Exogenous sources of regional industrial change

1 Arrival of new actors from outside the region

New path development can be triggered by the arrival of individual and organizational actors from outside the region. Inflow of new organizations can induce or support new growth paths as they bring new knowledge to the region. A powerful mechanism may include the resettlement of firms and other organizations such as R&D institutes. Neffke et al. (2014) have shown that it is relocating firms and not so much regional start-ups and incumbents that introduce most structural change (creation of new industries that require new capabilities instead of existing capabilities) in Swedish regions. New path development may also originate from the relocation of R&D organizations (Isaksen and Trippl (2016b) and the arrival of foreign direct investment (FDI) (O’Malley and O’Gorman, 2001). Martin and Sunley (2006: 423) note that ‘new knowledge brought into a region by the inward transplantation of firms from elsewhere (through FDI or takeover or merger) may be critical in initiating a new technological or industrial path locally’. MacKinnon (2012) argues that the relation between FDI and regional evolution is complex and path dependent in nature and may differ depending on the type of region under consideration. In particular in non-core regions the link between FDI and regional development can be negative, as

regions can become ‘locked-in’ to external networks…controlled by TNCs [transnational corporations]…as is evident from the experiences of branch-plant regions which became over-reliant on relatively low-value production plants, lacking more advanced functions and high-status employment as a result. (MacKinnon, 2012: 236)

Individual actors are carriers of knowledge, transferring expertise and know-how from one place to another by means of their mobility. Thus, the migration and mobility of highly skilled and resourceful people is a potential source for new path development (Kapur and McHale, 2005). Florida’s (2003) work points in this direction, suggesting that creative people move to places characterized by openness, diversity and tolerance and attract firms which seek ‘to draw from concentrations of talented people who power innovation and economic growth’ (Florida, 2003: 5).

Scholarly work on transnational(izing) entrepreneurship (Yeung, 2009) has sharpened our understanding of entrepreneurs moving across spaces and territories. Drori et al. (2009) find a positive impact of transnational entrepreneurship on the institutional environments of less-developed regions. Saxenian’s (2006) work on the ‘new argonauts’ is one of the few studies that have explicitly explored the link between the inflow of new knowledge through mobile individuals and new path development. Saxenian (2006) shows that the rise of ICT industries in various Asian countries has been triggered by returnees from the Silicon Valley, who transferred technical knowledge and provided advice to policy actors about the establishment of favourable institutional set-ups for new paths in their home countries (Saxenian and Sabel, 2008).

2 Extra-regional knowledge linkages

New development paths may also be supported through extra-regional knowledge linkages. Such linkages come in different forms (Tödtling et al., 2006), ranging from market linkages (buying of knowledge embodied in patents or intermediary goods), to formal (contract-based) networks (R&D and innovation collaboration) and informal linkages. Formal linkages are often labelled as global pipelines (Bathelt et al., 2004). Informal linkages rest on social networks, for instance, between former colleagues or study friends living and working elsewhere (Agrawal et al., 2006). Interactive learning over distance often involves a combination of social and cognitive proximity, as is the case in communities of practice (Wenger, 1998) or epistemic communities (Amin and Roberts, 2008). Attending conferences and trade fairs has also been found to be an important mechanism for knowledge acquisition (Bathelt and Gibson, 2015; Henn and Bathelt, 2015; Power and Jansson, 2008).

In the following sections we explore the nexus between non-local knowledge and regional industrial change. We advance the idea that the opportunities for and challenges of exogeneously-led path development vary between different types of RISs.

1 Needs and capacities to attract and absorb non-local knowledge

Organizationally thick and diversified RISs are often seen as centres of radical innovations and hot spots of new firm formation (Duranton and Puga, 2002). They are usually well endowed with a variety of different industries, a diverse set of strong research organizations and a diversified support infrastructure. Several mechanisms underlie localized knowledge circulation, in particular movements of skilled workers and face-to-face contacts between economic agents, which are pertinent when information is rapidly changing and knowledge is tacit (Storper and Venables, 2004). Heterogeneity in the industrial and support structures creates a basis for inter-industry crossovers (Asheim et al., 2011), which are seen as conducive to path branching processes (Boschma, 2015). In addition, the presence of strong research organizations and a good entrepreneurial climate offer opportunities for the rise of entirely new growth paths. Thick and diversified RISs are thus characterized by favourable conditions for new path development, not least due to their good endowment of endogenous (local) sources.

