New path development between innovation systems and individual actors

ABSTRACT

This special issue is devoted to studying mechanisms that may stimulate or hamper the renewal of existing industry paths and the growth of new paths. In this guest editorial, we look closely at the role of policy instruments in situations where the majority of regional industries are embedded in strong regional and/or national innovation systems. This theme is currently very relevant in (parts of) Norway, where the dominant oil sector is downsizing and new growth paths are required to maintain employment and welfare. The guest editorial presents a theoretical framework for new regional industrial path development, followed by a discussion on how two Norwegian policy tools contribute to new path development. The 10 articles of the special issue study different aspects of new regional industrial path development based on cases in Norway, Sweden and Spain. Some papers also discuss the role of policy in new path development. Based on the findings from these articles, we believe that new path development is fostered by policies that incorporate both actor-based and system-based elements. Such policy mix could provide a vital push towards new path development.

KEYWORDS:

• Path development
• innovation system
• innovation policy
• actor-based policy
• system-based policy

Introduction

Recent years have witnessed a growing interest in, and need for, the modernization of existing industry paths, and in particular the development of new industrial paths at the regional and national levels. Industry and innovation systems need to develop activities that go beyond path extension. Thus, the ways in which regions and nations are able to achieve path renewal and path creation are high on the research agenda in disciplines such as evolutionary economic geography (EEG) (Fløysand & Jakobsen, 2016; Isaksen, 2015; Martin & Sunley, 2006; Neffke, Henning, & Boschma, 2011). The EEG approach emphasizes that path renewal and creation are strongly rooted in the existing economic structure of regions and nations. ‘The local inherited knowledge and skill base of an industry can form the basis of the rise of related new local paths of industrial and technological activity’ (Martin, 2010, p. 19). This focus on historically developed skills and industrial structure is a reminder that regions cannot develop any kind of new industries and growth paths easily, or perhaps at all; for example, most regions cannot replicate initiatives and growth in dynamic high-tech clusters. However, historically developed structures can function as a double-edged sword; they can both hamper and stimulate new industrial growth paths.

On the one hand, functioning innovation systems that industries and their knowledge and support structures have jointly adapted over time strengthen firms’ innovation capabilities and competitiveness. This viewpoint is highlighted by Niosi (2010, p. 43), who argues that ‘countries possessing national and regional innovation systems in specific sectors will become wealthier’. Innovation systems provide firms with access to relevant competence and other resources, and include institutional frameworks that support their innovation activities.

On the other hand, strong innovation systems may have a weakness that becomes especially evident when the industry has to renew itself quickly. This weakness lies in the fact that innovation systems are mostly geared to support already strong industries. This is particularly apparent in small countries with limited human resources. Thus, innovation systems support path extension in particular, through continual improvements of already strong industries. This is the case when study programmes, R&D activity and policy tools are directed at supporting existing strong regional and national industries. The knowledge infrastructure, such as universities and R&D institutes, may initiate new growth paths, but there is still a significant risk that strong innovation systems will hamper rapid restructuring when the industry is well integrated into such systems.

This special issue is devoted to studying mechanisms that may stimulate or hamper the renewal of existing industry paths and the growth of new paths. The articles use Norwegian cases as their empirical test beds, in addition to two articles that build on cases from Sweden and Spain. In this guest editorial, we look closely at the role of policy instruments in situations where the majority of regional industry is embedded in strong regional and/or national innovation systems. This theme is currently very relevant in (parts of) Norway, where the dominant oil sector is downsizing and new growth paths are required to maintain employment and welfare. We start by providing a theoretical framework for new regional industrial path development. Thereafter, we refer to this framework in a discussion of how two Norwegian policy tools contribute to new path development. Finally, the editorial presents the major findings of the other articles in the issue and considers how these fit into the overall framework discussed in the editorial.

New path development

New regional industrial path development includes path renewal and path creation (Isaksen, 2015). Path renewal involves the growth of new activities and new industries via regional branching; existing knowledge and skills in a region are combined in new ways and may be linked to relevant, extra-regional knowledge to provide new knowledge for a region that enables innovations and entrepreneurship (Boschma & Frenken, 2011). Path creation represents the growth of entirely new industries for a region or nation. These are often created from scratch based on results from R&D activities or imported knowledge.

