Accepted Manuscript

A social capital paradox: Entrepreneurial dynamism in a small world clean technology

cluster

Deborah E. de Lange

PII: S0959-6526(16)31230-6
DOI: 10.1016/j.jclepro.2016.08.080
Reference: JCLP 7872

To appear in: Journal of Cleaner Production

Received Date: 2 June 2016

Revised Date: 1 August 2016
Accepted Date: 18 August 2016

Please cite this article as: de Lange DE, A social capital paradox: Entrepreneurial dynamism in a small
world clean technology cluster, Journal of Cleaner Production (2016), doi: 10.1016/j.jclepro.2016.08.080.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
Word count: 9841

A social capital paradox: Entrepreneurial dynamism in a small world clean technology

cluster

Abstract

PT
Clean tech entrepreneurs have struggled to gain investor confidence because of some particular

characteristics and circumstances of the industry. This research combines network and

RI
sustainable development literatures in the clean tech context to support the logic of a new

SC
investment approach that may stimulate sustainable investing in clean tech. Theory is proposed

to suggest that there may be an advantageous social capital paradox where strong ties in a cluster

U
lead to dynamism rather than decay. The clean tech industry provides a context where strong ties
AN
offer network stability in a small world cluster such that it is a value-creating organizational form

offering greater dynamism. Two related propositions are developed to support the social capital
M

paradox. They lead to a theoretical conclusion that long term integrated partner solutions where

partners are also resource constrained lead to successful alliances supporting a dynamic cluster
D

that will grow over time. A practical conclusion is that investing in a connected cluster of firms
TE

might be less risky compared to investing in a new firm or even a portfolio of well diversified
EP

assets. De-risking clean production investments may be achievable through a small world

network cluster-backed security.

C
AC

Keywords: Clean technology; small world networks; sustainable entrepreneurship; social

capital; sustainable investing; corporate social responsibility

1
 ACCEPTED MANUSCRIPT
Word count: 9841

1. Introduction

New investment models are needed by clean tech entrepreneurs who face greater funding

difficulties because venture capitalists (VCs) have been discouraged by recent firm failures in the

PT
risky clean technology industry (Pyper, 2014; Smith, 2013; Parad, Youngman, and Knowles,

RI
2013). Government subsidies and public-private partnerships have been helpful for supporting

clean tech product development and increasing investor confidence through early stage risk

SC
sharing (Bürer and Wüstenhagen, 2009; Loiter and Norberg-Bohm, 1999). However, as Olson

(2014: 73) states, “subsidies have not prevented numerous green technology bankruptcies

U
including the infamous 2011 closure of the California based solar panel producer Solyndra, and
AN
these failures have cost taxpayers and private investors billions in lost capital.” Many factors

have contributed to the challenges faced by clean tech including foreign competition, an
M

incumbent fossil fuel industry, and a multitude of competing technologies within subsectors such
D

as solar and energy storage. Moreover, as climate change and environmental pollution are
TE

increasingly urgent issues related to firm negative externalities, corporate social responsibility

(CSR) and sustainable entrepreneurship are more of a focus for all clean tech partners, whether
EP

SMEs, large corporate interests or venture capitalists and other investors.

Investors are beginning to understand the moral implications of their choices where their
C

investments may determine the survival of clean tech entrepreneurs’ startups that face a strong
AC

incumbent fossil fuel industry (Richardson, 2009). The industrialized West has been the primary

source for existing anthropogenic carbon emissions in the atmosphere (Bradshaw, 2010). Thus,

emerging and developing economies feel justified in adopting the same polluting path to

industrialization. As billions of people in lesser developed economies achieve a higher standard

2
 ACCEPTED MANUSCRIPT
Word count: 9841

of living by producing commensurate per capita levels of pollution as in the West, this leads to

future climate disaster. The West must lead the way to change so that others will follow.

Therefore, this research focuses on a Western context with the view that many modern thinking

investors are seeking impact investing opportunities so as to support clean tech while also

PT
achieving satisfactory returns (Jackson, 2013; Richardson, 2009).

RI
Accordingly, this article sets out to build on previous work in a special issue of the

Journal of Cleaner Production called Financing Cleaner Production and other work since then

SC
such as Bocken’s (2105) work on sustainable venture capital. Bocken (2015) explains that

venture capital has an important role to play in financing sustainable start-ups. Recognizing the

U
challenges for investors who must balance risk, return, and responsible behavior, this research
AN
develops network theory in support of a new small world network cluster model of investment.

This model could reduce investment risk, whether in clean tech or other industries that display
M

similar characteristics. It uses a deductive reasoning approach with supporting illustrations.

D

Theory development is achieved by combining network and sustainable development literatures

TE

related to clean technology and the financing of it. The interpretation of sustainable development

used in this work is that it “is development that meets the needs of the present without
EP

compromising the ability of future generations to meet their own needs.” (Brundtland, et al.,

1987). Also, in accordance with Gosens, Lu, and Coenen (2015: 379), clean tech includes:
C
AC

technologies that have a reduced environmental impact, i.e., have reduced

environmental emissions or natural resource use, when compared with

conventional technologies in providing similar products or services (cf. OECD,

2014; Truffer, 2012). It includes a wide variety of technologies for renewable

3
 ACCEPTED MANUSCRIPT
Word count: 9841

energy generation, energy efficiency and energy storage, sustainable water

management, sustainable mobility, waste management, and improved resource

efficiency (KPMG, 2013).

PT
This research proposes that a small world clean tech network cluster may build social

RI
capital that creates a base for dynamism rather than decay. It is theorized that a clean tech cluster

avoids an often cited problem of redundant information circulating among its strong ties such

SC
that the cluster becomes stagnant and decays for a lack of new revitalizing information (Capaldo,

2007; Granovetter, 1973; Uzzi, 1997). Instead, the strong ties offer stability so that the cluster is

U
resilient to change and individual firm failures. This stability affords the otherwise risky firms
AN
opportunities to be dynamic, meaning that they partner and become increasingly active in joint

innovation and business activity. In accordance with Nahapiet and Ghoshal (1998; 243), social
M

capital is defined as:

D
TE

the sum of the actual and potential resources embedded within, available

through, and derived from the network of relationships possessed by an

EP

individual or social unit. Social capital thus comprises both the network and the

assets that may be mobilized through that network (Bourdieu, 1986; Burt, 1992).
C
AC

This definition is relevant for this work because it is the network cluster formation together with

the interactions of the cluster participants, the clean tech firms, which produce the social capital.

So, this research proposes that in some contexts, such as clean tech, there is increasing long term

cooperative innovation in and growth of the cluster. A firm network cluster can become more

4
 ACCEPTED MANUSCRIPT
Word count: 9841

vibrant because it offers a base of strength that acts like a springboard enabling action. New

information is produced by the firm interactions and innovations. Thus, propositions are

developed to address the research question, “Under what conditions will a stable cluster of firms

build social capital that leads to dynamism (innovation and growth) rather than stagnancy or

PT
decay?” It will be proposed that dynamic clusters result when a fast changing industry requires

RI
long term cooperation for the development of complex, integrated, and novel long term solutions

under circumstances of extreme resource constraints. As a corollary to this, it will be proposed

SC
that the longer term, stable partnerships in dynamic clusters lead to increasing cluster size, within

limits (Fleming, King, and Juda, 2007; Gulati, Sytch, and Tatarynowicz, 2012). Thus, the theory

U
provides some insight as to how small world networks might evolve under the conditions such as
AN
those of the clean tech industry. This theory development will be used to support practical

financial innovation that reduces risk in clean tech investing so that the adoption of cleaner
M

production can proliferate.

D

The next two sections briefly introduce network theory and the clean tech context. The
TE

third following section presents the development of the theory and propositions. Finally, the

article concludes and makes recommendations for future research based on the theoretical
EP

development.
C

2. Network theory: Small worlds in the clean tech context

AC

Networks are relational structures where actors are connected through various types of

relationships such as business alliances as in this research (Wasserman and Faust, 1994).