At the same time, the economies of thick and diversified RISs are deeply rooted in external knowledge exchange (Scott and Storper, 2015). They function ‘as systems of dense local interactions imbricated in complex long-distance movements of people, goods and information’ (Scott and Storper, 2015: 7). These RISs exhibit a high capacity to attract globally available knowledge. They are the ‘preferred destinations’ of internationally mobile talented individuals and for research and innovation activities in multinational corporations.

The industrial development of thick and diversified RISs must be understood in the context of non-local knowledge exchange, or ‘in the context of distanciated economic flows and networks’ (Amin and Thrift, 2002: 53). Leading firms in these regions organize global production and innovation networks, and firms in some sectors depend on ‘project teams scattered around the world, linked up virtually and through placements’ (Amin and Thrift, 2002: 66). With lead firms in global networks, thick and diversified RISs have strong capacities to absorb knowledge from non-local sources and to exploit both local and non-local knowledge for path branching and new path creation. High absorptive capacity is also due to the presence of knowledge organizations with international links, diversity of industries, occupations and skills, and well-developed support structures with capacities to adapt to the needs of new industrial paths.

2 Arrival of new actors, global knowledge links and new path development

R&D organizations and firms in thick and diversified RISs are in general on the delivering rather than the receiving end of global knowledge pipelines. But as said above, new growth paths in this RIS type can rely on combinations of endogenous and exogenous resources. This is illustrated by the development of the biotech industry in Vienna (Trippl and Tödtling, 2007). The region has a long tradition in (bio-) medicine and a strong knowledge exploration capacity in this field due to an excellent scientific base. Lack of academic entrepreneurship (not least due to negative attitudes of university researchers towards new firm formation and collaboration with industry), prevalence of a bank-dominated landscape adverse to risk-taking, absence of venture capital, and a weakly developed public support infrastructure, however, undermined the capacity of regional actors to reap benefits from locally available scientific knowledge assets (Trippl and Tödtling, 2007), pointing to unbalanced exploration-exploitation capacities in Vienna’s RIS. The emergence of a new growth path around biotech activities was essentially triggered by the arrival of international companies with capabilities to transform the strong local research base into new regional industrial path development. Furthermore, the importation of managerial expertise through international recruitments of highly skilled people and various types of global knowledge linkages such as R&D partnerships have significantly contributed to the emergence of this sector. Regional and national policy interventions have played a major role in enhancing the absorption capacity. Public policy actions have thickened the organizational support structures to provide better conditions for academic spin-offs and to stimulate knowledge exchange between local firms, universities and the foreign-owned companies present in the region. This has become manifest in the creation of an ensemble of specialized organizations such as science parks, academic spin-offs centres, technology lisencing offices, public venture capital providers and cluster organizations facilitating networking activities (Trippl and Tödtling, 2007). This case demonstrates how new development paths in thick and diversified RISs can result from the ‘importation’ of exploitation capacities from outside and how their growth could lead to and be facilitated by a further thickening of RIS structures.

The oil and gas industry in Norway also represents a good example of how new growth paths can be initiated and supported by the attraction of non-local knowledge. The industry was transplanted into Norway from outside. Transplantation was required as nearly no knowledge of oil exploration and production existed in Norway when oil reserves were discovered in the North Sea in the late 1960s (Sæther et al., 2011). While the oil and gas industry in Norway started as path transplantation from outside, it continued by path branching through the diversification of ship yards, mechanical engineering and other firms, and it led to path creation when a totally new industry was formed. Inflow of knowledge from foreign sources was instrumental for this achievement.

Absorption capacity was increased through policy interventions. An important instrument in that respect was Goodwill agreements. These included that foreign oil companies got ‘goodwill points’ by contracting with Norwegian firms and research institutes. The amount of ‘goodwill points’ affected the result when concessions for developing new oilfields were passed to oil companies. The Goodwill agreements contributed to the development of a national innovation system with high research capacity and to the fact that the Norwegian oil and gas supplier industry has become globally competitive in a number of niches (Sæther et al., 2011). This example demonstrates the capacity to build new industrial paths on imported knowledge, conditioned on national receiver competence found in diverse industries in many parts of Norway. The growth of the oil and gas industry was accompanied by investments in the knowledge infrastructure in larger cities in Norway, which thus made their RISs thicker.

