Can the government create a vibrant cluster? Understanding the impact of cluster policy on the development of a cluster

Anne-Lorène Vernay

Grenoble Ecole de Management, France

Anne-Lorene.Vernay@grenoble-em.com

Beatrice D’Ippolito

The York Management School, The University of York, United Kingdom

Beatrice.DIppolito@york.ac.uk

Jonatan Pinkse*

Manchester Institute of Innovation Research, Alliance Manchester Business School,

University of Manchester, United Kingdom

Jonatan.Pinkse@manchester.ac.uk

*Corresponding author

Entrepreneurship and Regional Development, In Press

Abstract

Research has debated to what extent policy measures can facilitate or contribute to the

development of clusters. This paper contributes to this debate by questioning how the

government can create a cluster that is self-organising and vibrant but also maintain sufficient

influence to continue using the cluster as a policy instrument. Taking the perspective of

cluster members, the paper investigates how members perceive the ambiguous role of the

government in a government-supported cluster. It analyses to what extent cluster members

value a government-supported cluster and whether they perceive the government as one that

facilitates or hinders them in self-organising the cluster. Empirical evidence is derived from a

case study of a French cluster established as a result of a cluster policy initiative and which

has recently been required to fulfil a new set of objectives by the same government. The

findings suggest that government-supported clusters can self-organise if members are given

the opportunity, but with the consequence that it becomes difficult for the government to

fully control such clusters. To continue steering the cluster’s development, the government

would have to leverage the technology gatekeepers’ power by designing policies that allow

gatekeepers to translate government objectives into meaningful objectives for themselves.

Key words: cluster policy; cluster initiative; energy cluster; cluster participation

2

1 Introduction

Over the past two decades, policy-makers have increasingly tried to support the

development of clusters as a policy instrument to increase the competitiveness of regions,

accelerate entrepreneurship, and stimulate job creation (Huggins & Williams, 2011; OECD,

1999, 2001). Policies of this kind tend to either support existing clusters or create new ones

by establishing collaboration between firms that do not yet co-operate (Benneworth, Danson,

Raines, & Whittam, 2003). The policy measures used include direct and indirect financial

support, start-up support, aid for administration, networks and cooperation, as well as general

assistance for cluster activities (Brenner & Schlump, 2011). Across the world there are many

examples of governments implementing measures to stimulate cluster development (van der

Linde, 2003). The underlying assumption of cluster-based policies is that the government can

facilitate the development of an existing cluster (Hospers, Desrochers, & Sautet, 2009) or

create a new one from scratch (Fromhold-Eisebith & Eisebith, 2005; Palazuelos, 2005).

Whether policy measures can effectively contribute to the development of clusters is a

hotly debated topic. Some scholars are critical about the influence of governments on cluster

formation (Bresnahan, Gambardella, & Saxenian, 2001; Formica, 2003; Guinet, 2003; van

der Linde, 2003). There is evidence suggesting that cluster policy has a limited impact on

member firms’ business performance (Fromhold-Eisebith & Eisebith, 2008). The wish for a

vibrant cluster may justify government support, yet cluster policy could disturb cluster

development as well (Roelandt & den Hertog, 1999). Taking a more optimistic view,

evidence from Japan suggests that government-supported clusters which provide indirect

R&D support in the form of information, advisory services, and organisation of networking

events have a strong and positive impact on innovation and firm performance (Nishimura &

Okamuro, 2011). Similarly, Russian cluster policy has been effective in gaining access to

3

new knowledge through R&D partnerships with international partners (Islankina & Thurner,

2018).

This paper sheds further light on this debate by examining the acceptability and

effectiveness of cluster policy from the perspective of cluster members. First, we examine

how and why members participate in government-supported clusters. What makes such

clusters appealing for different types of firms to join and participate? Second, we investigate

how cluster members perceive the government’s role in cluster development. To what extent

do they believe cluster policy helps or hinders the cluster in effectively delivering its

mission? With cluster policy the government seems to pursue a paradoxical agenda. The

government supports the creation of a cluster with the aim to stimulate the local economy and

let the cluster function independently over time. At the same time, the government wishes to

continue steering the cluster’s direction to make sure it contributes to policy objectives. But,

how can the government create a cluster that at the same time is vibrant and self-organising

and keeps on meeting the policy objectives set out at its foundation? Governments would

have to give cluster members sufficient freedom to decide what the cluster does and how; yet,

they would also need to maintain some control to ensure policy objectives are met. This paper

thus explores to what extent cluster members value government-supported clusters and how

they perceive the government’s seemingly paradoxical role in cluster development.

To address these research questions, we conducted an in-depth case study of a French

cluster that was set up in 2005 as part of a broader government initiative aimed at fostering

the specialisation of different regions. While the national government sets the cluster’s main

objectives and direction for growth, cluster members have room for manoeuvre within these

constraints. The data collection was conducted a few months after the government had set

new objectives. Initially formulated to stimulate cooperation between firms and research

institutions, the renewed policy requires the cluster to develop a stronger focus on

4

commercialisation, generate a meaningful impact on local economies, and become more

financially independent. This change in the objectives made the cluster suitable to investigate

how members experience the government’s influence in steering the cluster’s development

path and whether they see cluster policy as helpful or not.

The paper makes several contributions to the cluster literature. First, the paper adds a

seldom-adopted member perspective (Perry, 2007) to the debate on the acceptability and

effectiveness of clusters and cluster policy (Mueller & Jungwirth, 2016). Our findings show

that members value government-supported clusters, but the perceived benefits strongly

depend on the size, connectedness and strategic objectives of each member (Lechner &

Leyronas, 2012; Perry, 2007). Notably, while smaller members might have more potential to

benefit, in practice many feel they do make the most of their participation in the cluster due to

a power imbalance with larger members. A move from a focus on research to

commercialisation is considered as potentially aggravating this problem because members

come into more direct competition with one another (Clarysse, Wright, Bruneel, & Mahajan,

2014). Second, our case evidence suggests that it does not seem feasible for the government

to keep on using a cluster as a policy instrument as soon as it has started to function as a self-

organising entity (Mueller & Jungwirth, 2016). Once a cluster matures, it develops a

relational form of governance that cannot easily be superseded by the hierarchical governance

put in place by the government (Bell, Tracey, & Heide, 2009). Third, the paper offers insight

for policy makers by drawing attention to the need for the government to strike a balance

between a top-down, interventionist approach and a bottom-up, laissez faire approach

(Jungwirth & Müller, 2014). While changing the direction of a cluster might not be possible

directly, our findings suggest that the government could leverage the position of the

technology gatekeepers to renew the direction of a cluster (Hervas-Oliver & Albors-Garrigos,

2014).

5

2 The nature of clusters and paths of development

The notion of a cluster is featured by different dimensions, which include geography,

firm heterogeneity, agglomeration economies, and diversity of local actors’ strategies. A

multitude of definitions exist and the concept has been criticised for being fuzzy and chaotic

(Martin & Sunley, 2003). Most definitions focus on the “localised” dimension of firm

agglomerations, the production processes of which are closely linked through the exchange of

goods, services, and knowledge (van den Berg, Braun, & van Winden, 2001). For this

research, we consider a cluster to be a “regional agglomeration of sector or value chain

related firms and other organisations (like universities, R&D centres, or public agencies)

which derive economic advantages from co-location and collaboration” (Fromhold-Eisebith

& Eisebith, 2005: 1251). Among a cluster’s main ambitions is the desire to increase local

competitiveness and innovativeness through lower transaction costs (Baptista & Swann,

1998; Iammarino & McCann, 2006) or stronger ability to attract R&D funds (Broekel,

Fornahl, & Morrison, 2015). In practical terms, this ambition means facilitating access to

specialised labour and infrastructure, enabling knowledge spillovers, and fostering inter-firm

cooperation and competition (Lublinski, 2003; Malmberg & Maskell, 2002; Porter, 1990,

1998; Wolfe & Gertler, 2004).

2.1 Top-down versus bottom-up clusters: blurring definitional boundaries

While clusters exist across the globe (Porter, 2000, 2003), local and/or national

culture represents an important factor in shaping cluster formation. Becattini (1991) and

Brusco (1982) have identified clusters as a set of relationsips that occur within a given

geographical area that bring unity to a production system, economic actors, and industrial

culture. The social and industrial context in which members are situated will thus influence

the dynamics of cluster development (Schiele & Ebner, 2013). The question remains, though,

6

whether clusters develop organically in an bottom-up manner or are the outcome of a

purposful act to create a cluster entity in a top-down manner (Fromhold-Eisebith & Eisebith,

2005).

