter-firm relations: the case of eco-industrial
development’, Local Economy, Vol. 18, No. 3, pp.222–236.
Gibbs, D., Deutz, P. and Proctor, A. (2002) ‘Sustainability and the local economy: the role of
eco-industrial parks’, Paper presented to the conference Ecosites and Eco-Centres in Europe,
Brussels, 19 June, http://www.hull.ac.uk/geog/research/EcoInd (accessed December 2004).
Gibbs, D.C. (2002) Local Economic Development and the Environment, London: Routledge.
Gouldson, A. and Murphy, J. (1997) ‘Ecological modernisation: restructuring industrial
economies’, in M. Jacobs (Ed.) Greening the Millennium. The New Politics of the
Environment, Oxford: Blackwell.
Hajer, M.A. (1995) The Politics of Environmental Discourse: Ecological Modernisation and the
Policy Process, Oxford: Clarendom.
Harris, S. and Pritchard, C. (2004) ‘Industrial ecology as a learning-process in business strategy’,
Progress in Industrial Ecology, Vol. 1, Nos. 1–3, pp.89–111.
Hasler, A. and Schwarz, E. (2001) ‘Recycling networks in Denmark and Upper Styria’, Discussion
papers, Institute for Innovation and Environmental Management, Karl Franzens University of
Graz, Austria.
Huber, J. (2000) ‘Towards industrial ecology. Sustainable development as a concept of ecological
modernization’, Journal of Environmental Policy and Planning, Vol. 2, No. 4, pp.269–285.
Huber, J. (2004) New Technologies and Environmental Innovation, Cheltenham: Edward Elgar.
Isaksen, A. (2001) ‘Building regional innovation systems: Is endogenous industrial development
possible in the global economy?’, Canadian Journal of Regional Science, Vol. 24, No. 1,
pp.101–120.
Korhonen, J. (2004a) ‘Industrial ecology in the strategic sustainable development model: strategic
applications of industrial ecology’, Journal of Cleaner Production, Vol. 12, pp.809–823.
Markusen, A. (1999) ‘Fuzzy concepts, scanty evidence, policy distance: the case for the rigour and
policy relevance in critical regional studies’, Regional Studies, Vol. 33, No. 9, pp.869–884.
Milchrahm, E. and Hasler, A. (2002) ‘Knowledge transfer in recycling networks: fostering
sustainable development’, Journal of Universal Computer Science, Vol. 8, No. 5, pp.546–556.
Mirata, M., et al. (2005) ‘Production systems aligned with distributed economies: examples
from energy and biomass sectors’, Journal of Cleaner Production, Vol. 13, Nos. 10–11,
pp.981–991.
Mol, A.P.J. (1997) ‘Ecological modernization: industrial transformations and environmental
reform’, in M. Redclift and G. Woodgate (Eds.) The Environmental Handbook of
Environmental Sociology, UK/USA: Edward Elgar.
Mol, A.P.J. and Spaargaren, G. (1993) ‘Environment, modernity and the risk society: the
apocalyptic horizon of environmental reform’, International Sociology, Vol. 8, pp.431–459.
Mol, A.P.J. and Spaarsgaren, G. (2000) ‘Ecological modernization theory in debate: a review’,
Environmental Politics, Vol. 9, No. 1, pp.17–49.
Morgan, K. (2004) ‘The exaggerated death of geography: learning, proximity and territorial
innovation systems’, Journal of Economic Geography, Vol. 4, pp.3–21.
Powell, W.W. and Smith-Doerr, L. (1994) ‘Networks and economic life’, in N. Smelser and
R. Swedberg (Eds.) Handbook of Economic Sociology, Princeton, NJ: Princeton University
Press, pp.368–402.
Rapley, T. (2007) Doing Conversation, Discourse and Document Analysis (Qualitative Research
Kit), London: Sage.

32 T. Doménech and M. Davies
Research Triangle Institute (2001) ‘Eco-industrial parks: a case study and analysis of economic,
environmental, technical and regulatory issues’, Executive summary, http://www.rti.org/units/
ssid/cer/parks.cfm (accessed October 2005).
Schwarz, E. and Steininger, K.W. (1997) ‘Implementing natures’s lesson: the industrial recycling
network enhancing regional development’, Journal of Cleaner Production, Vol. 5, Nos. 1–2,
pp.47–56.
Spiegelman, J. (2000) ‘Industrial ecology. Putting sustainable development to work’,
http://ejournal.nbii.org/archives/vol1iss2/o501-001.newman.pdf (accessed November 2005).
UNEP (1997) ‘The environmental management of industrial estates’, UNEP IE Technical
Report No. 39.
UNEP (2000) ‘The industrial symbiosis in Kalundborg, Denmark’, Training Materials,
http://www.unpeie.org (accessed 1 July 2005).
Valente, T. (1996) ‘Social network threshold in the diffusion of innovations’, Social Networks,
Vol. 18, pp.69–89.
Wallner, H.P. (1999) ‘Towards sustainable development of industry: networking, complexity and
eco-clusters’, Journal of Cleaner Production, Vol. 7, pp.49–58.
Williamson, O. (1991) ‘Comparative economic organisation: the analysis of discrete structural
alternatives’, Administrative Science Quarterly, Vol. 39, pp.269–296.
Web pages
www.unepie.org
www.symbiosis.dk
www.epa.gov
www.greeningofindustry.org
www.cein.ca
www.nisp.org.uk
Note
1 Note that the third condition relates to what Jones et al. (1997) specify as ‘task complexity’
and ‘human asset specifity’. In highly competitive contexts, uncertainty and need of adaptation
are central problems for the organisation of the industrial system. Under these circumstances
standard technological solutions may fail to address the challenges posed by a complex
context, that require ‘unique’ and customised solutions, including equipment, knowledge or
processes. In this sense, although the need for customised, ad hoc solutions can be partly
interpreted as a consequence of the previous two exchange conditions (uncertainty of rapidly
evolving regulatory framework and shortage of raw materials) it is also related to the existence
of competitive drivers that involve innovation and task complexity.