Research characterizes small world networks as a form of social organization that is made up of

many clusters of tightly interconnected actors having sparse connections between the clusters

5
 ACCEPTED MANUSCRIPT
Word count: 9841

(Watts, 1999; Kogut and Walker, 2001; Uzzi and Spiro, 2005). Figure 1 shows an example

diagram of a small world network that illustrates the network concepts discussed in this research.

In clean tech, a small world network is made up of connections that are inter-firm alliances based

on, “…voluntary arrangements among independent firms to exchange or share resources and

PT
engage in the co-development or provision of products, services, or technologies.” (Gulati,

RI
1998). The clusters in Figure 1 are the groups of firms connected by inter-firm alliances.

Research suggests that, “a cluster should be thought of as a set of nodes that has more and/or

SC
better connections between its members than to the remainder of the network.” (Leskovec, Lang,

Dasgupta, and Mahoney, 2011: 29). Olson (2014) provides an example of these clusters in clean

U
tech through his discussion of green innovation value chains (GIVCs) in the context of the solar
AN
panel industry. Also, SunRun’s solar panel supply (e.g., Canadian Solar and REC Group of

Norway supply SunRun) and extensive installation partnership network is a specific example of
M

the importance that this residential solar supplier places on relying on partners in its network
D

cluster over the long term (Sunrun.com).

TE

{Place Figure 1 about here}

C EP

This study is carried out at the network structural level considering a tightly knit cluster
AC

of firms not bound by geographic proximity. The small world cluster, as shown in Figure 1,

facilitates an interdependent group dynamic (Uzzi and Spiro, 2005), creating social capital

(Coleman, 1988; Walker et al., 1997) that may build a stronger and more stable entity for

investment than a new venture is on its own. The term social capital refers to the building of

beneficial social relationships that include the knowledge that circulates through the close

6
 ACCEPTED MANUSCRIPT
Word count: 9841

relationships (Aldrich, 1999; Coleman, 1988; Portes, 1998). Previous research has suggested that

cluster stability endures even when the cluster experiences external forces of change and/or

changes of individual members (Kogut and Walker, 2001). However, research has not

considered clusters as being able to create long term dynamic value. The inherent nature of

PT
sustainable technologies is that they will last and provide benefits over the long term. Network

RI
research does suggest that small world networks augment firms’ and the clusters of firms’

innovativeness because knowledge transfer is facilitated from cluster to cluster through weak ties

SC
(Fleming et al., 2007; Verspagen and Duysters, 2004; Schilling and Phelps, 2007). However, in

the case of within-cluster interactions, the literature tends to take the view that once the new

U
knowledge has been exhaustively shared the same information cycles around the group and the
AN
cluster becomes stagnant (Granovetter, 1973; Gulati et al., 2012; Uzzi, 1997).

This assumption in network literature that the strong ties within clusters circulate the
M

same information has remained generalized across contexts (Capaldo, 2007; Gulati, 1995a,
D

1995b). Thus, a cluster becomes insular and decays because of its stagnancy in terms of its
TE

learning abilities and level of business activity. It has succumbed to group-think (group members

adopt a set of homogeneous views) where only redundant information circulates (Granovetter,
EP

1973; Uzzi, 1997; Gulati et al., 2012). However, research finds that problem solving is improved

in these close relationships because they facilitate the exchange of fine-grained information
C

(Uzzi, 1996, 1997). But previous research also tends to take the view that new information is
AC

obtained through weaker bridging ties to other clusters in the small world network (Granovetter,

1973; McEvily and Zaheer, 1999; Burt, 2004). In small world networks, the within-cluster

connections are strong ties and the bridging ties are considered weak ties. Thus, weak ties

convey new information from different clusters whereas strong ties facilitate close working

7
 ACCEPTED MANUSCRIPT
Word count: 9841

relationships within clusters, but the partners are believed to share similar knowledge because

they are part of the same cluster (Walker, Kogut, and Shan, 1997; Watts, 1999; Uzzi and Spiro,

2005).

Within network clusters like those with the characteristics of the clean tech industry, this

PT
research discusses which attributes of networks are particularly relevant for clean tech and how

RI
they might differ compared to an expected general context. Previous research has not discussed

clusters as providing contexts that remain dynamic. The predominant assumption has been that

SC
these clusters decay due to a lack of new knowledge infusion. Some previous research has found

variation in the usefulness of strong ties across contexts (Rowley et al., 2000). Thus, this latter

U
research implies that the assertion made in this research that the behavior of small world clusters
AN
could change depending on contextual factors may be reasonable. Moreover, this examination of

network characteristics in the clean tech context could be instructive for other industries that do
M

not exactly conform to previous theoretical expectations.

D

Although a key assumption of network research is that firms learn through knowledge
TE

exchange, there is little recognition that they might generate significant new knowledge on their

own or jointly with existing partners (Gulati et al., 2012). Firms learn from a partner on making a
EP

connection (Burt, 2004; Gulati et al., 2012; Uzzi, 1997). If the new partner is from the same

cluster, then the learning will tend to be more incremental than if the new partner were from
C

another cluster (Burt, 2004; Gulati et al., 2012; Uzzi, 1997). After the information is shared, little
AC

or nothing new is exchanged or generated later. This research suggests that although this could

be the case in many established industries, it might be different in clean tech. In clean tech, long

term solutions (expected to last twenty to forty years and that may need maintenance and

upgrading over that time) require integrated solutions developed through partnerships of firms

8
 ACCEPTED MANUSCRIPT
Word count: 9841

having specialized knowledge and solutions. An illustration of these partnerships and the reasons

for them comes from Marra, Antonelli, Dell’Anna, and Pozzi (2015) where the

interconnectedness and complementarities of two hundred specialized clean tech start-ups in San

Francisco including the challenges they face are discussed. These SMEs (small and medium-

PT
sized enterprises) must work together to find solutions because they are individually resource

RI
constrained. It is the nature of their changing industry that they must continually innovate in their

areas of specialization. Innovation is costly but, they can find joint solutions when they work

SC
closely together.

3.
U
The clean tech context: Identification of challenges for clean tech compared to other
AN
high tech investment

Part of the aforementioned clean tech investment problem arises because VCs tend to
M

have investment return expectations based on the characteristics of the ICT (information and
D

communications technology) industry (Maake, 2012). ICT generally requires relatively low
TE

capital investment and offers quick exits, often through acquisitions, and high returns on

products often having shorter term lifecycles (Maake, 2012; Gersh, 2013). Clean tech has almost
EP

the opposite set of characteristics - high upfront costs and medium level returns over the long

term (CIE, 2014; Bocken, 2015). The capital intensity of many projects is illustrated with a solar
C

plant example. Installation of solar PV (photovoltaic) power plants based on current technologies
AC

that yield less than 15% efficiencies require large land areas compared to traditional power plants

(e.g., nuclear and fossil fuel), often in remote locations, and long transmission lines to populated

regions, generator and/or battery backup because of intermittent supply, and sometime subsidies

to compete with gas (Olson, 2014). As Bocken (2015) has discussed, sustainable start-ups face

9
 ACCEPTED MANUSCRIPT
Word count: 9841

challenges because of a lack of suitable investors, a strong incumbent industry (fossil fuels) and a

short-term investor mind-set. Others add that clean tech lacks support from financial institutions

and financial instruments are missing (Suzuki, 2015).

Moreover, these firms do not have clear exit strategies and thus, VCs struggle with

PT
whether they are attractive investments or not (Bocken, 2015). According to recent literature,

RI
some VCs are interested in sustainable investing and can be helpful if they make patient

investments that will ultimately prove clean tech’s capability to be successful on a triple bottom

SC
line basis (Bocken, 2015; Bürer and Wüstenhagen, 2009). In a Tech Crunch article, Rob Day

(2015), a venture capitalist with Black Coral Capital explains that exits and returns have been

U
achieved in clean tech but not at the expected levels. He thinks that investor strategies may be
AN
more of a problem than the actual market itself. Although many VCs have lost enthusiasm,

corporations such as GE and Google have been increasing their investments. The acquisitions
M

that VCs would like to see sometimes occur, but not at the expected rate because SME clean tech
D

firms are not as likely to buy out their partners.