Bottom-up clusters are initiated and governed by private firms willing to benefit from

regional cooperation (Su & Hung, 2009). Firms try to establish linkages and develop joint

strategies on their own accord (Andersson, Serger, Soervik, & Hansson, 2004). The key

actors around whom bottom-up clusters evolve tend to be referred to as technology

gatekeepers (Hervas-Oliver & Albors-Garrigos, 2014; Morrison, 2008) or anchor tenants

(Baglieri, Cinici, & Mangematin, 2012). Due to their relatively large size, technology

gatekeepers form the connection to external knowledge for many of the smaller firms.

Clusters thus tend to develop in a bottom-up way around technology gatekeepers as their

presence grants others access to knowledge not accessible otherwise (Hervas-Oliver &

Albors-Garrigos, 2014; Morrison, 2008).

Top-down clusters – also referred to as cluster initiatives (Fromhold-Eisebith &

Eisebith, 2008; Islankina & Thurner, 2018; Kowalski & Marcinkowski, 2014) – are instead

the result of government policies that promote regional competitiveness and are largely

financed and governed by public institutions (Fromhold-Eisebith & Eisebith, 2005; Su &

Hung, 2009). Government bodies set up a cluster management team to act as broker, create

appropriate structures, manage the cluster, and provide services to registered members

(Jungwirth & Mueller, 2014; McDonald, Tsagdis, & Huang, 2006; Sölvell, Lindqvist, &

Ketels, 2003).

The two cluster types have been compared and contrasted (Fromhold-Eisebith &

Eisebith, 2005; Jungwirth & Mueller, 2014; Su & Hung, 2009). A key characteristic of top-

down clusters is their ability to facilitate access to important external sources of funding

(Fromhold-Eisebith & Eisebith, 2005), although this means that clusters are at the mercy of

7

tightening budgets and changing political objectives because government support does not

last indefinitely. Beyond initial support, they are expected to gradually become financially

independent (Jungwirth, Grundgreif, & Müller, 2011; Jungwirth & Mueller, 2014). Switching

to private funding is challenging for the cluster management, because members are less

willing to pay for higher membership fees or additional services (Jungwirth et al., 2011).

Regarding decision-making, Fromhold-Eisebith and Eisebith (2005) have highlighted

that, while members can exert some influence on the cluster’s activies and direction of

growth, top-down clusters are centrally coordinated. Public authorities control the

performance of the cluster and assess the achievements of funded activities; cluster

management has to report on the results. Besides, Jungwirth and Mueller (2014) illustrated

how top-down and bottom-up clusters have different governance regimes, which may also

depend on the specificity of the industry. For bottom-up clusters, member firms assign

specific tasks to cluster managers, generally with the aim of increasing firms’

competitiveness and thus fulfilling private goals. For top-down clusters, task allocation lies

with the government and the management team has to fulfill two different goals: increasing

local attractiveness (i.e., a public goal) and improving member firms’ innovation capacity

(i.e., a private goal). Achieving both objectives requires very different strategies and attempts

to pursue them simultaneously may hinder efficient decision-making at the cluster level

(Jungwirth & Mueller, 2014).

Finally, top-down clusters are fairly inclusive. Any organisation can become a

member of the cluster, although administrative and spatial boundaries can set limits on

membership. Fromhold-Eisebith and Eisebith (2005) have argued that often only a few

members are highly motivated and active; the remaining are sitting members who hardly

participate in, or contribute to cluster activities. As a result, top-down clusters do not

necessarily create a very vibrant community because members may be formally part of the

8

cluster, yet lack a sense of belonging to it. Additionally, top-down clusters lack social capital,

are characterised by loose networks, and often have difficulties with developing linkages or

building a shared vision (Andersson et al., 2004; Su & Hung, 2009). In contrast, bottom-up

clusters tend to be characterised by a stronger sense of community and higher degree of

member participation (Beckeman & Skjöldebrand, 2007; Rosenfeld, 2003).

2.2 Cluster policy: creating and steering cluster development

Holding the promise to enhance local competitiveness, cluster policies have been

widely promoted by international bodies, such as the European Union and OECD, and have

proliferated worldwide both at regional and national levels (Benneworth et al., 2003; Sölvell

et al., 2003). By supporting cluster development, policy-makers expect to boost innovation,

raise competitivitiveness, and increase job creation (Feser, 2008; Huggins & Williams, 2011).

Still, given the large amount of government funding invested in cluster policy, the ongoing

debate about the impact of cluster policy is not surprising (Brenner & Schlump, 2011); even

more so given the ambiguous results about the impact of such policy. Evidence from global

surveys and meta-studies has indicated that cluster policy only plays a very minor role in

cluster creation (Enright, 2000; van der Linde, 2003). Nevertheless, some empirical evidence

suggests that government policy can influence cluster development (Brenner & Schlump,

2011; Rodríguez-Pose, 2013), for example, by adopting a brokering role (Roelandt & den

Hertog, 1999) or by creating conditions that facilitate cluster emergence and growth

(Bresnahan et al., 2001; Desrochers & Sautet, 2004). A positive influence unfolds typically

when governments invest in education or infrastructure and set up institutional frameworks

that support entrepreneurship (Feldman & Francis, 2004).

Nevertheless, few studies have rigorously evaluated cluster policy, and its impact is

not well understood (Uyarra & Ramlogan, 2016). Evaluating cluster policy impact is

9

challenging (Fromhold-Eisebith & Eisebith, 2008) and requires the mobilisation of different

methods depending on the purpose and scope of the evaluation (Schmiedeberg, 2010).

Among existing research, there are quantitative studies that try to evaluate the

macroeconomic impacts of cluster policy (Learmonth, Munro, & Swales, 2003) and compare

the innovation performance of firms within a top-down cluster to the performance of firms

not belonging to any cluster (Falck, Heblich, & Kipar, 2010). Several studies uncover

conditions under which cluster policies are successfully enhancing the performance of

member firms and inter-firm relationships (Fromhold-Eisebith & Eisebith, 2008). Nishimura

and Okamuro (2011), for instance, explored the impact of developmental programmes put in

place by a cluster and questioned which kind of cluster-level support programme contributes

to firm performance. Their findings reveal that not every support programme affects firm

performance positively. Firms may have to select the programme best aligned with their

aims. Molina-Morales and Expósito-Langa (2012) argued instead that the ability of cluster

firms to reap benefits from their R&D effort depends on the degree to which they are

involved in the cluster network.

To summarise, then, even though cluster policy is promoted and used widely across

the world (Benneworth et al., 2003; Sölvell et al., 2003), whether governments can indeed

have a meaningful impact on cluster development is still up to debate. No matter if a cluster

has been organised in a top-down or bottom-up manner (Fromhold-Eisebith & Eisebith,

2005), a cluster’s success is contingent on the extent to which various cluster members start

showing agency in terms of participating and collaborating in a cluster (Hallencreutz &

Lundequist, 2003). To leverage members’ agency, cluster policy will have to develop a

cluster that is to some extent self-organising with actively participating members who feel a

sense of ownership over the cluster. Since cluster policy has the underlying aim of pursuing

public goals such as stimulating regional competitiveness and job creation (Huggins &

10

Williams, 2011; Mueller & Jungwirth, 2016), the government also needs to maintain some

level of control to ensure that policy objectives are met. In the remainder of the paper, we

will explore empirically to what extent cluster members value government-supported clusters

and how they reflect on the paradoxical role of the government in such clusters. We thus aim

to provide a cluster-member perspective on whether cluster policy is misguided or not.

3. Research methodology

3.1 Research setting

The empirical setting of this research is Tenerrdis, one of the four French

competitiveness clusters that focus on energy. In the spring of 2004, the French government

initiated a nation-wide cluster policy called “Pôle de Compétitivité”, or competitiveness

cluster in English. The aim was to stimulate interaction between research centres and

universities, on the one hand, and large corporations and small and medium-sized enterprises

(SMEs), on the other. The policy’s purpose was to enhance local competitiveness, economic

growth and job creation (La Documentation Française, 2008). A consortium had to respond

to a public tender and, if selected, would be endorsed by the government as being a

competitiveness cluster (Shong, 2009).