New regional industrial paths develop through two micro-level processes: (1) formation of new ‘local’ firms or transplantation of firms from other places in new industries in the region or (2) commencement of new activities by existing local firms in new industries in the region. Both processes highlight the importance of actors such as entrepreneurs, firms and other organizations. The latter group includes for instance venture capitalists, incubator staff and support organizations who facilitate the transplantation of firms from outside the region. However, start-up and renewal of existing firms in new fields in the region will not necessarily lead to new path development. A new industrial growth path emerges in a region (1) when several functionally related firms are established, (2) when the firms face existing or potential demand and a market or (3) when the firms find input factors in a regional innovation system (RIS) and/or gain access to production and knowledge networks outside the region (cf. Binz, Truffer, & Coenen, 2016, p. 177). Firms are functionally related when they use corresponding knowledge and technology, or belong to the same value chain. This interpretation implies that the emergence of new growth paths demands more than just entrepreneurship and innovation activity; it demands the emergence of several related firms linked to supportive actors and institutions.

To simplify somewhat, the innovation system approach has been quite static and has not emphasized the importance of actors in system changes. Approaches that focus on individual actors such as entrepreneurs, on the other hand, have put little emphasis on the fact that actors often need support from those in the ‘environment’ to contribute to a robust renewal of a regional industry. This guest editorial discusses the ‘pure’ actor and innovation system approaches to new path development. We argue that a framework that combines the two approaches is the most appropriate for analysing new path development and the most relevant basis for designing policy to support new regional industrial paths. The editorial analyses the extent to which important actor-based and system-based policy programmes in Norway combine the two approaches and the extent to which they (can) lead to new path development.

Actor-based approaches

Entrepreneurs and firm leaders are certainly of vital importance for the renewal of regional industry. However, human agency is also central for the development of supporting institutional frameworks for new path development. For example, policy-makers, university representatives and other actors contribute to the creation of functioning RISs, in which global knowledge links are also often highly important (Saxenian, 2006).

A discussion of the role of human agency in new path development may begin with different conceptualizations of agents – from those embedded in social structures (Uzzi, 1997) to knowledgeable actors that are able to deviate mindfully from existing structures (Garud & Karnøe, 2001). The concept of embedded agents is similar to that in the path-dependent approach, which in some interpretations allows little room for human agency. The approach focuses on technological relatedness, knowledge combinations and branching out of new activities and firms from the existing industrial structure and competences in a region (Boschma & Frenken, 2011). Such an interpretation ‘downplays the influence of non-firm actors, institutions and public policy in creating and/or renewing industrial development paths in a region’ (Binz et al., 2016, p. 173). Rather, human agency has ‘to go with a flow of events that actors have little power to influence in real time’ (Garud & Karnøe, 2001, p. 2). As a result, actors mostly extend existing development paths, while ‘the emergence of novelty is serendipitous’ (Garud & Karnøe, 2001, p. 7).

At the other end of the spectrum, actors are believed to be capable of making ‘free’ decisions based on their own motivations and priorities. Alsos, Carter, and Ljunggren (2014, p. 97), for example, maintain that ‘entrepreneurship research traditionally views both the individual and the firm as decontextualized entities’. Approaches that focus on intentional and purposive action by individual agents consider that new pathways and the renewing of regional industry ‘require social action by knowledgeable pioneering individuals, universities, companies and/or governments’ (Simmie, 2012, p. 769). In a similar way, mindful deviation from existing structures by entrepreneurs is believed to constitute the heart of path creation (Garud & Karnøe, 2001, p. 6). The ‘mindful deviating entrepreneurs’ are still affected by social structures in several ways. First, new pathways may result from joint contributions by a number of actors, such as economic agents, policy-makers and potential customers (Simmie, 2012). Second, relevant actors can create favourable framework conditions and resources – for example, through policy actions – and can initiate new economic activity by mobilizing necessary resources. Third, the extent and importance of mindfulness by actors may differ from case to case and over time. Simmie (2012) describes how wind power industry entrepreneurs in Denmark employed and gradually developed local knowledge to supply local markets in rural areas and probably did not deviate from existing knowledge or business models. Later, the emerging cluster of wind power firms was supported strategically by government subsidies and tax relief.