TE

Thus, the networks tend to be maintained through cooperative strong ties built over the

lifetimes of projects. For example, the University of California San Diego had a micro-grid built
EP

cooperatively with OSISoft, Viridity, and Power Analytics (UCSD Case Study). Other

technology partners in solar on the same grid are SoiTech and Sanyo systems. In addition to this
C

specific example, at OSISoft’s website one can find a directory called OSIsoft EcoSphere
AC

Partners which lists hundreds of the firm’s technology partners (OSISoft, 2016). Such a network

grows over time through a great deal of effort. More evidence of the importance of partnering

and the longevity of these relationships is illustrated by NREL’s (The National Renewable

Energy Lab in the United States) stated orientation towards partnering over decades. This lab

10
 ACCEPTED MANUSCRIPT
Word count: 9841

seeks to establish what it calls CRADAs (cooperative research and development agreements) so

that its innovations are developed in partnership with market players (NREL, 2016). Partners in

clean tech work together over the long term because of the many years required for

multidisciplinary product development and integration into existing markets. Not even a large

PT
government lab like NREL seeks to do it alone. Instead, it aspires to be a wide reaching partner

RI
supportive of an entire industry. These illustrations are demonstrative that partnerships in clean

tech are of high importance, they do not tend towards decay, and remain in place over time.

SC
Prior network theory development has rarely used such a context. Network research has

used the ICT industry context (Coviello, 2006; Gulati et al., 2012; Rowley, Behrens, and

U
Krackhardt., 2000) and other industry contexts (Beckman, Haunschild, and Phillips, 2004; Bell,
AN
2005; Madhavan, Gnyawali, and He, 2004) or has been essentially industry agnostic, sometimes

specifying innovation or R&D networks broadly (Baum, Cowan, and Jonard, 2010; Eguíluz,
M

Zimmermann, Cela‐Conde, and Miguel, 2005).

D

Biotech has been used in network studies (Walker, Kogut, and Shan, 1997) and although
TE

it is also unlike ICT, it is still different from clean tech. This needs explaining because the two

industries are often erroneously likened to each other. The differences are important because they
EP

are related to industry patterns of and time frames for partnering and exits such as acquisitions.

Biotech has established industry patterns where a small biotech firm is usually acquired by a
C

large pharma firm (Behnke and Hültenschmidt, 2007; Giniatullina, Boorsma, Mulder, and van
AC

Deventer, 2013).Venture capital tends to invest in a biotech firm so it can demonstrate early

efficacy and then find an exit through pharmaceutical firm takeover. Thus, long term

partnerships are not needed because the pharmaceutical firms have the resources to acquire

biotech and complete the drug approval stages (Behnke and Hültenschmidt, 2007; Giniatullina et

11
 ACCEPTED MANUSCRIPT
Word count: 9841

al., 2013). Venture capitalists enjoy this pattern in biotech as it reduces uncertainty for exits and

returns on their investments (Behnke and Hültenschmidt, 2007; Giniatullina et al., 2013).

A contrasting set of examples in clean tech comes from a US Department of Energy

funding initiative called the DOE Grid Modernization Laboratory Consortium (GMLC) where

PT
many partnerships have been formed to work on multiple US electric grid projects all across the

RI
country (St. John, 2016). Examples abound of collaborative projects including many private

sector partners engaged in public-private partnerships (PPPs). On DOE Project 16 called,

SC
“Stabilizing the Power System in 2035 and Beyond: Evolving from Grid-Following to Grid-

Forming Distributed Inverter Controllers”, the partners include: UC Santa Barbara, University of

U
Minnesota, Arduino, SunPower, HECO, and Schneider Electric working with the National
AN
Renewable Energy Laboratory (NREL). Although few of these projects include small firms

likely because of the scale of these research projects, the list of these PPPs makes it evident how
M

important partnerships are for this industry even when many larger players are involved. The
D

biotech model will only sometimes apply when a large corporation buys a smaller clean tech
TE

firm, which happens, but VCs cannot count on this as recent history has shown (Day, 2015).

Whatever the barriers, investors are increasingly considering CSR and sustainability
EP

important to an extent that they are influencing firms towards cleaner production through their

investing choices (Huhtala, 2003a, b; O’Rourke, 2003). Opportunities for sustainable (often
C

referred to as “ethical” or “impact”) investment often take the form of screened mutual funds
AC

holding portfolios of large firms, and many of these funds have been providing competitive

returns (O’Rourke, 2003). Ethical investors can also choose to invest in clean-tech start-ups and

encourage the larger firms that they are invested in to invest in these start-ups (Huhtala, 2003a).

These investors together with the venture capitalists that Bocken (2015) identifies as potentially

12
 ACCEPTED MANUSCRIPT
Word count: 9841

supportive may be more encouraged if a new investment vehicle that reduces the risk in clean

tech investing were devised. In fact, Huhtala (2003b: 616) makes a key recommendation “for the

adoption of cleaner production investments world-wide” and says that there is a “need for the

financial services sector to identify cleaner production as an investment opportunity through

PT
financial innovation.”

RI
Through the development of an understanding of how network theory applies to the clean

tech context, this research suggests a financial innovation which is a ‘small world cluster-backed

SC
security’. Theoretical propositions are advanced to support the logic of this practical financial

innovation. In addition, these conceptual developments imply contributions to CSR, sustainable

U
investment, and sustainable entrepreneurship theory on funding models. The stable small world
AN
network cluster could suggest an alternative investment approach preferable to a different

audience of longer term investors. They may be sustainable (or ethical) investors, but not
M

necessarily as it is hoped the rationale will be appealing to a wider range of investors looking for
D

lower risk medium level returns. Thus, sustainable investment might become mainstream and
TE

clean tech entrepreneurs will more easily find an interested investor audience.
EP

4. Network cluster theory and proposition development

In the following section, two propositions are developed (See Figure 2). One of them
C

delineates conditions that could lead to a small world cluster becoming a longer term dynamic
AC

value-building organizational form. This supports the idea that a clean tech cluster could be

worth investing in as a less risky investment compared to clean tech firm on its own. The second

proposition suggests that a corollary to the discussion is increasing cluster size. Thus, the clean

tech cluster is resilient and active to an extent that it grows by attracting more firms.

13
 ACCEPTED MANUSCRIPT
Word count: 9841

{Place Figure 2 about here}

The literature relating to small worlds examines how networks evolve and which

structures are effective for productive collaboration. Small world network configurations are

PT
widely believed to be efficient at supporting a variety of types of innovative activity (Cowan,

Jonard, and Zimmerman, 2007). These configurations are highly clustered, from a network

RI
perspective, but have low overall density and short characteristic path lengths. A characteristic

SC
path length is the mean geodesic length (average shortest distance) in the network (Watts, 1999;

Robins, Pattison, and Woolcock, 2005). Given a random network having n actors and k

U
relationships with limiting values for the average path length, ln(n)/ln(k), and clustering
AN
coefficients1 much greater than k/n, the network is considered a small world (Watts and Strogatz,

1998). For example, in the clean tech industry context, n would be the number of clean tech
M

firms in the network and k would be the number of alliances connecting all of those firms in the

network. If the clean tech network conforms to the specifications in the aforementioned
D

definition, then it is a small world network with the clusters, as previously discussed and
TE

illustrated in Figure 1.

As the cluster gains more members, they will have some redundant capabilities so if one
EP

partnership fails another cluster firm can fill the void quickly (Burt 2004; Granovetter 1973;
C

Schilling and Phelps, 2007). This makes the cluster resilient to an overall breakup or failure of
AC

the cluster because it is self-healing.