To support competitiveness clusters, the government introduced a specific R&D

funding programme, the “Fonds Unique Interministériel (FUI)”. Funding can only be granted

to collaborative R&D projects that (i) involve both firms and research institutes and (ii) have

received a label from a competitiveness cluster. The government also allocated a specific

budget to financially support cluster management teams, representing about 20% of the total

budget. For the remaining share, clusters are funded via local government resources and

membership fees. The fee is proportional to a firm’s size: start-ups pay a few hundred euros

to join a cluster while the contribution of large firms is slightly more than €20,000. Since

11

their creation, the clusters accompanied the development of more than 1,200 R&D projects,

which accounted for a total public spending exceeding €2.3 billion. Competitiveness clusters

are evaluated every three to five years to decide whether they can maintain their government

endorsement.

Tenerrdis was established in 2005 as a response to the government initiative to set up

competitiveness clusters. It was a joint effort of the Rhône-Alpes region, the Department of

Savoie, a Grenoble-based engineering school, and a national research lab. Some firms that

already played an active role in the region welcomed the initiative and decided to join the

cluster as founding member. The cluster’s foundations are closely connected to those of the

founding members, at least with respect to the focus and governance structure. The cluster

covers six themes that correspond to the research activities of the research lab and the

engineering school: biomass, hydrogen, solar PV (research lab), smart grids, energy

efficiency in buildings (research lab and engineering school), and hydropower (engineering

school). Regarding its governance, a representative from one of the local industrial actors has

chaired the cluster and general managers have either been selected from within the research

lab (until December 2013) or from a major local industrial actor. Since its creation Tenerrdis

has grown significantly from around 50 members at its foundation to 178 members at the

time of our research. The cluster shows great resemblance with top-down clusters that have

been described in the literature (Fromhold-Eisebith & Eisebith, 2005). Even though the

initiative taken by the founders may be seen as bottom-up, for the large majority of the

members that joined later, Tenerrdis is the product of a top-down initiative of the national

government. The top-down nature is reflected in the cluster being regularly evaluated by the

government based on objectives it has decided upon. Nonetheless, Tenerrdis does have some

characteristics of a bottom-up cluster as well because the founding members were already

collaborating on research projects before being formalised as a cluster initiative.

12

In 2013, the French government set out a new objective for the competitiveness

clusters to make sure that funded research projects give rise to marketable products or

services, both in France and internationally. In using the French Ministry’s terminology,

clusters were required to shift their overall mission from managing a “usine à projets”

(factory of projects) to managing a “usine à produits” (a factory of future products). The

government would like to see jobs being created as a result of the significant investment

made into R&D projects. Although support for the development of R&D projects will remain

a core mission, clusters now have to make sure that funded projects lead to the creation of

new products or services. Of importance to reach this objective is the specific support that

clusters should give to SMEs to facilitate their access to external funds and

internationalisation (Andersson, Evers, & Griot, 2013), and to allow them to better anticipate

needs for new competences. These changes in the mission of the clusters are reflected in the

objectives stated in the “performance contract” signed between each cluster and the (local)

government(s). At the time of the data collection no transition had happened yet, but the

cluster management team and member firms were already reflecting on the on-going changes.

3.2 Data collection and analysis

To explore how members value government-supported clusters and perceive the

government’s role in their further development, we rely on interview as well as archival data

sources. In line with the exploratory nature of the research, data collection followed a loose

timeline which overlapped with the data analysis, a common feature of theory building that

draws on a case study (Eisenhardt, 1989). This approach enriches the validity of the data and

facilitates the adjustment of objectives that are pre-established in a deductive way with

elements that are identified later on, according to an inductive logic. The project started at the

outset of 2014 with the collection of archival data about the establishment of the

13

competitiveness clusters. In particular, we gathered documents about cluster policies of the

French government and benchmark reports about the performance of existing French clusters.

These were complemented by data about the specific cluster our research focused on,

including results of a survey conducted to assess members’ satisfaction as well as the latest

performance contract signed between the cluster and the French Government. Finally, we

analysed how the specifications written in the calls for FUI projects evolved over time.

Next, we conducted semi-structured interviews with 20 members of the cluster. A

representative set of cluster members were pre-selected jointly with the cluster management

team by taking account of the following criteria: size, date of joining the cluster, member

firms’ engagement in activities organised by the cluster, and business focus. Three members

that act as technology gatekeepers – two incumbent firms and the national research lab –

were included in the selection (Hervas-Oliver & Albors-Garrigos, 2014). With these criteria

we sought to prevent a bias in the findings. We also conducted semi-structured face-to-face

interviews with cluster management team representatives, including the general manager, the

innovation and partnership coordinator, and the business development and IT project

manager. Interviews were conducted between July 2014 and December 2014 and lasted 60 to

90 minutes (see Table 1 for details about the interviews). Interviews were transcribed

verbatim and analysed using the Dedoose data analysis platform. Transcripts were sent to all

interviewees to verify these were consistent and to obtain their consent.

------TABLE 1 ABOUT HERE------

The data analysis followed a thematic approach and tried to unpack two main themes:

(i) in what ways do members participate in the cluster and to what extent is the cluster self-

organising and (ii) how do members view the government’s intention to manage the cluster

14

from the top-down by steering it in a new direction? In other words, we aimed to capture the

experiences and the perceptions of cluster members as the new policy was being

implemented. We carefully read and re-read our interview transcripts to define and agree

upon a coding structure that we would use for analysing the interview data. Consultation

between the three researchers was essential at this stage. To address the first theme, we

structured codes around four aspects: 1) the extent to which members take part in the various

activities organised by the cluster; 2) whether they are involved in strategic decision-making

about the cluster’s development; 3) whether they consider themselves as receiving from

and/or contributing to the cluster; and 4) whether they feel a sense of belonging to the cluster.

These aspects gave us insight into whose cluster it actually is, to what extent the cluster is

vibrant, and whether it is self-organising.

To address the second theme, we structured the codes in a way that would capture

how members react to the attempt of the government to steer the cluster’s development. We

conducted our analysis in two steps. First, we coded whether members agree or disagree with

the initiative. This allowed us to appreciate how they feel about the attempt of the

government to change the direction of the cluster. Second, we focused on understanding what

members do, or intend to do, as a response to the attempt of the government to change the

cluster’s direction. Our code list included the different attitudes that members have regarding

the new government objectives. We coded, for instance, whether they ignore, (passively)

resist, embrace or opportunistically use the government’s attempt to change the cluster’s

direction. This gave us additional insights about whether members perceive the government

as hindering their chance to self-organise or as facilitating their participation instead. We

used the archival data to triangulate the analysis of the interview data (Yin, 2009).

15

4 Findings

This section introduces the main findings of our analysis. In Section 4.1 the focus will

be on investigating whether cluster members actively participate to the cluster’s activities, in

what way they value being part of a government-supported cluster, and to what extent the

cluster appears vibrant and self-organising. Section 4.2 will address whether members

perceive the government as facilitating or hindering their capacity to self-organise. Based on

these findings, the discussion in Section 5 will explore cluster members’ views on the value

of government-supported clusters and whether and/or how governments can foster cluster

development.

4.1 Members’ participation to the life of the cluster

First, we were interested in understanding to what extent members participate in the

activities of the cluster and as a result whether the cluster has the capacity to self-organise.

We found considerable diversity in our sample. On one end of the spectrum, we found that

the national government lab and the two large incumbent firms (i.e., the technology

gatekeepers) are actively participating in the cluster. They have employees dedicated to

representing them in the cluster and attending many of the events that are organised. We

could observe that these large members are well aware of the extent to which they contribute

to the cluster. They regularly contribute to cluster events by sharing their knowledge with

other members and they offer free support for various cluster activities such as providing

feedback on R&D project proposals submitted to obtain the cluster’s label. These technology

gatekeepers are also aware of being among the major funders. For the research lab this

awareness went as far as them stating that: “if there was a large cluster focused on new

energy technologies with a [research lab] that would say ‘We look elsewhere, that’s not

interesting for us’, it would simply not be credible.” As members of the Board of

16

Administrators, the gatekeepers have the possibility to “give [their] opinion about the

orientation […] and the choices made by the cluster.” They may propose themes that should

be included in the cluster and thereby influence the cluster’s path of development in a major

way. One of the incumbent firms, for instance, explained that they “accompanied the

evolution of the themes covered by Tenerrdis. […] The topic of energy efficiency for instance

is something that we integrated in Tenerrdis, together with the smart grid.” This clearly

shows that large members play an important role in steering the cluster’s direction.