Based on this short review of various approaches, we consider that human agency lies somewhere between ‘dependent’ actors that are strongly influenced in their decision-making by opportunities and restrictions set by regional and social structures, and actors that make ‘free’ decisions based on their own motivations and priorities. This interpretation corresponds with a view of entrepreneurship as ‘the result of the interaction between individual actors and the surrounding environment’ (Bosma, Schutjens, & Stam, 2011, p. 484). Actors ‘identify, evaluate and exploit entrepreneurial opportunities within certain structures, while at the same time influencing these structures’ (Bosma et al., 2011, p. 482).

Such considerations lead us to consider new regional industrial growth paths initiated by actors that can both utilize pre-existing regional capabilities and develop or mobilize completely new knowledge and skills in a region. A new path also induces ‘changes to the relevant institutional contexts of the emerging industry’ (Binz et al., 2016, p. 179). Binz et al. (2016) describe steps in the development of an OST (on-site water recycling technology) industry in Beijing, including initiatives by various actors and ‘system development’. First, foreign companies entered Beijing to utilize a small niche market created by regional actors. This motivated ‘highly skilled returnee entrepreneurs to found de novo start-ups and local actors to start spin-offs from related sectors and local universities’ (Binz et al., 2016, p. 191). Thereafter, the imported technology led to a learning-by-doing process, and the development of an RIS with international links formed around the OSTs.

System approach

A system approach to path development departs from the fact that renewal, innovation and new firm formations rely on not only the internal competence of entrepreneurs, firms and organizations, but also networks and systemic configurations that stimulate co-operation and provide actors with new competences and critical resources. There are two main schools of thought concerning regional path development as a systemic phenomenon: RIS theory and cluster theory.

The concept of an RIS was developed during the 1990s and is linked to a broader line of literature on innovation systems that also includes studies of national systems of innovation and sectoral systems of innovation (Asheim & Isaksen, 1997; Cooke, 1992; Edquist, 1997; Jakobsen et al., 2012; Lundvall, 1992). The importance of geography and the regional level in the RIS literature is based on the observation that knowledge spillovers, which are essential in processes of interactive learning and innovation, tend to be spatially bounded and to decrease with distance. It is also true that regions differ with respect to industrial specialization, institutional architecture and patterns of innovation (Isaksen & Trippl, 2016; Tödtling & Trippl, 2005). Thus, different types of RIS need different policies to stimulate upgrading and renewal of the industry.

An RIS can be seen ‘ … as a specific framework in which close inter-firm interaction, knowledge and policy support infrastructure, and socio-cultural and institutional environments serve to stimulate collective learning, continuous innovation and entrepreneurial activity’ (Isaksen & Trippl, 2016, p. 70). However, RISs differ in many aspects, for instance, in knowledge foundation, type of firms and type of institutions. Asheim and Gertler (2005) distinguish between a broad understanding of RISs that covers all aspects of the (regional) economic structure and institutional set-up (i.e. all firms and the wider system of organizations supporting learning and innovation) and a narrow understanding that considers RISs to be interactions between knowledge organizations (such as universities, colleges or research institutions), knowledge-intensive firms, technology transfer organizations and other supporting organizations. Furthermore, we can differentiate between organizationally thin RISs and organizationally thick RISs (Tödtling & Trippl, 2005). The former type has a low number of firms and organizations in knowledge generation and diffusion, while the latter has a large number of firms and knowledge organizations.

The challenge for achieving new path development varies between types of RISs. In organizationally thin RISs, renewal and new firm formation often presuppose the establishment of links to external expert milieus and the import of new ideas and technology. This can provide existing firms with new competences and new solutions, while a new industry path can be created by importing products and technological solutions developed in other regional settings. In organizationally thick RISs, new path development to a larger extent is stimulated through spin-offs from knowledge institutions, by intensifying co-operation between knowledge organizations (such as universities, colleges and research institutions) and the industry and by the branching of existing industries into new and related industries (Boschma & Frenken, 2011; Isaksen, 2015).

Cluster theory departs from Porter’s (1998, 2000) seminal work on industry clusters. He defined a cluster as a geographical concentration of interconnected and related companies and supporting organizations. Thus, a cluster system is narrower than an RIS, which can include several clusters. The basic idea of cluster theory is that geographical concentrations of similar and related firms and organizations stimulate competitiveness and innovation capabilities, and make firms that are part of a cluster more competitive than those that are not. The rationale for this is the presence of an ‘upgrading mechanism’ within a cluster, with knowledge spillovers, trust, complementarity, collective learning and rivalry as the most important components (Fløysand, Jakobsen, & Bjarnar, 2012).