Resilience is also facilitated by reputational knowledge that helps firms find substitute

partners. Reputational information travels in the cluster so that even if two cluster firms are not

1
“The local clustering coefficient is measured as the number of actual links connecting all
neighbors of a focal actor with one another, divided by the number of all possible ties among
those nodes. The measure is subsequently averaged over all the actors in the network and shows
whether one actor’s direct contacts typically also know each other.” (Gulati et al., 2012: 450)
14
 ACCEPTED MANUSCRIPT
Word count: 9841

directly connected, they will likely know of each other and each other’s connections (Ahuja,

2000; Baum, Shipilov, and Rowley, 2003; Gulati et al., 2012 ). This reputational knowledge is

limited to the local cluster (Johanson and Vahlne, 2003; Pattison and Robins, 2002; Rivkin and

Siggelkow, 2007). Because the cluster can offer this resilience, this research does not require an

PT
assumption that a stable cluster cannot have firm failures and alliance break-ups. Alliances have

RI
a high rate of failure (Schrank and Whitford, 2011). However, it is proposed that the cluster

remains stable overall because its social capital enables it to be self-healing (Kogut and Walker,

SC
2001). For example, the CrunchBase profile for a solar company called Brightsource Energy

provides evidence of the firm’s dense cluster of partners that offer it stability in the face of high

U
risks. It has a list of fifteen equity investors including a recent Alstom investment. It is also
AN
partnering to deliver customer solutions. A Greentechmedia article on the Brightsource

CrunchBase profile mentions its investment with NRG Energy and Google in a solar plant in the
M

Mojave Desert that is delivering electricity to Pacific Gas & Electric (PG&E) (Lacey, 2015).
D

One can find more stories of clean tech resilience. A123 Systems was a failing lithium
TE

ion battery firm in 2012. Fisker Automotive which is a luxury electric car maker comparable to

Tesla had used the A123 Systems batteries as a sole supplier. It also later sought bankruptcy, but
EP

Fisker was relaunched with BMW battery systems. Thus, another battery supplier in the network

was found that could replace A123 Systems. A123 Systems also bounced back with new owners
C

and sells microhybrid batteries to new partners including Mercedes-Benz, Porche, and others.
AC

This illustrates how interfirm network knowledge can facilitate the survival of clean tech firms

by matching them with new and replacement partners. Another observation from an investor’s

perspective is that one may have chosen to invest in Tesla over Fisker if one were to have placed

15
 ACCEPTED MANUSCRIPT
Word count: 9841

greater focus on the partnership networks rather than solely on knowledge about the individual

firms.

Some literature points to the cooperation and coordination required for learning complex

knowledge leading to alliance stability and success (Cowan, Jonard, and Zimmerman, 2007;

PT
Hamel, 1991; Garcia-Pont and Nohria, 2002). Although it is rare that research considers a long

RI
lasting alliance as valuable or successful (Arino, 2003) some has (Shrank and Whitford, 2011).

In clean tech, there is a necessity for longer term cooperation in partnerships because joint

SC
projects must be maintained and upgraded, including infrastructure and integrated software

systems, for decades (CIE, 2014). The earlier UC San Diego case is exemplary of this where

U
OSIsoft, Viridity and Power Analytics worked together to make the micro-grid function via
AN
integrated software systems over the long term. The other previous example where Brightsource

Energy, Google and NRG Energy were in partnership to deliver solar energy to PG&E is also
M

exemplary of long term partnering. When partners deliver these products and services together
D

they collectively learn how to improve their offerings, thus information and knowledge is
TE

generated through the joint projects. This further illustrates how long term cooperation and

reliability are critical in this industry – it enables success for all partners and infuses them with
EP

knowledge and experience for the next project.

Repeated partnering with the same entity is advantageous especially if available follow-
C

on work extends the last project with the same customer or another customer needs somewhat
AC

similar solutions (Gulati, 1995a). For example, after the UC San Diego micro-grid project, an

announcement was made that OSIsoft together again with Viridity, Power Analytics, UC San

Diego, and SDG&E (and additional new partners, Conner Networks and Spirae) would work

together on a US Department of Defense (DOD) contract for a secure cluster of micro-grids at

16
 ACCEPTED MANUSCRIPT
Word count: 9841

three U.S. Navy facilities in the San Diego area (Brockman, 2013). See Figure 3, below, for a

pictorial illustration of the dynamism that can grow in a clean tech small world cluster. Notice in

the diagram, that partners think in advance about how they may work together beyond the

current project and build future strategies together. Thus, competition within clean tech

PT
partnerships is rarely a consideration (Khanna et al., 1998; Aggarwal et al., 2011) being

RI
counterproductive and leading to network reputational damage (Uzzi, 1996, 1997). Instead,

reciprocity becomes an unwritten rule (Capaldo, 2007).

SC
{Place Figure 3 about here}

U
AN
Partnerships exist under circumstances of extremely constrained resources, as has been

illustrated by highly publicized news reports of many clean tech firm failures (e.g., Solyndra,
M

A123 Systems, Fisker) and venture capital shying away (Fehrenbacher, 2014; Pyper, 2014). The

constant requirement to invest in innovation is costly. Thus, searching out new partners for each
D

new project and learning to work with them is an unwanted diversion. They choose partners
TE

carefully because they can only manage a select few at any time, thus they prefer that

partnerships pay off over the long term (Gulati, Nohria, and Zaheer, 2000; San Diego
EP

discussions, 2012). In fact, research has demonstrated that a good reputation in a cluster and the
C

capability to maintain long-term alliances are competitive advantages valued by the market
AC

(Baum et al., 2003; Gulati et al., 2012; Kale, Dyer, and Singh, 2002). Moreover, maintaining a

technological lead through agility requires that the SMEs do not become large and, hence they do

not often buy each other out. Instead, partners keep their ties over the long term to coordinate on

projects that will require their unique capabilities and joint attention over decades.

17
 ACCEPTED MANUSCRIPT
Word count: 9841

Thus, social capital in a dynamic cluster is valuable for finding partners and for building

cooperative and innovative relationships that offer long-lasting competitive advantage (Greif,

1993; Gulati et al., 2012). Stable cluster relationships that support dynamism create a base for

firm growth. See Figure 4, below, which depicts the new dynamic small world cluster theory

PT
developed here in the clean tech context as compared to the predominant theoretical view in

RI
network theory that has been generalized across contexts. Notice that under the new theory, a

network cluster does not need to receive information from other clusters to avoid decay.

U SC
{Place Figure 4 about here}
AN
This long term growth and development can be attractive to investors and thus this discussion
M

provides logic for their investment in a cluster. The following proposition summarizes the
D

preceding discussion by tying industry characteristics, where integrated partner solutions over
TE

the long term are required in an extremely resource-constrained environment, to a dynamic

cluster.
EP

P1: A small world network cluster will tend to become more dynamic as integrated partner
C

solutions tend towards the longer term and when the partnering firms are more resource
AC

constrained.

As has been established, the working history between firms that results because of the

aforementioned industry characteristics leads them to work on future projects together and

jointly innovate, thus extending, broadening, and/or deepening their relationships (Uzzi, 1997).

18
 ACCEPTED MANUSCRIPT
Word count: 9841

The UC San Diego micro-grid partners who later obtained the DOD’s contract is an excellent

example of this kind of success. Cluster partners build go-to-market strategies together. In this

process of lengthening their relationships, they dynamically move ahead and build positive

reputations in the cluster (Baum et al., 2003).