Examining Oslo’s performance in the path development process outlined above, however, also offers insights into how unbalanced exploration-exploitation capabilities might constrain path development in thick and diversified RISs. The engineering part of the oil and gas sector grew up and was highly concentrated in the Oslo region until the mid-1990s (Isaksen, 2003). The oil engineering sector from then on declined sharply in Oslo while it grew in other parts of Norway. The sector in Oslo faced difficulties in mastering the new technological development of subsea installations as it became locked-in to the ‘old’ technological path of building large, fixed and floating platforms. Oslo lagged behind in the construction of subsea installations because of the absence of advanced manufacturing firms in the region with research-based and non-codified, experience-based knowledge that could contribute to developing subsea technology (Isaksen, 2003). This points to a more general industrial development challenge in Oslo and some other large cities, which relates to little experience-based manufacturing competence necessary for ‘pushing’ and processing new scientific knowledge towards commercial utilization (Manniche, 2012: 1835). Thus, metropolitan areas and advanced technology regions may be thick and diverse with regard to the knowledge infrastructure and the presence of consulting, cultural and other service industries, but may be less well endowed with manufacturing production capacities, revealing unbalanced exploration-exploitation capabilities (Isaksen and Trippl, 2016a). As several scholars remind us, there might also be other constraints to harness endogenous and exogenous sources of path development in this RIS type. These may range from limited knowledge exchange due to cognitive distance resulting from too much diversity (Boschma, 2005) to competition between newly emerging and established paths over public support, human capital, finance and other scarce resources (Miörner and Trippl, 2017).

While exogenous sources potentially contribute in many ways to new industrial path development in thick and diversified RISs, it can be expected that high complementarities exist in particular if external impetus relates to internal knowledge exploitation capacities. These RISs have high capacities in knowledge exploration, and are globally embedded centres for research excellence. A question is then how to make this knowledge economically viable, which can be done through endogenous entrepreneurial experimentation. However, capacities for knowledge exploitation can also be brought in by external actors, who have the necessary absorption capacity, access to global markets and financial resources, either internally or through their network embeddedness. In this case, external actors could contribute to the creation of new paths by exploiting (but also adding to) the strong regional knowledge infrastructure.

1 Needs and capacities to attract and absorb non-local knowledge

Organizationally thick and specialized RISs exhibit a strong economic specialization pattern in only one or a few clusters as well as knowledge and support organizations that are well adapted to the needs of the dominating industries. A high degree of specialization can be positively linked to regional dynamism, particularly in early development phases of an industry or cluster, in which positive externalities and increasing returns fuel regional growth. In later stages, the structures, configurations and practices established in earlier phases may, however, become a source of rigidity, undermining regional development (see, for instance, Grabher, 1993; Martin and Sunley, 2006).

This section focuses on thick and specialized RIS configurations in old industrial areas (OIAs) (Grabher, 1993; Hassink, 2005; Trippl and Otto, 2009). These regions are characterized by a strong specialization in well-established (and often outdated) industrial and technological trajectories. They tend to be ‘myopic’ to opportunities that lay beyond existing growth paths and face challenges to switch to new ones (Grabher, 1993; Maskell and Malmberg, 2007).

Thick and specialized RISs of OIAs lack the internal diversity of industries and supporting organizations that are considered as being critically important for new path development (Asheim et al., 2011; Boschma and Frenken, 2011). Due to the dominance of a few clusters, the degree of intra-regional related variety is low and only few opportunities for combining or recombining diverse knowledge bases at the regional scale exist (Boschma, 2015). In other words: the availability of local assets is lower when compared to diversified RISs and, as a consequence, the need for exogenous sources of path development is more pronounced.

However, thick and specialized RISs seem to exhibit a rather low capacity to attract knowledge from outside the region. This particularly holds true for internationally-mobile, highly-skilled people such as entrepreneurs, scientists, and other members of the creative class (and to some extent also for firms and other organizations). These actors are usually lured to places characterized by diversity, openness and excellence in emerging fields. Thick and specialized RISs of OIAs hardly offer such assets, and as a consequence their capacity to attract these exogenous sources of new path development is limited. The capacity to absorb knowledge that has been explored elsewhere is also lower in this RIS type when compared with thick and diversified RISs. Absorption capacity is confined to knowledge assets that are similar or closely related to the region’s narrow knowledge base. Due to the strong specialization of industrial and support structures the capacity to absorb and exploit unrelated knowledge for new path development is rather weak.