The evolution of clusters as a systemic configuration is linked to processes of scale and scope. The issue of cluster scale has been thoroughly debated in recent literature. The literature agrees that external links are crucial for the renewal of mature clusters. To be innovative, strong clusters are dependent on access to new knowledge, information and networks (Bathelt, Malmberg, & Maskell, 2004; Nadvi & Halder, 2005). It has been widely acknowledged that knowledge links need to be balanced between the local and the global; the embeddedness of both clusters and their firms has been extensively investigated (e.g. Fornahl & Tran, 2010; Kramer & Diez, 2011; Montagnana, 2010). For instance, Bathelt et al. (2004) argue that the coexistence of intense local networking and a high number of external linkages facilitate collective learning processes that trigger innovation in a cluster.

The scope of a cluster concerns the type of industry actors and associated institutions that should constitute it. As with the debate over scale, there is widespread discussion about how specialized or diverse a cluster should be. Porter (2000) originally argued that a cluster should include firms and institutions in related industries, while others have argued that more narrowly defined and specialized clusters are beneficial for development (Reve & Sasson, 2012; Sölvell, Lindqvist, & Ketels, 2003). New lines of research have been more sceptical (Cooke, 2012), as in fact, specialization can hamper innovation. It is widely understood that innovations are stimulated by new combinations of dissimilar knowledge types. In this context, related variety is a useful concept (Boschma & Frenken, 2011). Aarstad, Kvitastein, and Jakobsen (2016) found that related industry variety promotes both innovation and productivity in clusters.

Traditionally, public cluster programmes in Norway and other Western countries have emphasized that clusters should be regional and specialized (Fløysand et al., 2012; Isaksen, 2009; Njøs, Jakobsen, Aslesen, & Fløysand, 2016; Sölvell et al., 2003). In addition, there has been a strong focus on strengthening external cluster linkages, and a consensus that external links are crucial for cluster evolution and growth. However, it can be argued that cluster policy is often ‘more of the same’. It concerns strengthening the development of already strong industrial milieus, so it is closer to path extension than to new path development (Njøs & Jakobsen, 2016).

Two main policy approaches

Based on the above distinction between actor-based and system-based interpretations of regional industrial development, we distinguish between two main types of innovation policy approaches: those that primarily target individual actors and those mainly intended to improve the way in which ‘systems’ support the innovation capabilities of industry. Here, as in Nauwelaers and Wintjes (2003), we explicitly refer to regional systems and clusters while being well aware that national and international systems and knowledge links are also important for the innovation activities of most firms. Based on the above discussion, Table 1 proposes key strengths and weaknesses of the two types of innovation policy strategies with regard to the development of new regional industrial growth paths. Actor-based policy tools may support the innovation activity of individual entrepreneurs and firms. However, innovation activities risk receiving little support from existing RISs and clusters that are mostly geared towards currently strong industries. System-based policies, on the other hand, may assist in the restructuring of innovation systems and clusters so that education programmes, R&D activities and policy instruments stimulate new growth paths more efficiently. However, if regional industry and entrepreneurs have limited absorptive capacity for new knowledge, few effects are expected, which indicates that knowledge organizations act as ‘cathedrals in the desert’ (Hansen, 1992).

Table 1. Characteristics of actor-based and system-based regional innovation policy.

Policy approaches	Characteristics	Strength with regard to stimulating new path development	Weakness with regard to stimulating new path development
Actor-based	Support entrepreneurs and firms’ innovation projects	May identify and support actors with innovation activities that are new to a region	Little backing of actors with new ideas and projects from the existing clusters and RISs
System-based	Improve the functioning of clusters and RISs; fix system failures	Restructure and further develop clusters and RISs to provide better support for innovation activity in new industries in a region	Actors with relevant competence to initiate new growth path may not be present, that is, ‘cathedral in the desert syndrome’

In the following section, we present two important Norwegian policy tools – one actor-based and the other system-based – and discuss the extent to which the framework in Table 1 characterizes the strong and weak parts of these tools.