PT
Also of note about these clusters is that they offer safeguards against complete failure, in

RI
addition to resilience when a single or a few firms experience difficulties. This happens because

allied partners bring complementarities to projects so a cluster is rarely solely composed of firms

SC
in the same subsector or focused niche. Clusters offer some diversification alongside

interdependencies because some firms are not “pure plays” solely devoted to a particular clean

U
technology. For example, a manufacturer that is a cluster partner could make clean tech
AN
apparatus as well as other goods. Alternatively, a large strategic investor like GE, Siemens, or

Google could be invested in clean tech and other sectors. OSIsoft, a medium sized company
M

often associated with clean tech is a technological project partner working across many
D

industries. Other examples of technology partners that diversify the cluster may include drone
TE

companies looking to oversee the maintenance of solar and wind farms, software firms involved

in the energy management and maintenance systems, and smart grid firms among many other
EP

possibilities. They are all likely cluster partners which will diversify their customer bases and are

unlikely to go down because their pure play partners fail. Although it is possible that all firms in
C

a cluster could have problems at once, as could happen in a broad based financial crisis, a cluster
AC

investment is potentially a better investment strategy than many of the usual options. A cluster

offers better diversification than does a single firm. Moreover, a cluster that is doing well is

better than a randomized portfolio of investments as is next explained.

19
 ACCEPTED MANUSCRIPT
Word count: 9841

Traditional portfolio investing offers an almost opposing philosophy over cluster

investing and some comparative discussion is provided here to highlight the benefits of the new

possible approach (Markowitz, 1952; Perold, 1984). Normally, a set of independent firms is

chosen by an investor as part of a diversified portfolio, purposely avoiding the interdependencies

PT
that an interconnected cluster of firms offers. The point in traditional portfolio theory is to reduce

RI
risk by making a variety of investments in select firms in industries that move in opposing

directions so that when one industry is down the others are up. With interdependencies, as a

SC
cluster has, one may first think that risk is very high because when the clean tech industry is

down, so is the entire cluster. However, this potential scenario is unlikely as explained earlier

U
due to the variation in types of partners and when considering the long term. A single firm might
AN
go bankrupt in the short term. Over the long term, the cluster investment may be safer and offer

better returns because long term projects continue to offer returns over their lives (decades in
M

clean tech) and the cluster is resilient, as explained earlier. Moreover, a cluster offers greater
D

upside potential. By avoiding full diversification, the upside potential is increased because gains
TE

are not countered by balancing losses.

A set of clean tech cluster investments can act more like a long term bond offering
EP

potentially higher value than a typical corporate bond. A related diversified corporation may

offer integrated value (and it has been shown to be more often of higher value than an unrelated
C

diversified firm), but a cluster is not the same (Bettis, 1981). Corporations often do not choose
AC

the right combination of businesses to be in and continue without clarity on this matter, perhaps

because a hierarchy controlled by boundedly rational actors attempts direction over and

coordination of all of the assets (Simon, 2010). However, a cluster’s alliances are self-organizing

and experience more market pressure to remain effective, as discussed earlier. Thus, clean tech

20
 ACCEPTED MANUSCRIPT
Word count: 9841

alliances are disciplined by a stronger market mechanism than are difficult-to-measure corporate

synergies built among complementary subsidiaries (Tsai, 2002). Moreover, clean tech cluster

firms that are not large corporations cannot afford as much slack which is usually part of

corporations (Cyert and March, 1963; Singh, 1986). Slack can foster inertia rather than

PT
nimbleness and, in fact, corporations often strategically partner with small innovative firms for

RI
revitalization (de Lange, 2011; Nohria and Gulati, 1996; Pyper, 2014). Thus, a clean tech cluster

offers a set of more strongly related assets without the excess slack of a corporation. Ultimately,

SC
the long term projects and reputational effects that bind a stable clean tech cluster suggest that

investing in a cluster is a safe long term bet offering medium level returns.

U
In addition to the aforementioned cluster benefits, it is possible that the dynamic behavior
AN
among partners in the cluster becomes widely adopted behavior – thus, the cluster becomes

homogenous in this sense (Capaldo, 2007; Gulati et al., 2012). A culture of dynamism builds in
M

the cluster and it is self-reinforcing because the productive cooperative relationships generate
D

continuing success. Cooperation becomes a model for behavior amongst cluster partners.
TE

Moreover, the general success found through partnering is a basis for a positive reputation for the

entire small world cluster within the small world network. Reputation builds because brokers
EP

within the cluster relay positive information to other clusters which spreads to other parts of the

network (See Figure 1). Therefore, previous theory that says that strong ties in a cluster lead to
C

homogenization through their cohesion is applicable here (Capaldo, 2007; Gulati et al., 2012).
AC

However, these clusters exhibit dynamic homogenous behavior over the long term, not behavior

that leads to stagnation (Gulati, Nohria, and Zaheer, 2000).

Furthermore, firms are under pressure to maintain this culture of success through

cooperation for themselves and their cluster. Deviating from established cluster norms by

21
 ACCEPTED MANUSCRIPT
Word count: 9841

betraying a partner for short term gains is information that would travel around to other cluster

members. Such a deviating firm would lose the confidence of others in the cluster and would be

expelled in the sense that others would not ally with it (Brass, Galaskiewicz, Greve, and Tsai,

2004; Larson, 1992). The others understand that their norms of long term cooperation affect their

PT
own and their cluster’s likelihood of survival and success (Gulati et al., 2000). Maintaining

RI
cluster social capital is firm survival (Lin, 1999).

Moreover, the social capital offers a benefit to cluster firms where they can be

SC
knowledgeable about firms in other clusters. Brokers may facilitate this knowledge exchange.

Brokers are firms that reach across structural holes (structural holes exist where there is a single

U
tie connecting two clusters by a broker firm on each side) and can learn about other potential
AN
partners in other clusters in the small world network (Burt, 2004; Kogut and Walker, 2001;

Obstfeld, 2005). (Please see Figure 1 which shows the role of brokers in the network.) OSIsoft’s
M

directory of network partners is illustrative of its brokerage position across industries, not only
D

within clean tech. When new partners are needed, they can be found either within the same
TE

cluster or in another connected cluster through a broker firm that can recommend reliable

partners based on reputational knowledge. For example, in the case of the aforementioned DOD
EP

micro-grid contract, new partners (Conner Networks and Spirae) were added to work with the

original partners. Cluster growth ensues as new firms join the cluster whether because they are
C

attracted by the positive reputation of the cluster and seek a partner within it or because they are
AC

discovered through the network and are invited to join as a partner within the focal cluster.

Although the net result of failures, departures and additions to the cluster is difficult to predict, if

the general pattern is to maintain long term relationships, then the net changes to the cluster are

expected to result in an increase in cluster size.

22
 ACCEPTED MANUSCRIPT
Word count: 9841

Limits to cluster size would be found in the unlikelihood that an entire small world

network would aggregate into a single cluster (Gulati et al., 2012). The clusters would not

aggregate this way unless the industry became focused on a large monopoly firm which is

unlikely in the foreseeable future. Also, some firms are not technologically compatible or

PT
complementary in anyway so, they will not be in the same cluster. Moreover, strong, direct

RI
competitors may view it as in their interests to remain clear of each other. Larger firms such as

those in the electric car sector might keep certain partners exclusive so that proprietary

SC
knowledge is not leaked from one firm to another. However, Tesla has released all of its patents

for wide use and Elon Musk has stated that all cars are going to be electric in the future so he

U
does not see electric vehicles as a unique competitive sector (Musk, 2015). Among smaller clean
AN
tech firms, strong competition is found today in rooftop solar (Tong and Mickey, 2016).

However, solar firms are losing sales more often for reasons of customer risk aversion than due
M

to competition. Evidence shows that more competition in the sector is leading to more awareness
D

of solar as a distributed energy option which is positive for these firms (Tong and Mickey, 2016).
TE

Thus, the industry tends to be relatively cooperative. Where there are concerns about competition

for reasons of protection of intellectual property and employee retention, competitors are
EP

predicted as less likely to be in the same cluster of partners.