2 Arrival of new actors, global knowledge links and new path development

The arguments raised above, however, do not imply that this RIS type is cut off from any inflow from people, organizations and knowledge from outside the region. Analyses of OIAs demonstrate that under certain circumstances the entry of people and organizations can induce regional industrial change. Work on the role of FDI in OIAs has shown that the arrival of foreign companies can open up new development paths, if these firms feature high value-added functions and embed themselves in the local economy by forming long-term linkages to regional suppliers and partners. This has been observed in Styria (Austria), where the arrival of MNCs has fuelled the creation of a new path in the automotive industry (Trippl and Tödtling, 2008). However, new path development based on FDI can be vulnerable if the big players decide to leave the region (MacKinnon, 2012). Finally, recent evidence on new path creation processes in the offshore wind industry in the UK regions also suggests that ‘transplantation’ of new knowledge from outside the region has played a major role (Dawley et al., 2015). This holds true in particular for Scotland, where

transplantation is a key mechanism of path creation in the…wind sector, providing some parallels with the development of other recent growth sectors such as electronics and oil and gas, although…transplantation extends to research and development in offshore renewables, in contrast to the branch plant nature of electronics in particular. (Dawley et al., 2015: 11)

Importantly, this case also yields interesting insights into the role of regional, national and supra-national policy actors and influences in new path development. Dawley (2014) and Dawley et al. (2015) show that the interplay of policy interventions at multiple scales, ranging from policies toward attracting firms and investments from elsewhere, to horizontal policies, sector-based vertical policies, and local development initiatives, has essentially shaped path development activities in the Scottish wind sector. Similar findings have been reported for the Styrian automotive industry. Its rise and evolution has been influenced by the attraction of FDIs with the help of public subsidies combined with regional policy measures. The latter included the establishment of a technical college and various research institutes as well as a cluster policy initiative, enhancing collaboration between foreign firms and local suppliers (Trippl and Otto, 2009). The examples discussed above reveal that under certain circumstances new paths can be developed in OIAs, triggered by extra-regional actors and linkages and facilitated by diversification processes of the organizational support structure of the RIS.

Thick and specialized RISs found in OIAs may also benefit from being connected to distant knowledge sources through various types of linkages. It has often been argued that these regions are characterized by inward-looking networks that close them off from extra-regional knowledge and reinforce path extension at the cost of new path development (see, for instance, Grabher’s (1993) seminal contribution on the Ruhr area in Germany). Trippl and Tödtling (2008) claim that new path development in these regions requires the establishment of new networks and socio-cultural shifts. Work on renewal processes in the Styrian metal industry (diversification from mass products towards manufacturing of high-quality forgings for the aviation and space industry, ultra-long weld-free rails) clearly shows that these modernization and branching processes were facilitated by a reconfiguration of local networks and new connections to actors outside the region (Trippl and Otto, 2009). Collaborative linkages in R&D and innovation have been established with distant firms and research institutes to get access to complementary knowledge assets not available in the region. This case offers insights not only into the link between new path development and extra-regional knowledge connections in OIAs but also into policy initiatives and their selective impact. National and regional policy actors have supported the formation of these connections through a variety of initiatives providing incentives for knowledge exchange in formal R&D and innovation projects and collaboration in jointly run competence centres. However, it is important to note that it has first and foremost been larger firms who have established such non-local knowledge connections, whilst smaller ones have been found to be more confined in their capacities to search for and absorb knowledge from distant sources. Policy programmes set up to stimulate path modernization and path branching activities in Styria’s metal industry through non-local knowledge linkages have thus mainly benefitted larger firms (Trippl, 2004).

The empirical examples presented above suggest that new path development in thick and specialized RISs of OIAs often requires the inflow of extra-regional knowledge. In contrast to thick and diversified RISs, there is little potential to combine different types of local knowledge and resources outside the existing specializations. Hence, accessing such knowledge externally will be crucial for new path development. Such processes imply that new non-local knowledge relates to and complements existing industrial specializations and contributes to the diversification of thick and specialized RISs.