Commercializing R&D results – an actor-based policy tool

Research-based innovation is a key concept in Norwegian innovation policy. In its strategy document for 2015–2020, the Research Council of Norway maintains that ‘research-based innovation is central for the development and restructuring of Norwegian industry. Therefore, the Research Council has a role to play in industrial politics’.¹

The FORNY (commercializing R&D results) programme began in 1995 and since then has been the central policy tool for stimulating commercialization of publicly funded research in Norway. From the start, the aim of this policy tool was to improve the ability to commercialize research-based business ideas at universities and research institutes. The main instrument was the establishment of so-called commercialization units to help researchers to manage different parts of the commercialization process. The work is now performed by seven Technology Transfer Offices (TTOs) run by universities or by other actors in which the universities have shares. The TTOs help with patenting, licensing and start-ups, but their activities are seldom linked to other activities at the universities (Spilling, Borlaug, Iversen, Rasmussen, & Solberg, 2015).

An evaluation of the FORNY programme in 2009 criticized its relatively limited achievements (Borlaug et al., 2009). From 1996 to 2009, the programme supported about 300 start-ups, with a total of approximately 700 employees and a turnover of 900 million NOK (ca. 100 million euros) in 2009. The 13 TTOs involved in the FORNY programme at that time employed staff equivalent to 100 man-years. Most of the supported firms survived but on a small scale with turnovers of up to one million NOK (ca. 110,000 euros) and with entrepreneurs as sole employees. Only three or four firms could be characterized as successful and accounted for a large proportion of the total number of jobs and turnover.

The next evaluation in 2015 (Spilling et al., 2015) did not make any similar assessment of results. The evaluation considered the organization of the commercialization support system. It found that the system functioned quite well at the universities and at some of the research institutes and that it was regarded as more focused and targeted than that described in the 2009 evaluation report. The 2015 evaluation found a marked increase in patenting with researcher participation, a clearly more positive attitude towards commercialization among researchers and a steady growth in technology sales and licence agreements from 2011 to 2014. However, no significant change in the number of start-ups was found. An analysis in 2013 (described in Spilling et al., 2015) of the portfolio of 474 firms and 424 agreements of technology sales and licence agreements in the FORNY programme found a number of promising projects. However, commercialization of research results takes time, and the 2013 analysis also found that the programme mainly produced small firms. This suggests that the results of the FORNY programme in terms of the number of jobs generated remain modest, despite the fact that the organization and implementation of commercialization activities are described as good.

The FORNY programme may contribute to new path development, and in particular to path creation. The policy tool helps to commercialize results from research milieus, so it may give rise to innovations and industrial activity that are new to a region. The new innovative firms are mainly spin-offs from universities and research institutes; therefore, they should benefit from ongoing R&D activities at their ‘mother organizations’. The programme could also lead to reconfigurations of the R&D system with new and stronger research groups and possibly new education programmes. On the other hand, the evaluation document indicates fairly modest results. It may be pertinent to question whether modest results reflect the linear view of innovation in the programme, whereby research results are ‘pushed onto’ the market. A system approach entails putting market potential and contributions from potential customers as well as possible technology suppliers and producers higher on the agenda early in the innovation process. A broader set of actors is then brought into the innovation and commercialization process, which can ensure that products based on research target the market and can be produced efficiently. However, this demands a broader view of innovation than the linear approach that seems to dominate the programme and the activities of the TTOs.

The Norwegian Centre of Expertise programme – a system-based policy tool

Our example of a system-based approach is a cluster programme. Promoting the development of clusters has a strong foothold in the innovation policy of Norway. The main scheme is the Norwegian Innovation Clusters programme, run by Innovation Norway. There are three subprogrammes under this scheme. The most important of these is the Norwegian Centre of Expertise (NCE) programme, initiated in 2006 and designed for mature clusters with a strong international position. Financing for projects is granted for up to 10 years. The intention of the programme is to ‘enhance sustainable innovation and internationalization processes in the most dynamic and growth-oriented Norwegian clusters’ (nce.no). In October 2016, there were 14 NCE cluster projects in operation. In addition, the ARENA programme is aimed at emerging and immature clusters. Cluster projects can receive ARENA status and financing for 3–5 years. At present, 20 ARENA cluster projects are in operation. The public funding for these clusters is substantially smaller than that for the NCE projects. A third cluster level has recently been initiated in Norway: Global Centres of Expertise (GCE). Three GCE projects are running, all of which were previously NCE projects. Status and financing are granted for up to 10 years.