Overall, the long term productive alliances that foster a positive reputation for the
C

dynamic cluster will lead it to attract additional partners, so it will grow in size. From a
AC

sustainable investor’s perspective, this cluster growth is further indication of the increasing value

of the cluster as a whole. Therefore, rather than solely report on the growth of individual firms,

investment analysts could begin to map clean tech networks and watch clusters’ growth (and

contraction) in aggregate over time. They could also watch as firms migrate from one cluster to

23
 ACCEPTED MANUSCRIPT
Word count: 9841

another. This sophisticated knowledge would inform savvy investors on industry trends and

better investment choices. Remembering an earlier example, if one were considering which firm

to invest in, Tesla or Fisker, their networks might provide additional valuable information along

with their individual financial information. Furthermore, because of the long term nature of these

PT
growing clusters, cluster-backed securities could be developed and invested into with this

RI
information. The following proposition reflects the aforementioned relationship relating the

longevity of the partnerships to cluster growth.

SC
P2: As the longevity of network cluster partnerships increases, more firms will join the cluster

thus increasing cluster size.

U
AN
5. Conclusions, Limitations and Future Research
M

This research has made several contributions to theory by combining network and
D

sustainable development literatures. It has shown how clean tech is an applicable context for
TE

small world network theory as well as how some network theory may be modified for

sustainable technology contexts. Although a cluster might homogenize, it could build a common
EP

culture and set of behaviors that support dynamism among the cluster firms rather than

stagnancy. Thus, this article has put forward the theory that a small world cluster of firms can
C

build beneficial long term social capital using illustrative examples taken from previous
AC

academic literature and directly from the clean tech industry. This occurs when firms are

connected by alliances and where expectations are to build joint value over the long term, as is

often the case with clean tech. At the same time, many cluster clean tech SMEs are constrained

by a shortage of resources so, they do not often opt to buy their alliance partners. This places the

24
 ACCEPTED MANUSCRIPT
Word count: 9841

connected firms under more pressure to maintain cooperation and rely on each other for

additional value building activities. Moreover, the cluster is self-healing. Thus, the cluster, as a

whole, remains stable over time and may grow, offering the firms a base for value creation that is

not necessarily only incremental. Instead, the shared long term cluster resources enable firms to

PT
discover solutions that may otherwise not be reachable via weak ties because of the specific

RI
assets they have built together. Additional contributions that fall out of this context are: 1) an

instance where long term alliances can be associated with success and value building, and 2) a

SC
mechanism for supporting long term alliances that being the small world cluster under the

conditions as outlined.

U
From a sustainable development perspective, the network theory has important
AN
implications for sustainable investment and funding. The small world network cluster of firms

offers stable support for the clean tech entrepreneur. The entrepreneur can seek out resources
M

easily and safely within the cluster community because reputation is so important for becoming
D

part of and maintaining long term projects. Moreover, a cluster of entrepreneurs could bundle
TE

their firms together as an investment asset for investors who would otherwise consider corporate

bonds, offering them higher returns with less risk than a corporation. A cluster is a set of many
EP

firms connected by market forces, as compared with a corporation having many parts not

necessarily connected by market rationale and therefore, with some inefficiencies. Some VCs
C

interested in supporting sustainable entrepreneurship may find an option here although they may
AC

have to accept longer term, medium level returns.

In any case, VCs have been stepping away from clean tech so, clean tech may need to

find another audience for investment. The cluster may be attractive to a wider audience including

large institutional investors, even fitting into a mutual or pension fund. Also, larger corporate

25
 ACCEPTED MANUSCRIPT
Word count: 9841

investors can use this vehicle to simultaneously invest in new clean technology and in CSR.

Impact investors are seeking investment opportunities that address our moral dilemma with

respect to the West’s responsibility to lead on the wide spread implementation of climate change

mitigation technologies that the developing world must also adopt. This issue may start to affect

PT
the decisions of large institutional investors as public concern regarding climate change increases

RI
since many of these large investors ultimately serve public stakeholders such as pensioners.

This work suggests many future research opportunities. First, this theory may be

SC
generalizable to other existing or future contexts that involve new sustainable technologies,

especially those that address market failures and face challenges attracting the investment that

U
they require to succeed. Another context to consider is eco-industrial parks where geographically
AN
proximate firms build networks through waste exchange for reuse, recycling, and reduction

rather than through informational ties. Whether these localized networks of firms could be
M

interesting investment targets for venture capitalists or other investors would need investigation.
D

As a theoretical limitation, this paper did not consider hierarchy associated with
TE

governance of a small world cluster. The network was assumed to have a flat structure in the

sense that no partner was particularly powerful over the cluster as a whole, often reflected by
EP

high centrality. Thus, an aspect to consider for future research is governance of or within clusters

(Baum et al., 2003; Eguíluz et al., 2005). For example, management of a clean tech small world
C

cluster may add value to it. More investigation is also needed into the factors that explain how
AC

and why small world systems evolve. This work added to this research dialogue in terms of

cluster growth.

Overall, this article presents the basic theory to support the possibility of a new

investment and funding model for clean tech. Thus, follow-on work could develop the

26
 ACCEPTED MANUSCRIPT
Word count: 9841

practicalities of how to create a small world cluster-backed security and demonstrate the viability

of this financial innovation to entrepreneurs and investors with investment simulations. The

simulations would be instructive prior to making real investments. They could use historical data

that compares the rate of return and risk of a clean tech cluster of firms with a standard

PT
diversified investment portfolio and with individual firm performance. Clean tech clusters today

RI
are often composed of a mix of public and private firms, so gathering the financial data is

currently a challenge. Private firms often do not want to share sensitive financial information

SC
related to performance. However, when the industry has a lengthier history and the historical

financial information is no longer consequential to competitive positions, perhaps this

U
information will be accessible for research that can test the theory proposed by this research. In
AN
the meantime, wealthy investors who are industry savvy and can afford to experiment could go

ahead and see if this cluster investing approach works for them.
M
D

References
TE

Aggarwal, V.A., Siggelkow, N., Singh, H., 2011. Governing collaborative activity:
EP

interdependence and the impact of coordination and exploration. Strategic Management

Journal 32, 705 –730.

C

Ahuja, G., 2000. Collaboration networks, structural holes, and innovation: a longitudinal study.
AC

Administrative Science Quarterly 45, 425-455.

Aldrich, H.E., 1999. Organizations evolving. London: Sage.

Arino, A., 2003. Measures of strategic alliance performance: an analysis of construct validity.

Journal of International Business 34, 66-79.

Baum, J. A., Cowan, R., Jonard, N., 2010. Network-independent partner selection and the

27
 ACCEPTED MANUSCRIPT
Word count: 9841

evolution of innovation networks. Management Science 56(11), 2094-2110.

Baum, J.A.C., Shipilov, A.V., Rowley, T. J., 2003. Where do small worlds come from?

Industrial and Corporate Change 12(4), 697-725.

Beckman, C.M., Haunschild, P. R., Phillips, D. J., 2004. Friends or strangers? Firm-specific

PT
uncertainty, market uncertainty, and network partner selection. Organization Science

RI
15(3), 259-275.

Behnke, N., Hültenschmidt, N., 2007. New path to profits in biotech: Taking the acquisition exit.

SC
Journal of Commercial Biotechnology 13, 78–85.

Bettis, R. A., 1981. Performance differences in related and unrelated diversified firms. Strategic

Management Journal 2(4), 379-393.

U
AN
Bocken, N. M. P., 2015. Sustainable venture capital–catalyst for sustainable start-up success?

Journal of Cleaner Production 108, 647-658.

M

Bourdieu, P., 1986. The forms of capital. In J. G. Richardson (Ed.), Handbook of theory and
D

research for the sociology of education: 241-258. New York: Greenwood.

TE

Bradshaw, M. J. 2010. Global energy dilemmas: a geographical perspective. The Geographical

Journal 176(4), 275-290.