The dominance of the larger members does not mean that smaller members are

completely left out of the process; some of the latter also proactively participate in the cluster

activities. These smaller members contribute to the cluster dynamics by proposing specific

initiatives, contributing to the organisation of events, and regularly sharing their expertise

(e.g., knowledge of specific markets and experience in setting up collaborative projects).

Nevertheless, the smaller members are conscious that they are too small to dedicate much

time and resources to the cluster, which weakens their belief as to whether they can steer the

cluster’s direction. This tension is clearly highlighted in the way the following member

describes their participation: “We are not just active in the sense that we participate in things

that are organised. We are also a driving force because […] we push projects forward”.

However, the interviewee also drew attention to some constraints they were facing: “we are

ten people, so we can’t afford to be present at each event. […] As I said, when we are small

we have difficulties to have someone dedicated to these aspects.”

In the midst of the spectrum, we found firms that participated in the cluster as

opportunistic receivers without contributing much to the cluster. This was particularly the

case for start-ups that mainly use their membership to benefit from the services offered by the

cluster (e.g., help in obtaining a bank loan or advice about intellectual property management).

They may participate in some events, if they think these have a positive impact on their

17

business or research objectives. However, they would not propose activities or share their

knowledge with others. One start-up, for instance, explained: “till now, besides getting

support to obtain a loan […] I clearly did not take enough advantage of Tenerrdis’ network.”

Here, we also find SMEs that are not yet part of the local network and participate in events to

gain visibility, build a network (e.g., discuss shared objectives), or gain legitimacy (e.g.,

those interested in entering the energy business and see the cluster as an opportunity to meet

relevant actors). An SME interested in offering energy services for example stated: “if you

have developed software that helps to optimise energetic gains backed by a cluster with a

strong image in energy […] you can certainly become more visible and commercialise it

efficiently.” While quite a few members already collaborated before Tenerrdis came into

being (i.e., especially the founding members), the formal nature of the competitiveness

cluster allowed firms in the region not yet part of the informal network to also access it.

On the other end of the spectrum, we found firms that hardly participate in the cluster

for various reasons. First, some members do not have a genuine interest in the cluster apart

from obtaining the label (i.e., providing access to FUI funding), or using it as a marketing

tool to gain visibility, especially in the eyes of potential stakeholders such as non-member

firms. Firms in this group do not take part in any of the collective initiatives. Several firms

even mentioned that they became member “by accident”. For these firms, being involved in a

research project with other organisations already member of the cluster meant that they felt

the pressure to also join the cluster. The following quote exemplifies some members’

apparent lack of interest: “It allowed me to get the label I needed to present my project in

front of the government. Beyond that […], it is difficult to pinpoint the added value [of

Tenerrdis].”

Second, we found start-ups that became part of the cluster because it represented

another venue for them to meet firms already part of their network. Not joining the cluster

18

might thus have been perceived as awkward. The following member explained that they

joined the cluster because “the president […] was from [a local industrial firm], a friend of

mine. I already knew a lot of people that were members here and there.” Despite the strong

motivation for committing to the cluster, their daily routine would leave them little time to

actively contribute to the cluster’s activities. As one interviewee put it: “I really don’t have

much time to dedicate [to Tenerrdis].” Hence, being part of the informal network in the

region pushed these firms into becoming an official member of the cluster, even if the

benefits of doing so were not obvious.

Third, some firms felt that their business focus was only vaguely overlapping with

that of the cluster. By seeing other cluster members as distant, these firms gave the

impression that they did not feel they belong to the cluster. For example, one member stated:

“we don't take advantage of this because the topics managed by Tenerrdis are not very much

in line with our current business.” Fourth, we also found firms that were initially

participating actively in the cluster but eventually reduced their involvement because they

experienced a misfit between their expectations when joining the cluster and the possibility to

benefit from their membership. In this group, we found small members for which the cluster

did not generate any concrete benefit despite the hope of finding potential customers. This

mismatch is clearly exemplified in the way the following member talks about his experience

in the cluster: “as of today - it’s been three years we are member of Tenerrdis – I spent many

hours having fascinating conversations. Generated turnover: zero!” Among the main

critique they had on the cluster, interviewees described the cluster as having been made “by

researchers for researchers”. That is, they criticised the cluster for paying less attention to

more applied research projects like their own.

A significant share of the members in our sample also felt that the process of

participating in the cluster was not as straightforward as one might expect. Smaller firms

19

expressed the intention to be more active, yet their willingness to participate in the cluster

was discouraged by the presence of larger, technology gatekeeper-type firms playing a

deterring role. On the one hand, smaller firms are in direct competition with some of the

incumbents and, as a result, are given very little visibility in the cluster. One member for

instance stated: “this large firm is everywhere, so we are overshadowed.” On the other hand,

incumbent and small firms may not have a constructive relationship, which leaves hardly

room for the latter to benefit from the services or network offered by the cluster. One of the

smaller firms, for example, explained being “blacklisted by the [large firm], which means

that they would direct potential clients to competitors providing a service similar to ours.” As

a result, the engagement of some small firms within the cluster is limited to attending some

cluster meetings only. Moreover, we found that geographic and mental distance may act as

deterring factors hindering members to fully participate. Some interviewees who openly

criticised Tenerrdis for being very ‘Grenoblois’ felt that not being part of the local, informal

network that pre-dated the formal competitiveness cluster made it hard to participate in and

contribute to the cluster activities.

In summary, our findings show that a few members – mostly the technology

gatekeepers and some SMEs – actively participate in the cluster’s activities and have made

the cluster their own. However, many members do not participate or do so only

opportunistically. The presence of many sitting members suggests that the cluster may not

have enough members actively involved to be able to fully self-organise. Nevertheless, there

is a clear drive from members that do participate to shape the cluster in ways that matches

their interests both in terms of content and regarding the types of events that are organised.

20

4.2 Members’ perspectives on the government as a facilitator or hindrance

In the previous section, we presented evidence of the different degrees by which

members participate in the cluster’s activities. This provides us with initial insights on whose

cluster it actually is and whether the cluster is starting to stand on its own legs. In this section,

we question the role of the government as it tries to steer the cluster development in a new

direction. As explained in Section 3.1, the French government has required competitiveness

clusters to make a transition from “usine à projets” (factory of projects) to “usine à produits”

(factory of future products). Here, we analyse how members perceive the government’s

attempt to make this transition in having the cluster function in line with the new policy

objectives. This sheds light on whether cluster members perceive the government as

hindering or facilitating their capacity to self-organise and use the cluster to achieve their

own private goals.

Our analysis reveals that members hold very different perspectives on the role of the

government as it tries to steer the cluster in a new direction. A few members did not express

any opinion in favour or against the attempt of the government to assign new objectives to the

cluster. Here, we mostly see members who joined the cluster to obtain the label. They neither

participate in the cluster’s activities nor have an opinion about the role of the government.

However, there are also quite a few members that are sceptical about the feasibility of the

government plan. These members do not perceive the government as having much influence

on the cluster’s path of development and thus also not the cluster members’ capacity to self-

organise. In this group, we found incumbent firms that do not believe that the government

can push the cluster to facilitate the commercialisation of products that are created via

subsidised R&D projects because “they don’t have the [required] competence; they will

never have it.” Engaging in activities aimed at the commercialisation of their output would

require specific knowledge (e.g., about the market) and competences (e.g., marketing), which

21

the cluster currently lacks. They doubt that the cluster could organise training or conduct

studies and charge members for it. Two smaller members also shared the scepticism about the

government’s ability to steer the cluster towards commercialisation. One small firm in the

sample thought it would be interesting if the cluster could help them commercialise their

product, yet the interviewee could hardly envision how this might take place in practice. He

stated: “It is interesting but I really don’t see how they could help us.” Another interviewee

went as far as stating: “are there really projects that lead to solutions that are marketable? I

don’t have the impression”. This member questioned whether the cluster could ever support

the emergence of collaborative R&D projects that would result in delivering marketable

products.