If we examine more closely the most important part of the Norwegian Innovation Clusters scheme – the NCE programme – the description states that NCE clusters should encompass a clear concentration of firms, both small and medium sized enterprises and large specialized suppliers, and many of these should also be globally oriented firms (2014). Furthermore, they must represent a ‘specialized, attractive labour market in the cluster’s regional area of impact’ (p. 2; our translation). Thus, the aims of the programme are to improve the conditions for value creation and to strengthen the position of the cluster firms in national and global value chains. Furthermore, it is especially emphasized that ‘connections between different suppliers and connections to buyers and users are crucial for well-functioning systems and solutions’ (p. 21). The programme highlights the importance of co-ordinating and strengthening vertical linkages in value chains as a key characteristic of dynamic clusters. Hence, it can be claimed that the NCE programme emphasizes ‘value chain thinking’ (Njøs & Jakobsen, 2016).

The findings from evaluations and analyses of the NCE programmes reveal that the projects have mainly nurtured the existing value chains in which the firms operate (Econ Pöyry, 2009; Jakobsen et al., 2012). It has also been pointed out that new cluster members have been recruited from existing niches (Oxford Research, 2013, p. 29). In general, this observation supports our initial assumption that system-oriented policy tools contribute to the further development and strengthening of existing structures and networks. By nurturing already strong clusters, the NCE programme mainly contributes to path extension.

However, when we investigate the programme more closely, there are several indications that it can also contribute to path renewal and even path creation. In the 2015 programme description of the NCE initiative, ‘cluster-to-cluster co-operation’ is listed as one of four priority areas, and all of the NCE projects are expected to use resources to stimulate such co-operation (Norwegian Centre of Expertise, 2015). In some cases, this co-operation can occur between clusters in the same branch or sector, but there are also several examples of co-operation between clusters in different sectors. For instance, NCE Subsea (now a GCE) and NCE Seafood, both located in the Western part of Norway, have launched several ‘crossover initiatives’ together. As one example, they have supported projects where competences and technology from the oil and gas industry have been used to solve challenges in the aquaculture industry. Such cross-industry innovation practices have the potential to renew existing industry paths (www.gcesubsea.no/). Innovation Norway, which is running the NCE programme, has also recently launched an initiative called ‘the cluster as a restructuring engine’. Three NCE cluster projects have been granted funding, and they will use well-known and proven competences and technology from their cluster firms to promote innovation and growth in other sectors. This also demonstrates how system-based tools can stimulate path renewal and path creation².

The content of the special issue

The 10 papers that follow study different aspects of new regional industrial path development based on cases in Norway, Sweden and Spain. Some papers also discuss the role of policy in new path development. The article by Doloreux and Porto Gomez includes a comprehensive assessment of the literature on RISs between 1998 and 2015, and discusses how this field of study should move forward. Supporters of the RIS approach aim for policy relevance and particularly stress the importance of building and upgrading regions’ knowledge bases and of the interaction and knowledge flow between private and public actors. Nevertheless, Doloreux and Porto Gomez argue that policy-makers find only limited guidance in the RIS literature, as it has not developed any ‘standard model’ for the development of RISs. Instead, they argue that there is no one-size-fits-all policy, calling for analytical and policy development competence at the regional level rather than blind replication of so-called best-practice policy prescriptions. However, according to Doloreux and Porto Gomez, RIS research should develop a ‘true’ evolutionary dimension that considers how new actors can emerge and existing actors can alter their roles in RISs. The authors call for more research on the link between firm innovations, regional economic growth and the transformation of RISs.

Several articles address this challenge. The paper by Mæhle, Hauge, Kyllingstad and Schulze-Krogh investigates cross-industry innovation capability (CIIC) building in firms and discusses how conditions for innovation and learning in regions drive this process. CIIC is the firm’s ability to transform knowledge and ideas from different industries into new products, processes and systems, and/or its ability to adapt existing products, processes and systems to new industries. The authors predict that if a substantial number of firms in a region possess high CIIC, path renewal will take place. Their study of CIIC building draws on case studies of firms in western and southern Norway, an area highly dependent on oil. The authors observe that firms that institutionalize cross-industry innovation thinking in their strategy are better off developing CIIC. They also find that some regional characteristics favour the development of CIIC, while others do not. They conclude that successful policies cannot focus solely on pushing technology and R&D investment to promote innovation, but should also be concerned with learning processes and the development of a strategic mindset at the firm level.