EP

Brass, D. J., Galaskiewicz, J., Greve, H. R., Tsai, W. 2004. Taking stock of networks and

organizations: A multilevel perspective. Academy of Management Journal 47(6), 795-

C

817.
AC

Brockman, T., 2013. Company announces partners for ESTCP microgrid project. Power

Analytics Press Releases, December 17th. [Accessed

http://www.poweranalytics.com/company-announces-partners-estcp-microgrid-project/

on February 15, 2016.]

28
 ACCEPTED MANUSCRIPT
Word count: 9841

Brundtland, G., Khalid, M., Agnelli, S., Al-Athel, S., Chidzero, B., Fadika, L., ..., Singh, M.

(1987). Our common future (\'brundtland report\').

Bürer, M. J., Wüstenhagen, R. 2009. Which renewable energy policy is a venture capitalist's best

friend? Empirical evidence from a survey of international cleantech investors. Energy

PT
Policy 37(12), 4997-5006.

RI
Burt, R. S., 1992. Structural holes: The social structure of competition. Cambridge, MA: Harvard

University Press.

SC
Burt, R.S., 2004. Structural holes and good ideas. American Journal of Sociology 110, 349–99.

Capaldo, A., 2007. Network structure and innovation: The leveraging of a dual network as a

U
distinctive relational capability. Strategic Management Journal 28(6), 585-608
AN
Cleantech Incubation Europe (CIE), 2014. Cleantech Incubation Policy and Practice. European

Union, June.
M

Coleman, J.S., 1988. Social capital in the creation of human capital. American Journal of
D

Sociology 94, S95-S120.

TE

Coviello, N.E., 2006. The network dynamics of international new ventures. Journal of

Intenational Business Studies 37(5), 713-731.

EP

Cowan, R., Jonard, N., Zimmermann, J. B., 2007. Bilateral collaboration and the emergence

of innovation networks. Management Science 53(7), 1051-1067.

C

Cyert, R. M., March, J. G., 1963. 3. A behavioral theory of the firm. Englewood Cliffs, N.J.:
AC

Prentice-Hall.

Day, R., 2015. Cleantech is dead, long live cleantech. TechCrunch, April 18th [Accessed

http://techcrunch.com/2015/04/18/cleantech-is-dead-long-live-cleantech/ on February 15,

2016].

29
 ACCEPTED MANUSCRIPT
Word count: 9841

de Lange, D., 2011. High-tech cluster revolution from an organizational ecology perspective.

Chapter 8 in Marcus, A., Shrivastava, P., Sharma, S. and Pogutz, S. (eds.) Cross-Sector

Leadership for the Green Economy: Integrating Research and Practice on Sustainable

Enterprise. New York: Palgrave MacMillan, 141-156.

PT
Eguíluz, V.M., Zimmermann, M.G., Cela-Conde, C.J., Miguel, M.S., 2005. Cooperation and the

RI
emergences of role differentiation in the dynamics of social networks. American Journal of

Sociology 110 (4), 977-1008.

SC
Fehrenbacher, K., 2014. Even for investors with cleantech home runs like Tesla, digital energy is

more attractive these days. Gigaom.com, October 13th.

U
Fleming, L., King, C., Juda, A.I., 2007. Small worlds and regional innovation. Organization
AN
Science 18(6), 938–954.

Gersh, L., 2013. The velocity of obsolescence. Forbes, July 29th.

M

Giniatullina, A., Boorsma, M., Mulder. G. and van Deventer,S., 2013. Building for big pharma.
D

Bioentrepreneur, March 21st.

TE

Gosens, J., Lu, Y., Coenen, L., 2015. The role of transnational dimensions in emerging

economy ‘Technological Innovation Systems’ for clean-tech. Journal of Cleaner

EP

Production 86, 378-388.

Granovetter, M., 1973. The strength of weak ties. American Journal of Sociology 78, 1360
C

–1380.
AC

Greif, A., 1993. Contract enforceability and economic institutions in early trade: The Maghribi

traders’ coalition. American Economic Review 83(3), 99–118.

Gulati, R., 1995a. Does familiarity breed trust? the implications of repeated ties for contractual

choice in alliances. Academy of Management Journal 38, 85-112.

30
 ACCEPTED MANUSCRIPT
Word count: 9841

Gulati, R., 1995b. Social Structure and alliance formation patterns: a longitudinal analysis.

Administrative Science Quarterly 40, 619-652.

Gulati, R., 1998. Alliances and networks. Strategic Management Journal 19, 293–317.

Gulati, R., Nohria, N., Zaheer, A., 2000. Strategic networks. Strategic Management Journal 21,

PT
203-215.

RI
Gulati, R., Sytch, M., Tatarynowicz, A., 2012. The rise and fall of small worlds: Exploring the

dynamics of social structure. Organization Science 23(2), 449-471.

SC
Hamel, G., 1991. Competition for competence and interpartner learning within international

strategic alliances. Strategic Management Journal 12, 83-103.

U
Huhtala, A., 2003a. Special issue on cleaner production financing. Journal of Cleaner
AN
Production 11(6), 611-613.

Huhtala, A., 2003b. Promoting financing of cleaner production investments—UNEP

M

experience. Journal of Cleaner Production 11(6), 615-618.

D

Jackson, E. T. 2013. Interrogating the theory of change: evaluating impact investing where it
TE

matters most. Journal of Sustainable Finance & Investment 3(2), 95-110.

Johanson, J., Vahlne, J.R., 2003. Business relationship learning and commitment in the
EP

internationalization process. Journal of International Entrepreneurship 1(1), 83-101.

KPMG, D., EUPD Research, 2013. Cleantech-standortgutachten 2013 (in German;

C

English summary). Available via: dcti.de/en/cleantech/definition.html.

AC

Kale, P., Dyer, J. H., Singh, H., 2002. Alliance capability, stock market response, and long-term

alliance success: The role of the alliance function. Strategic Management Journal 23(8),

747-767.

Khanna, T., Gulati, R., Nohria, N., 1998. The dynamics of learning alliances: competition,

31
 ACCEPTED MANUSCRIPT
Word count: 9841

cooperation, and relative scope. Strategic Management Journal 19, 193-210.

Kogut, B., Walker, G., 2001. The small world of German corporate networks in the global

economy. American Sociological Review 66, 317–35.

Lacey, S., 2015. NRG could default on its contract with PG&E for the Ivanpah Concentrating

PT
Solar Plant. Greentechmedia.com, December 21. [Accessed

RI
http://www.greentechmedia.com/articles/read/nrg-could-default-on-its-contract-for-the-

ivanpah-concentrating-solar-power on February 15, 2016]

SC
Larson, A. 1992. Network dyads in entrepreneurial settings: A study of the governance of

exchange relationships. Administrative Science Quarterly 37, 76-104.

U
Leskovec, J., Lang, K. J., Dasgupta, A., Mahoney, M. W. 2009. Community structure in large
AN
networks: Natural cluster sizes and the absence of large well-defined clusters. Internet

Mathematics 6(1), 29-123.

M

Lin, N., 1999. Building a network theory of social capital. Connections 22(1), 28-51.
D

Loiter, J. M., Norberg-Bohm, V. 1999. Technology policy and renewable energy: public roles in
TE

the development of new energy technologies. Energy Policy 27(2), 85-97.

Maake, M., 2012. ICT gets lion's share of venture capital funding. Business Times, September,
EP

30th.

Marra, A., Antonelli, P., Dell’Anna, L., Pozzi, C., 2015. A network analysis using metadata
C

to investigate innovation in clean-tech–Implications for energy policy. Energy Policy 86,

AC

17-26.

Markowitz . H.,1952. Portfolio selection. Journal of Finance 7 (1)77-91.

McEvily, B., Zaheer, A., 1999. Bridging ties: A source of firm heterogeneity in competitive

capabilities. Strategic Management Journal 20(12), 1133–1156.