Such concerns show how some members simply do not believe that the government

could actually transform the nature of the cluster. In some of the interviewees’ opinion, the

cluster policy has not been properly designed to allow the government to play an influential

role. Although the government has spelled out new objectives for the clusters, these

objectives have not been equipped with additional leveraging power. As one of the

incumbents argued: “I think that the mission of the clusters is complicated because they are

asking to generate real products without having a large budget, simply by creating

incentives.” The need to meet the research objectives set out by FUI still remains the main

driver of clusters and their members. In fact, the main change in the FUI research calls

brought about recently consists of reducing the time for project commercialisation from five

to “three years, with exceptions”.

Furthermore, about one third of the sampled members, including small members and

the research lab, outright disagree with the new policy objectives. As stated by one

interviewee: “they would lose their federalising nature if they started getting involved in

being a Usine à Produits”. According to them, clusters have been created to foster

22

cooperation and networking opportunities between firms and research institutions and should

continue doing so. The following quote exemplifies this perspective: “it is maybe not their

role. For me their role is more to develop partnerships.” More generally, we realised that

many interviewees took the function of the cluster as a breeding ground for collaborative

research projects mostly for granted. Members’ requests to receive support for the

commercialisation of their technologies was often ill received or not well understood. This

particularly concerned firms in which it is mainly the R&D department that maintains the

relations with the cluster. These members resist the idea of a cluster bearing a purpose

different from the one set out at the cluster’s foundation (i.e., a focus on research). Of interest

is the fact that many of these sceptical members were among the most active contributors to

the cluster’s activities.

Focusing on how they react to the attempt of the government to steer the cluster, we

could identify two responses. First, the research lab tries to leverage its power as a founding

member to make sure that the new government objectives do not take the cluster away from

its original mission. The following quote is exemplary in this regard: “The [clusters] had a

meaning when they were seen as factories of projects. Now they are more factories of

products. But we should not stop the logic of the factories of projects. I have a strong feeling

about this and it is the position I defend when I go to the board of Tenerrdis.” This shows

that one of the technology gatekeepers perceives the government as hindering the cluster’s

capacity to self-organise and internal lobbying is considered to minimise the impact of the

new government objectives. Second, the interviews also revealed that by shifting the

objectives of the cluster towards commercialisation, the government might actually

discourage some members from maintaining their level of participation with the cluster’s

current activities. One member for instance states: “Help commercialising, for me, it is not

23

the role of a cluster. I did not understand it like this. If the status has changed okay that’s

very good. But it moves further away from our interest.”

Still, nearly half of the interviewees responded positively to the attempt of the

government to impose new objectives onto the cluster. While firms generally agreed that the

cluster should not play a direct role in their commercialisation activities, they would welcome

a cluster that could help them reach the market. The incumbents, despite their overall

scepticism, see in the policy an opportunity to develop business relations with promising

SMEs and implement their open innovation strategy. As one incumbent explained: “Suppose

that one SME from Tenerrdis develops a great product in the energy sector. Clearly this SME

will have direct access to our company to go and present their idea. If it is worthwhile, we

can even invite them to come and talk to us before it is confirmed on the public place because

we would have an interest in doing that.” For large members the new policy objectives could

thus be an opportunity to strengthen their position in the cluster, in particular because they

provide other cluster members with new market opportunities. As stressed by a small

member, after all “the best way for large firms that are part of a cluster to help other

members is to give them an order”.

For some of the smaller members the new policy objectives also create an opportunity

to participate more in the cluster’s activities. In this category we find many members that

hardly participate now, because they have not found what they were looking for. If the new

objectives better mirror their expectations it may motivate them to participate more regularly

in the cluster activities. As the following quote shows, the new objectives may better mirror

some of the members’ expectations: “I only have two arms, two legs and one brain. I cannot

be everywhere, and I need to make choices depending on […] my availabilities and the return

I can expect in terms of business opportunities.” These members seemed to welcome the new

objectives because the new direction would allow them to move away from their more

24

marginalised position, as they are not part of the more informal networks in the cluster that

have formed around research activities. Of those that respond positively to the new policy

objectives, we found members able to translate the objectives into something meaningful for

them. By doing so, these members enact the government’s objectives and make sure that the

output of this process (e.g., organisation of activities or events) fits their own interests.

In summary, our findings show no consensus on whether the proactive approach of

the government to steer the cluster’s development is helping or hindering members to self-

organise. The perspective of members depends on their own private objectives for being part

of the cluster. While there is clearly much resistance against moving away from the research

focus, almost half of all members interviewed react to the new policy objectives more

positively. Notably, the members resisting the new policy objectives most vehemently were

also among those currently most active in the cluster. Not surprisingly, the national research

lab expressed unease with a new direction focused on commercialisation instead of basic

research. The two other technology gatekeepers took a more pragmatic approach, as the new

purpose of the cluster could fit their open innovation strategy. Finally, for some of the smaller

members the new purpose provides a window of opportunity to move into a more central

position in the cluster.

5 Discussion

5.1 Discussion of the findings

Scholars have previously addressed the paradoxical role of the government in cluster

development (Fromhold-Eisebith & Eisebith, 2008; Menu, 2012; Sydow, Lerch, Huxham, &

Hibbert, 2011; Zagorsek, Svetina, & Jaklič, 2008). In this paper, we sought to explore this

paradoxical role from the perspective of cluster members. We explored to what extent cluster

members value government-supported clusters and how they perceive the government’s

25

seemingly paradoxical role in the development of a cluster. We thus tackle a fundamental

question of whether members start seeing a cluster as a vibrant community to which they feel

a sense of belonging, even when the government has played a pivotal role in setting it up.

Our findings reveal that many cluster members value a government-supported cluster

but, reflecting Perry’s (2007) findings, differ in what they see as perceived benefits for

different reasons related to their size and strategic objectives. Perry (2007) argued that

‘[c]lusters are not automatically a win-win opportunity for all participants.’ Likewise,

Lechner and Leyronas (2012) showed that the competitive advantage derived from cluster

membership is firm-specific because it also depends on firms’ embeddedness in networks

beyond the cluster. Our study provided support for these observations by showing how a

power imbalance between large and small members affected the perceived benefits of the

cluster for each of them. The large members were among the founders and thus consider

themselves as an integral part of the cluster. They also made sure that the cluster’s mission

was well aligned with their own objectives, so they are more likely to benefit from their

participation. While these members might be the cluster’s technology gatekeepers (Hervas-

Oliver & Albors-Garrigos, 2014), they do not depend on the cluster solely for resources,

legitimacy, or access to other firms. Their perceived benefit is in gaining access to local

players in the region for collaborative R&D projects, but their size gives them more power in

the cluster, not only because their voice is heard more clearly, but also because they are less

dependent on this cluster alone for resources than is the case for some of the smaller members

(Casciaro & Piskorski, 2005; Lechner & Leyronas, 2012; Pfeffer & Salancik, 1978).

Smaller members might depend much more on participating in this specific cluster to

get access to local actors, to obtain government funding, or to be seen as a legitimate partner

by being associated with the cluster (Lublinski, 2003; Malmberg & Maskell, 2002; Porter,

1990; Wolfe & Gertler, 2004). The findings suggest, however, that the top-down nature of the

26

cluster has created many ‘sitting members’ for whom the cluster seems to have very limited

benefits in practice (Fromhold-Eisebith & Eisebith, 2005). While there might be more

potential benefits for smaller members whose strategic objectives fit the cluster’s mission,

ironically, they seem hesitant to use this potential due to their relatively high dependence on

the cluster for resources and legitimacy. Unlike the incumbents, who can spread their effort

across clusters, the small members have a limited geographical scope which increases their

dependence on a cluster in their own region (Perry, 2007). As a result, the small members

feel that they might not fully benefit from their participation in the cluster due to a power

imbalance vis-à-vis the technology gatekeepers which could make them vulnerable to

misappropriation by these larger members (Casciaro & Piskorski, 2005; Katila, Rosenberger,

& Eisenhardt, 2008). The proposed shift of the cluster's focus from research to

commercialisation might potentially aggravate this problem because members would more

directly compete with each other (Clarysse et al., 2014). For quite a few other small

members, the perceived benefits of participating in the cluster are low, just because their

strategic objectives do not have a good fit with the cluster’s mission. These members only use

the cluster opportunistically for the label to get access to government funding or because they

happen to be in the region anyway. Yet, their dependence on the cluster is limited and they do

feel threatened by the power of the incumbents as some of the others do.