Gjelsvik and Aarstad analyse the role of entrepreneurship and financial institutions in regional development in south-western Norway. Regions evolve in path-dependent ways, but the degree of dependence or novelty may vary over time and between regions. Using register data, they investigate whether new firm formations reproduce the regional industry structure. They observe a steady decline in new firm formation in both related and unrelated industries after 1998, indicating a pattern of path extension but not diversification. This trajectory coincides with a steady and substantial boost in the oil price. This trend of path extension is stronger for south-western Norway than for the rest of the country. Furthermore, qualitative and quantitative data from financial institutions in the regions indicate that these institutions have contributed more to path extension than to path renewal or path creation. Regarding the policy implications of their study, the authors stress that there is no quick fix to the problem of strong entrepreneurial path dependence. However, policy-makers should be aware of the challenge and should elaborate potential strategies to stimulate sectoral diversification in new firm formation.

The article by Billington, Karlsen, Mathisen and Pettersen also explores the relationship between firm development and the regional environment. Their approach focuses on four firms that have demonstrated resilience by surviving over a long period and through times of external turbulence. The authors argue that the firms have built organizational resilience. The development of resilience is seen not only as the result of long-term investment in competence and workplace culture, but it also reflects values, resources and networks embedded in the region. The regional milieu supports the firms’ resilience, and the firms contribute to the development of regional resilience. This conclusion also underlines the need to analyse both the actions of individual firms or entrepreneurs and the characteristics of a region or locality to understand the long-term resilience and innovation capability of regional industry.

One potentially important group of actors in new regional growth paths is multinational companies (MNCs). Aslesen, Hydle and Wallevik explore how MNCs can act as extra-regional sources of new path development in two Norwegian regions with thick RISs and industrial specialization in the oil and gas sector. Based on investigations of 15 MNCs, the authors find that the local units of these companies are highly dependent on acquiring competence from the MNCs’ networks and RISs. Although some signs of path renewal are found, path extension through incremental innovations dominates. One reason is that many innovation activities in the local units of MNCs depend on approvals from the MNCs. Another reason is that the RISs are geared to supporting their existing specializations. Thus, path renewal may be difficult to achieve because of the organizational structure of the MNCs and the lock-in tendencies of the RISs. One way to achieve new growth may be to develop the RISs further; for example, through new educational programmes and policies that can spur investment by the MNCs in new areas.

Fløysand, Njøs, Nilsen and Nygård analyse a similar topic. Their article investigates the linkages between foreign direct investment (FDI) and the renewal of industries through case studies of the development of four regional industry paths. They argue for a perspective where the reciprocity between FDI and renewal of industries is the outcome of both material and discursive processes. Hence, they show how interdependence between materiality and discourse has different effects upon industry development. Path exhaustion and path extension are observed when FDIs are capital dominated and contain limited spillover effects, and when anti-FDI narratives and discourses dominate, while path renewal and path creation occur when FDI arrives as a capital–network–knowledge package and there are substantial regional spillovers and pro-FDI narratives and discourses. Policy initiatives are observed to be important for the development of two of the selected industry cases. Here ‘local content policy’ encourages MNCs, the key actors in FDI, to engage in projects that result in regional spillovers. Furthermore, the article indicates how directed policy action can influence the reciprocity between discourse and materiality. Consequently, the authors argue that policy should be considered as crucial in directing the interplay between FDI and the renewal of industries in a way that enriches both the MNC and regional industry as such.

Several other articles focus more specifically on the role of innovation policy in new regional industrial pathways. Miöner and Trippl focus on how a regional environment that constrains the development of new development paths can be transformed into an enabling environment. Such transformations include changes in the policy support structure, which can take three main forms. Layering consists of gradually adding new elements (such as policies and other institutions) to the support structure. Adaptation includes changes in, or reorientation of, existing institutions and organizations, for example, tailoring education programmes and incubation activities to emerging industries. A novel application is the utilization of existing policy instruments and organizations for new purposes, such as using existing policy tools in new industries. Miöner and Trippl study the emergence of the digital games industry in the Scania region of Sweden. They find that the growth of this industry results from combinations of three types of change in the policy support structure described above and that the changes were implemented by a few key individuals operating in the newly emerging paths. The case illustrates the importance of changes in the RIS for new growth paths to emerge as well as the role of key actors in undertaking new initiatives.