32
 ACCEPTED MANUSCRIPT
Word count: 9841

Musk, E. 2015. Elon Musk on why Hydrogen fuel cell is dumb (2015.1.13). [Accessed

https://www.youtube.com/watch?v=Y_e7rA4fBAo on July 29, 2016].

NREL. 2016. [Accessed http://www.nrel.gov/continuum/partnering/tech_to_market.html,

https://www.nrel.gov/workingwithus/partnership-opportunities.html,

PT
http://www.nrel.gov/docs/gen/fy14/60986.pdf ,

RI
http://www.nrel.gov/news/press/2016/24654 and

http://www.nrel.gov/news/features/2014/15407 on July 26, 2016].

SC
Nahapiet, J., Ghoshal, S., 1998. Social capital, intellectual capital, and the organizational

advantage. Academy of Management Review 23(2), 242-266.

U
Nohria, N., Gulati, R., 1996. Is slack good or bad for innovation? Academy of Management
AN
Journal 39(5), 1245-1264.

OECD, 2014. Online Glossary of Statistical Terms, Definition of Clean Technology.

M

https://stats.oecd.org/glossary/detail.asp?ID¼2988.
D

O’Rourke, A., (2003). The message and methods of ethical investment. Journal of Cleaner
TE

Production 11(6), 683-693.

OSISoft. 2016. OSIsoft EcoSphere Partners. [Accessed http://www.osisoft.com/StartPartners/ on

EP

February 15, 2016.]

Obstfeld, D., 2005. Social networks, the tertius iungens orientation, and involvement in
C

innovation. Administrative Science Quarterly 50, 100-130.

AC

Olson, E. L., 2014. Green innovation value chain analysis of PV solar power. Journal of Cleaner

Production 64, 73-80.

Parad, M., Youngman, R., Knowles, V., 2013. The (continued) rise of the corporation in

cleantech investment, Cleantech Group, April.

33
 ACCEPTED MANUSCRIPT
Word count: 9841

Pattison, P.E., Robins, G.L., 2002. Neighborhood based models for social networks. Sociological

Methodology 32, 301–37.

Perold, A. F., 1984. Large-scale portfolio optimization. Management Science 30(10) 1143-1160.

Portes, A., 1998. Social capital: Its origins and applications in modern sociology. Annual Review

PT
of Sociology 24, 1–24.

RI
Pyper, J., 2014. Strategic or VC Investors? Where Cleantech Startups Are Finding Funding.

Greentechmedia.com, December 1st.

SC
Richardson, B. J. 2009. Climate finance and its governance: moving to a low carbon economy

through socially responsible financing?. International and Comparative Law Quarterly

58(03), 597-626.
U
AN
Rivkin, J.W., Siggelkow, N., 2007. Patterned interactions in complex systems: implications for

exploration. Management Science 53(7), 1068–1085.

M

Robins, G., Pattison, P., Woolcock, J., 2005. Small and other worlds: global network structures
D

from local processes. American Journal of Sociology 110 (4), 894-936.

TE

Rowley, T., Behrens, D., Krackhardt, D., 2000. Redundant governance structures: An analysis of

structural and relational embeddedness in the steel and semiconductor industries.

EP

Strategic Management Journal 21(3), 369-386.

San Diego discussions. 2012. Confidential discussions with clean tech executives were
C

conducted to explore and validate the ideas in this article.

AC

Schilling, M. A., Phelps, C. C., 2007. Interfirm collaboration networks: The impact of large-scale

network structure on firm innovation. Management Science 53(7), 1113-1126.

Schrank, A., Whitford, J., 2011. The anatomy of network failure. Sociological Theory 29(3),

151-177.

34
 ACCEPTED MANUSCRIPT
Word count: 9841

Simon, H. A., 2010. A behavioral model of rational choice. Competition Policy International

6(1), 241-258.

Singh, J. V., 1986. Performance, slack, and risk taking in organizational decision making.

Academy of Management Journal 29(3), 562-585.

PT
Smith, R., 2013. A humbled Kleiner Perkins adjusts its strategy. The New York Times, May 7,

RI
p. B1.

St. John, J., 2016. Breaking down DOE’s new $220M grid modernization plan.

SC
Greentechmedia.com, January, 22nd [Accessed

http://www.greentechmedia.com/articles/read/breaking-down-does-new-220m-grid-

U
modernization-plan on February 17, 2016].
AN
Sunrun.com. [Accessed http://www.sunrun.com/how-it-works/faq#item-2406 on

February 17, 2016].

M

Suzuki, M., 2015. Identifying roles of international institutions in clean energy technology
D

innovation and diffusion in the developing countries: matching barriers with roles of the
TE

institutions. Journal of Cleaner Production 98, 229-240.

Tong, J., Mickey, A. 2016. How to Turn Solar ‘Considerers’ Into Solar Adopters.
EP

Greentechmedia.com, June 20th. [Accessed

http://www.greentechmedia.com/articles/read/How-to-Turn-Solar-Considerers-Into-
C

Solar-Adopters on July 29, 2016]

AC

Truffer, B., 2012. The need for a global perspective on sustainability transitions. Env.

Dev. 3, 182e183.

Tsai, W. 2002. Social structure of “coopetition” within a multiunit organization: Coordination,

35
 ACCEPTED MANUSCRIPT
Word count: 9841

competition, and intraorganizational knowledge sharing. Organization science 13(2),

179-190.

UCSD Case Study. Case Study: University of California San Diego Microgrid. [Accessed

www.osisoft.com/WorkArea/DownloadAsset.aspx?id=9345 on February 15, 2016]

PT
Uzzi, B., 1996. The sources and consequences of embeddedness for economic performance of

RI
organizations. American Sociological Review 61, 674–698.

Uzzi, B., 1997. Social structure and competition in interfirm networks: the paradox of

SC
embeddedness. Administrative Science Quarterly 42, 35-67.

Uzzi, B., Spiro, J. 2005. Collaboration and creativity: the small world problem. American

Journal of Sociology 111(2), 447-504.

U
AN
Verspagen, B., Duysters, G., 2004. The small worlds of strategic technology alliances.

Technovation 24(7), 563-571.

M

Walker, G., Kogut, B., Shan, W., 1997. Social capital, structural holes and the formation of an
D

industry network. Organization Science 8(2), 109-125.

TE

Wasserman, S., Faust, K., 1994. Social network analysis: Methods and applications (Vol. 8).

Cambridge university press.

EP

Watts, D.J., 1999. Small worlds: the dynamics of networks between order and randomness.

Princeton University Press: Princeton, N.J.

C

Watts, D.J., Strogatz, S.H., 1998. Collective dynamics of ‘small-world’ networks. Nature 393,
AC

440-442.

36
 ACCEPTED MANUSCRIPT
Word count: 9841

PT
Figure 1: A diagrammatic example of a small world network and some associated concepts

RI
U SC
AN
M
D
TE
C EP
AC

37
 ACCEPTED MANUSCRIPT
Word count: 9841

PT
RI
Figure 2: A diagram of the main propositions

U SC
AN
M
D
TE
C EP
AC

38
 ACCEPTED MANUSCRIPT
Word count: 9841

PT
RI
Figure 3: A pictorial diagram of the dynamism in a small world clean tech cluster

U SC
AN
M
D
TE
C EP
AC

39
 ACCEPTED MANUSCRIPT
Word count: 9841

Figure 4: A comparative diagram of the predominant network cluster theory versus the
new dynamic network cluster theory

PT
RI
U SC
AN
M
D
TE
C EP
AC

40
 ACCEPTED MANUSCRIPT

Highlights

• Sustainable investors and entrepreneurs find that risk in clean tech is a barrier.
• Small world network clusters of firms can be a value-creating organizational forms.

PT
• In industries like clean tech, clusters may be dynamic rather than stagnant.
• Clean tech might be less risky through a small world cluster-backed security.

RI
• This financial innovation may resolve low investment in clean technology.

U SC
AN
M
D
TE
C EP
AC