Although there are many sitting members (Fromhold-Eisebith & Eisebith, 2005), the

findings still indicate that government-supported clusters can self-organise to some extent.

Consistent with Ingstrup and Christensen’s (2017) findings, the technology gatekeepers and,

in our case also some smaller members, actively participate in the cluster and shape its

functioning and activities. The case suggests that a cluster is more likely to self-organise if

members are given the opportunity to proactively participate in the strategic development of

the cluster and can propose initiatives that align with their needs or those of the local

27

community. Our findings also reveal that cluster members’ perception about the role of the

government as facilitating or hindering their capacity to self-organise depends on the private

objectives that they pursue (Lechner & Leyronas, 2012). Those who embrace the initial

mission of the cluster and already participate actively see the new objectives as a hindrance

and would prefer a government that does not attempt to shape the direction of the cluster

around new objectives. Yet, some members see clear opportunities in the new policy

objectives. The incumbent firms seem to have the pragmatism to translate the new policy

objectives into something meaningful for themselves to strengthen their position as core

cluster members. For those that do not participate that actively yet, the new objectives

facilitate a renewed interest from their side and more active future participation.

5.2 Implications for theory and policy

Existing research has explored the impact of cluster policy of governments on the

development of clusters, showing rather ambiguous results. Some studies have found a

positive impact (Brenner & Schlump, 2011; Islankina & Thurner, 2018; Rodríguez-Pose,

2013), while others concluded that cluster policy only plays a rather minor role (Enright,

2000; van der Linde, 2003). Our paper contributes to this line of research, not by assessing

policy impact directly, but rather by providing an in-depth investigation of how members of a

cluster reflect on the influence of the government. By taking a cluster member perspective

(Perry, 2007), we sought to gain more insight into to what extent members value government-

supported clusters and whether they believe the government can have a genuine impact on

cluster development. We were specifically interested in understanding how members reflect

on the dilemma that the government seems to face of having to strike a balance between

controlling and letting go of a cluster initiative (Mueller & Jungwirth, 2016).

28

Our research could be regarded as a follow-up of the study of Menu (2012) who

analysed the French cluster policy of the “Pôles de Compétitivité” a year after it was

launched. She argued that the way these clusters had been set up did not support cluster

members to take the lead in cluster development. One major reason was that the hierarchical

nature of these clusters did not allow members to self-organise (cf. Bell et al., 2009). Our

research complements Menu (2012) by investigating the effectiveness of the policy eight

years after its launch. The findings indicated that there now is a group of members that

actively takes part in the cluster activities and have managed to self-organise to some extent.

At first sight, these active members do not seem to be much disturbed by the hierarchical

governance of the competitiveness clusters, as they have built their collaborative efforts on

strong relations, some of which pre-date the cluster (Bell et al., 2009). This seems to bode

well for clusters that came into existence in a top-down manner; that is, even if many cluster

members were not part of its foundation, they can still develop a sense of ownership over the

cluster.

However, as Bell and colleagues (2009) argue, the co-existence of hierarchical and

relational governance also seems cause for conflict in a cluster. Here, the co-existing

governance structure seems to hinder the government’s attempts to change the direction of

the cluster. While the cluster policy might have imposed hierarchical governance onto the

region to organise collaborative research activities, once a cluster matures more relational

forms of governance develop due to an increase in trust between the members (Mesquita,

2007). In the specific case studied in this paper, relational governance was even stronger as

many of the actors in the region were already collaborating before the cluster was formally

created as a cluster initiative. As a consequence, once a cluster matures there seems to be a

blurring of the line between a top-down and a bottom-up cluster (Fromhold-Eisebith &

Eisebith, 2005; Jungwirth & Mueller, 2014; Su & Hung, 2009). It is this blurring line that

29

causes the paradoxical role of the government, i.e., of trying to create a vibrant, self-

organising cluster that would at the same time follow ‘government orders’, to be particularly

problematic. Our case suggests that the government has trouble overruling the relational

governance that has come to dominate the cluster over time with its own hierarchical

governance and therefore cannot steer the cluster in the direction it desires. Since the

relational governance developed around collaborative research projects, replacing this

research focus by a commercialisation focus appears a difficult mission.

The case of Tenerrdis also draws attention to the double-edged sword of government-

driven regional policies. Whilst originally set up with the best of intentions – that is, fostering

economic and technological progress often through the promotion of initiatives that may

attract different types of actors – by trying to keep using the cluster as a policy tool, the

government may create a cluster that cannot stand on its own feet. As the government tries to

impose a new direction, it discourages certain cluster members to maintain their level of

participation. In turn, the new direction may encourage others from partaking in the cluster’s

activities. However, if the policy is not implemented or ends up not working well, it also runs

the risk of losing active members thereby losing cohesion.

Mueller and Jungwirth (2016: 438) argued that ‘[w]hile the government may set the

framework conditions, the cluster facilitators need to be free to take cluster-specific decisions

within these conditions in order to adapt the cluster goals and activities to the firms’ needs.’

Our findings lend support to this argument; it does not seem possible for the government to

take full control of the cluster and its development, thus questioning the future of cluster

policies. One might argue that cluster policy is becoming outdated because large firms have

found other ways to build connections with innovative start-ups. The two incumbent firms in

our sample hinted at these developments by seeing the cluster as part of their open innovation

strategy (Chesbrough, 2006). Large firms might no longer need the government to create the

30

structures that bring together different types of research-intensive organisations, they now do

this themselves by creating their own innovation ecosystems (Iansiti & Levien, 2004). While

large firms’ creation of innovation ecosystems might replace the need for cluster policy,

government policy could continue to play a complementary role in structuring regional

networks (Clarysse et al., 2014). Brenner and Schlump (2011) argue, for example, that the

needs of clusters change depend on the cluster life cycle. Different policy measures may be

adequate in different stages of cluster life cycle. This paper shows that depending on the stage

of the cluster life cycle, government may be more, or less capable to implement policies to

steer the cluster. As the cluster matures and members are given the opportunity to self-

organise, they take ownership of the cluster and it becomes increasingly difficult for the

government to mould it. Nevertheless, this opportunity to self-organise does not mean that

the government cannot exert any influence at all. Our findings suggest that the government

may become influential if it is able to leverage the power of influential cluster members

(Clarysse et al., 2014), i.e., the technology gatekeepers (Hervas-Oliver & Albors-Garrigos,

2014).

Interestingly, this scenario raises a new dilemma for the government: Which

objectives should be set out and what would be the most effective way to implement them?

While it is not the aim of this research to propose a solution, the analysis above brought to the

fore that, as cluster members start to self-organise, for the government to be able to continue

using the cluster as a policy instrument, it should design policy that allows for some

‘interpretative flexibility’ (Pinch & Bijker, 1984; Smith & Kern, 2009). Building in

interpretative flexibility increases the possibility for technology gatekeepers to translate new

policy objectives into something meaningful for them. Accordingly, the probability that the

policy initiatives will be implemented as intended and the cluster changes in the desired

direction increases. This argument also suggests that it would be difficult to effectively

31

implement policy initiatives that are not well aligned with the motivations that originally led

technology gatekeepers to join and actively contribute to the cluster (Hervas-Oliver &

Albors-Garrigos, 2014).

6 Conclusions

This paper investigated to what extent a government can use cluster policy to create or

aid the further development of a vibrant cluster. We argued that governments have a

seemingly paradoxical role in cluster development due to the need of continuously striking a

deal between a top-down, interventionist approach (i.e., using the cluster as a policy tool) and

a bottom-up, laissez faire approach (i.e., let cluster members use the cluster to self-organise

their collaborations). This research thus questioned how the government, in implementing

cluster policy, could guarantee, on the one hand, that the cluster is run effectively and meets

policy objectives, and, on the other hand, that members feel part of it so that they actively

contribute to the cluster activities and make the cluster into a vibrant community. Based on

findings of a case study, we argued that it seems difficult for the government to use a cluster

as a policy instrument once it has started to function as a self-organising entity. Nonetheless,

the government could try to change a cluster’s direction indirectly by leveraging the role of

technology gatekeepers, who have a more central position within the cluster (Hervas-Oliver

& Albors-Garrigos, 2014). Future research may explore how this process would unfold and

further question the role of the government in top-down clusters.