Holmen and Fosse come to a similar conclusion through a somewhat different approach. These authors analyse the role of regional agency in the early phase of path creation. Their study investigates the constitution of two new industry paths in the western part of Norway, a region highly dependent on the oil and gas sector. The study shows that two types of agency are vital in the processes of path creation. First, there is the regional policy agency that contributes to the creation of regional frames for new industry paths. Second, there is a need for a strong entrepreneurial agency that can push the process forward. Such agents can be a person, a firm or a network. In their investigation, Holmen and Fosse found that cluster projects, as a collective organization, created a common arena in which to exercise this type of entrepreneurial agency. Policy lessons from their study are that policy and entrepreneurial agency are interdependent in constitution processes, and policy initiatives need to gain strong support from firms or network organizations to be successful.

Fitjar and Timmermans emphasize the need for context-specific policy approaches. They employ a new method to study related variety in regional industrial structures. The authors develop a regional skill-relatedness measure that builds on labour mobility flows across Norwegian industries, and they identify the regional industries to which each industry in a region is related. Their analyses of Norwegian regions demonstrate a highly differentiated pattern with regard to how many and which industries are related through labour mobility flows, and this applies even to regions of similar size. They find more clearly related industries in urban regions than in small peripheral ones, but the data also reveal large differences within each category of region. This finding causes the authors to warn against policies such as treating all rural or peripheral regions in the same way or dishing out the same policy solutions to all large urban regions. Rather, policies must depart from specific analyses of regional contexts in individual regions.

The article by Elola, Valdaliso, Lôpez and Franco also puts policy at the forefront of analyses, and they investigate the role of public policies in the origin and evolution of six industry clusters in the Basque Country (Spain). Not surprisingly, they observe that industrial dynamics and market conditions have been the most influential factors in the development of the clusters. Public policy has also been important in the emergence and development of the clusters, but in general, public policy has played a secondary role. They also find that a broad-based general policy rather than a specific cluster policy has had the greatest influence on the emergence and development of these clusters. This can partly be explained by the fact that they have conducted historical studies tracing the development of the clusters back to the nineteenth and early twentieth centuries. Cluster policy as a distinct policy field was not introduced until the 1990s. Furthermore, they argue that there is a need for a more specific focus on the linkages between broad-based general policies and cluster policies. In general, cluster policies need to be better co-ordinated with wider policy to be more effective.

Summary

Several articles in this Special Issue indicate that many Norwegian regions are dominated by path extensions. Empirical studies demonstrate a steady decline in new firm formations in both related and unrelated industries over the past 20 years. RISs are mostly geared to supporting existing specializations. The activities of MNCs also frequently strengthen path extension. Local units of MNCs must often stick to the product and market niches appointed by higher levels, and many innovation activities in the local units are quite dependent on approval by the MNCs.

Policy recommendations drawn from the articles suggest that path renewal and path creation require co-ordinated activities by several actors. These demand learning processes at the entrepreneur and firm levels, and the development of a strategic mindset by these actors. MNCs can also stimulate path renewal and path creation when they act as technological gatekeepers and engage in substantial regional spillovers. Moreover, it is illustrated that formalized networks can provide arenas for collective entrepreneurship. However, there is also a need to develop RISs further, for example, with new educational programmes, new R&D activities, new initiatives for co-ordination between subsystems and new proactive policy tools. Policies to develop RISs should also be better co-ordinated with wider policy to be effective.

Based on the findings from the articles, we believe that new path development is fostered by policies that incorporate both actor-based and system-based elements. Such an approach is not really seen in the current discussion about policy to create new growth in the wake of the reductions in the numbers of jobs and activities in the oil and gas sector in Norway. Two main strategies are most often proposed by the government, ministries and other key stakeholders, which are to increase the number of entrepreneurs by providing better access to capital and competence, and to support high-quality basic research that will lead to new research results that can be commercialized. These are mostly actor-based strategies targeting entrepreneurs and research centres and teams of high quality. Thus, we have advocated more emphasis on system-based initiatives. Such a broad innovation policy, consisting of a mix of complementary actor-based and system-based tools, could provide a vital push towards new path development.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Research Council of Norway (P233737/o50).

Notes

1. Quoted from the Strategy of the Research Council of Norway, pp. 7–8. Authors’ translation.

2. http://www.nceraufoss.no/startside/aktuelt/item/779-klynger-som-omstillingsmotor