References

Andersson, S., Evers, N., & Griot, C. 2013. Local and international networks in small firm internationalization: cases from the Rhône-Alpes medical technology regional cluster. Entrepreneurship & Regional Development, 25(9-10): 867-888.

Andersson, T., Serger, S. S., Soervik, J., & Hansson, W. E. 2004. Cluster Policies Whitebook: International Organisation for Knowledge Economy and Enterprise Development (IKED).

32

Baglieri, D., Cinici, M. C., & Mangematin, V. 2012. Rejuvenating clusters with ‘sleeping anchors’: The case of nanoclusters. Technovation, 32(3): 245-256.

Baptista, R., & Swann, P. 1998. Do firms in clusters innovate more? Research Policy, 27(5): 525-540.

Becattini, G. 1991. Italian industrial districts: problems and perspectives. International Studies of Management & Organization, 21(1): 83-90.

Beckeman, M., & Skjöldebrand, C. 2007. Clusters/networks promote food innovations. Journal of Food Engineering, 79(4): 1418-1425.

Bell, S. J., Tracey, P., & Heide, J. B. 2009. The organization of regional clusters. Academy of Management Review, 34(4): 623-642.

Benneworth, P., Danson, M., Raines, P., & Whittam, G. 2003. Confusing clusters? Making sense of the cluster approach in theory and practice. European Planning Studies, 11(5): 511-520.

Brenner, T., & Schlump, C. 2011. Policy measures and their effects in the different phases of the cluster life cycle. Regional Studies, 45(10): 1363-1386.

Bresnahan, T., Gambardella, A., & Saxenian, A. 2001. 'Old Economy’ inputs for ‘New Economy’ outcomes: cluster formation in the New Silicon Valleys. Industrial and Corporate Change, 10(4): 835-860.

Broekel, T., Fornahl, D., & Morrison, A. 2015. Another cluster premium: innovation subsidies and R&D collaboration networks. Research Policy, 44(8): 1431-1444.

Brusco, S. 1982. The Emilian model: productive decentralisation and social integration. Cambridge Journal of Economics, 6(2): 167-184.

Casciaro, T., & Piskorski, M. J. 2005. Power imbalance, mutual dependence, and constraint absorption: A closer look at resource dependence theory. Administrative Science Quarterly, 50: 167-199.

Chesbrough, H. W. 2006. Open innovation: The new imperative for creating and profiting from technology. Boston, MA: Harvard Business School Press.

Clarysse, B., Wright, M., Bruneel, J., & Mahajan, A. 2014. Creating value in ecosystems: Crossing the chasm between knowledge and business ecosystems. Research Policy, 43(7): 1164-1176.

Desrochers, P., & Sautet, F. 2004. Cluster-based economic strategy, facilitation policy and the market process. Review of Austrian Economics, 17(2-3).

Eisenhardt, K. M. 1989. Building theories from case study research. Academy of Management Review, 14(4): 532-550.

Enright, M. J. 2000. Survey on the characterization of regional clusters: Initial Results. Hong Kong: Institute of Economic Policy and Business Strategy, Competitiveness Program, University of Hong Kong.

Falck, O., Heblich, S., & Kipar, S. 2010. Industrial innovation: direct evidence from a cluster-oriented policy. Regional Science and Urban Economics, 40(6): 574-582.

Feldman, M. P., & Francis, J. L. 2004. Homegrown solutions: Fostering cluster formation. Economic Development Quarterly, 18(2): 127-137.

Feser, E. 2008. On building clusters versus leveraging synergies in the design of innovation policy for developing economies. In U. Blien, & G. Maier (Eds.), The Economics of Regional Clusters: Networks, Technology and Policy: 177-199. Massachusetts: Wiley.

Formica, P. 2003. Corporate governance of cluster development agencies: the case for market orientation. In J. Bröcker, D. Dohse, & R. Soltwedel (Eds.), Innovation Clusters and Interregional Competition: 241–271. Berlin: Springer.

33

Fromhold-Eisebith, M., & Eisebith, G. 2005. How to institutionalize innovative clusters? Comparing explicit top-down and implicit bottom-up approaches. Research Policy, 34(8): 1250-1268.

Fromhold-Eisebith, M., & Eisebith, G. 2008. Looking behind facades: Evaluating effects of (automotive) cluster promotion. Regional Studies, 42(10): 1343-1356.

Guinet, J., 2003. Drivers of economic growth: the role of innovative clusters. In: Br¨ocker, J., et al. (Eds.), Innovation Clusters and Interregional Competition. Springer, Berlin, pp. 150–160. 2003. Drivers of economic growth: the role of innovative clusters. In J. Bröcker, D. Dohse, & R. Soltwedel (Eds.), Innovation Clusters and Interregional Competition: 150-160. Berlin: Springer.

Hallencreutz, D., & Lundequist, P. 2003. Spatial clustering and the potential for policy practice: experiences from cluster-building processes in Sweden. European Planning Studies, 11(5): 533-547.

Hervas-Oliver, J.-L., & Albors-Garrigos, J. 2014. Are technology gatekeepers renewing clusters? Understanding gatekeepers and their dynamics across cluster life cycles. Entrepreneurship & Regional Development, 26(5-6): 431-452.

Hospers, G.-J., Desrochers, P., & Sautet, F. 2009. The next Silicon Valley? On the relationship between geographical clustering and public policy. International Entrepreneurship and Management Journal, 5(3): 285-299.

Huggins, R., & Williams, N. 2011. Entrepreneurship and regional competitiveness: The role and progression of policy. Entrepreneurship & Regional Development, 23(9-10): 907-932.

Iammarino, S., & McCann, P. 2006. The structure and evolution of industrial clusters: Transactions, technology and knowledge spillovers. Research Policy, 35(7): 1018-1036.

Iansiti, M., & Levien, R. 2004. The keystone advantage: what the new dynamics of business ecosystems mean for strategy, innovation, and sustainability. Boston, MA: Harvard Business School Press.

Ingstrup, M. B., & Christensen, P. R. 2017. Transformation of cluster specialization in the wake of globalization. Entrepreneurship & Regional Development, 29(5-6): 500-516.

Islankina, E., & Thurner, T. W. 2018. Internationalization of cluster initiatives in Russia: empirical evidence. Entrepreneurship & Regional Development: 1-24.

Jungwirth, C., Grundgreif, D., & Müller, E. 2011. How to turn public networks into clubs? The challenge of being a cluster manager. International Journal of Entrepreneurial Venturing, 3: 262-280.

Jungwirth, C., & Mueller, E. F. 2014. Comparing top-down and bottom-up cluster initiatives from a principal-agent perspective: what we can learn for designing governance regimes. Schmalenbach Business Review, 66: 357-381.

Jungwirth, C., & Müller, E. F. 2014. Comparing top-down and bottom-up cluster initiatives from a principal-agent perspective: What we can learn for designing governance regimes. Schmalenbach Business Review, 66(3): 357-381.

Katila, R., Rosenberger, J. D., & Eisenhardt, K. M. 2008. Swimming with Sharks: Technology Ventures, Defense Mechanisms and Corporate Relationships. Administrative Science Quarterly, 53(2): 295-332.

Kowalski, A. M., & Marcinkowski, A. 2014. Clusters versus cluster initiatives, with focus on the ICT sector in Poland. European Planning Studies, 22(1): 20-45.

La Documentation Française. 2008. La politique des pôles de compétitivité 2005-2008.

34

Learmonth, D., Munro, A., & Swales, J. K. 2003. Multi-sectoral cluster modelling: the evaluation of Scottish Enterprise cluster policy. European Planning Studies, 11(5): 567-584.

Lechner, C., & Leyronas, C. 2012. The competitive advantage of cluster firms: the priority of regional network position over extra-regional networks–a study of a French high-tech cluster. Entrepreneurship & Regional Development, 24(5-6): 457-473.

Lublinski, A. E. 2003. Does geographic proximity matter? Evidence from clustered and non-clustered aeronautic firms in Germany. Regional Studies, 37(5): 453-467.

Malmberg, A., & Maskell, P. 2002. The elusive concept of localization economies: towards a knowledge-based theory of spatial clustering. Environment and Planning A, 34(3): 429-449

Martin, R., & Sunley, P. 2003. Deconstructing clusters: chaotic concept or policy panacea? Journal of Economic Geography, 3(1): 5-35.

McDonald, F., Tsagdis, D., & Huang, Q. 2006. The development of industrial clusters and public policy. Entrepreneurship & Regional Development, 18(6): 525-542.

Menu, S. 2012. The role of cluster policy on leadership: evidence from two Pôles de compétitivité. Environment and Planning C: Government and Policy, 30: 816-834.

Mesquita, L. F. 2007. Starting over when the bickering never ends: Rebuilding aggregate trust among clustered firms through trust facilitators. Academy of Management Review, 32(1): 72-91.

Molina-Morales, F. X., & Expósito-Langa, M. 2012. The impact of cluster connectedness on firm innovation: R&D effort and outcomes in the textile industry. Entrepreneurship & Regional Development, 24(7-8): 685-704.

Morrison, A. 2008. Gatekeepers of knowledge within industrial districts: who they are, how they interact. Regional Studies, 42(6): 817-835.

Mueller, E. F., & Jungwirth, C. 2016. What drives the effectiveness of industrial clusters? Exploring the impact of contextual, structural and functioning determinants. Entrepreneurship & Regional Development, 28(5-6): 424-447.

Nishimura, J., & Okamuro, H. 2011. Subsidy and networking: The effects of direct and indirect support programs of the cluster policy. Research Policy, 40(5): 714-727.

OECD. 1999. Boosting Innovation: The Cluster Approach. Paris: Organisation for Economic Co-operation and Development.

OECD. 2001. Innovative Clusters: Drivers of National Innovation Systems. Paris: Organisation for Economic Co-operation and Development.

Palazuelos, M. 2005. Clusters: Myth or realistic ambition for policy-makers? Local Economy, 20(2): 131-140.

Perry, M. 2007. Business environments and cluster attractiveness to managers. Entrepreneurship and Regional Development, 19(1): 1-24.

Pfeffer, J., & Salancik, G. R. 1978. The external control of organizations. A resource dependence perspective. New York: Harper & Row, Publishers.

Pinch, T. J., & Bijker, W. E. 1984. The social construction of facts and artefacts: or how the sociology of Science and the Sociology of Technology might benefit each other. Social Studies of Science, 14(August): 399-441.

Porter, M. E. 1990. The Competitive Advantage of Nations. London: Macmillian.Porter, M. E. 1998. Clusters and the new economics of competitiveness. Harvard Business

Review, 76(November-December): 77–90.Porter, M. E. 2000. Locations, clusters, and company strategy. In G. L. Clark, M. Gertler, &

M. P. Feldman (Eds.), The Oxford handbook of economic geography: 253-274. Oxford: Oxford University Press.

35

Porter, M. E. 2003. The economic performance of regions. Regional studies, 37(6-7): 549-578.

Rodríguez-Pose, A. 2013. Do institutions matter for regional development. Regional Studies, 47(7): 1034-1047.

Roelandt, T. J. A., & den Hertog, P. 1999. Cluster analysis and cluster-based policy-making: the state of the art In T. J. A. Roelandt, & P. den Hertog (Eds.), Boosting Innovation: The Cluster Approach: 413-427. Paris: Organisation for Economic Co-operation and Development (OECD).

Rosenfeld, S. A. 2003. Expanding opportunities: cluster strategies that rech more people and more places 1. European Planning Studies, 11(4): 359-377.

Schiele, H., & Ebner, A. 2013. The role of domestic and international external cluster linkages explored on the example of buyer–supplier relations in learning regions: a cross-functional assessment. European Planning Studies, 21(5): 683-699.

Schmiedeberg, C. 2010. Evaluation of cluster policy: a methodological overview. Evaluation, 16(4): 389-412.

Shong, F. 2009. Le pilotage chemin faisant. Emergence des modes de gouvernance et de pilotage des pôles des competitivité. Université de Paris Dauphine, Paris.

Smith, A., & Kern, F. 2009. The transition storyline in Dutch environmental policy. Environmental Politics, 18(1): 78-98.

Su, Y.-S., & Hung, L.-C. 2009. Spontaneous vs. policy-driven: The origin and evolution of the biotechnology cluster. Technological Forecasting and Social Change, 76(5): 608-619.

Sydow, J., Lerch, F., Huxham, C., & Hibbert, P. 2011. A silent cry for leadership: Organizing for leading (in) clusters. The Leadership Quarterly, 22(2): 328-343.

Sölvell, Ö., Lindqvist, G., & Ketels, C. 2003. The Cluster Initiative Greenbook. Gothenburg: Ivory Tower.

Uyarra, E., & Ramlogan, R. 2016. The effects of cluster policy on innovation. In J. Edler, P. Cunningham, & A. Gok (Eds.), Handbook of Innovation Policy Impact: Edward Elgar.

van den Berg, L., Braun, E., & van Winden, W. 2001. Growth of clusters in European cities: an integral approach. Urban Studies, 38(1): 185-205.

van der Linde, C. 2003. The demography of clusters - findings from the cluster meta-study. In J. Bröcker, D. Dohse, & R. Soltwedel (Eds.), Innovation Clusters and Interregional Competition: 130-149. Berlin: Springer.

Wolfe, D. A., & Gertler, M. 2004. Clusters from the inside and out: local dynamics and global linkages. Urban Studies, 41: 1071-1093.

Yin, R. 2009. Case study research: design and methods (4th ed.). California: Thousand Oaks.

Zagorsek, H., Svetina, A. C., & Jaklič, M. 2008. Leadership in clusters: attributes of effective cluster leader in Slovenia. Transformation in Business & Economics, 7(2): 98-113.

36

Table 1: Details of interviews Firm Size Market Reach Sector Informant Interview

dateLength

Member 1 Large Global Across thematic/Electricity producer

1) Integration director2) Expert biomass

09/07/2014

131 min

Member 2 Large Global Energy Innovation and Partnership Manager

16/10/2014

72 min

member 3 Large Global/Lyon Components R&D Manager 25/06/2014

84 min

Member 4 Medium Global/outside RA Components 1) Sales Engineers2) Sales and Marketing Manager

24/07/2014

51 min

Member 5 Medium Global/Grenoble Technology producer/CSP

1) Partnership Director2) Project Manager

15/07/2014

56 min

Member 6 Medium National/Grenoble Software/Energy efficiency

Business Development Manager

15/07/2014

74 min

Member 7 SME Regional/Grenoble System’s components

General Manager 03/09/2014

75 min

Member 8 SME Regional/Lyon Components/(PV) Market Manager 29/07/2014

74 min

Member 9 SME Regional/Chambery

Software; Consultancy /Energy efficiency-PV

Firm Director 18/07/2014

66 min

Member 10 SME Regional/Chambery

Consultancy 1) Head of the Regional Business Unit2) Knowledge and Innovation Manager

21/07/2014

37 min

Member 11 SME Global/Chambery Energy demand/energy efficiency

R&D Manager 23/07/2014

99 min

Member 12 SME Global/Grenoble Components/smart grid (storage)

R&D Coordinator 30/06/2014

61 min

Member 13 SME Global/Grenoble Software/Energy efficiency

Firm Director 24/07/2014

62 min

Member 14 Start-up Regional/Grenoble Technology producer

Chief Operating Officer 15/07/2014

65 min

Member 15 Start-up Regional/Lyon Components/(PV) Founder 24/07/2014

60 min

Member 16 Start-up Regional/Chambéry

Technology producer/Biogas

Founder 30/07/2014

56 min

Member 17 Start-up Regional/Grenoble Technology producer/hydro

Founder 02/09/2014

91 min

Member 18 Start-up Regional/Chambéry

Software/Energy efficiency

Founder 07/07/2014

82 min

Member 19 Start-up Regional/Chambéry

Technology producer/(biomass)

Head of Strategy, Finance and Administration

05/09/2014

81 min

Member 20 Large National Research and development

Scientific Director 17/12/2014

62 min

(20 interviews in total, 23h59m). Interviews took place face-to-face except those with Member 4, 10 and 15.

37