1 Needs and capacities to attract and absorb non-local knowledge

Organizationally thin RISs exhibit a lack of organizations with the required knowledge bases and resources to engage in localized learning processes. In contrast to thick RISs, they do not have a critical mass in any specific industry, let alone opportunities to benefit from variety in the regional industrial structure. Arguably, a few strong and innovative firms may exist in thin RISs (Shearmur and Doloreux, 2016); however, they lack local partners with whom to exchange knowledge (Grillitsch and Nilsson, 2015). Often such regions are dominated by SMEs active in traditional and resource-based industries. Innovation activities tend to be low and typically of an incremental type, improving existing processes and products. Furthermore, thin RISs have per definition deficits as regards knowledge and support infrastructures such as universities, R&D institutes, or technology transfer centres. The endogenous potential for new regional industrial path development tends to be low and the need for exogenous sources high when compared to other RIS types.

The capacity to attract external knowledge is rather constrained in such regions. Investments in thin RISs often relate to natural resources, or cheap labour and land. However, resource-seeking investments do not require embedding in the local milieu and exchanging knowledge with local actors (Dunning and Lundan, 2008). Moreover, path dependencies in global production networks may generate lock-ins in peripheral regions (MacKinnon, 2012). Typically, the relationships are highly asymmetrical where regions offer rather ubiquitous resources such as low-skilled labour and consequently have limited bargaining power vis-à-vis the multinational enterprise investing in the region. The exploitation of global knowledge linkages tends to be hampered due to the strength of bonding social capital and the relative homogenous knowledge bases and worldviews (Westlund and Kobayashi, 2013). Too much bonding social capital and region-mindedness lower the likelihood for actors to search for complementary knowledge globally (Fitjar and Rodríguez-Pose, 2011; Malecki, 2011).

Organizationally thin RISs are thus characterized by a high need for exogenous sources but low capacities to attract and absorb knowledge generated elsewhere. However, this does not imply that new path development is impossible in this RIS type. Empirical studies provide evidence for path creation in peripheral areas, arguing that the absence of a critical mass of strong actors may offer geographical niches for new paths and room for experimentation (Simmie, 2012) and underlining the importance of exogenous stimuli (Varis et al., 2014; Virkkala, 2007; Isaksen and Trippl, 2016b). The questions are then how exogenous knowledge could be attracted to thin RISs and why such RISs could simultaneously build up the regional knowledge bases and absorption capacity that would allow for the emergence of new paths and the transition from a thin to a thicker and specialized RIS.

2 Arrival of new actors, global knowledge links and new path development

Mudambi and Santangelo (2016) study the arrival of MNEs in peripheral regions and highlight the importance of building up local knowledge bases over time. Their starting point is that even in the periphery some untapped knowledge resources may exist, typically bundled within a strong organization. The first MNE to move to a specific peripheral region typically acquires a strong local firm and thereby gains control over the most attractive local resources. The first mover provides legitimacy to the location. However, as the resource pool is shallow and the most attractive resources bound by the first mover, the location’s knowledge pool becomes less attractive for second movers. Second movers may still consider it valuable to enter the peripheral region, e.g. in order not to miss out on opportunities to supply to the first mover. The investments of the first and second movers, spillovers between them and potentially few local firms, combined with public investments in the knowledge and support infrastructure, may increase the attractiveness of the location over time and thereby lead to the emergence of a new regional growth path related to the industry of the first mover (and acquired firm). This would then correspond to a transition from a thin RIS to a thicker (and specialized) one.

Isaksen and Trippl (2016b) also show that the arrival of new actors can be essential for new path creation in peripheral regions. In Mühlviertel, one of the least developed regions in Upper Austria, the settlement of research institutes has over time led to spin-offs, the attraction of other research centres and firms, as well as the establishment of publicly-funded support infrastructure. In Arendal-Grimstad, located in the southern part of Norway, two pioneering firms in the electronics sector were established in the early 1960s. Starting with simple contract production, these firms transformed and upgraded over time to automated production lines and advanced software production. The key firms went through merger and acquisition processes, as well as downscaling of foreign-owned subsidiaries. Against the backdrop of these dynamics, a supporting infrastructure was established which included new study programmes, a technology park and an incubator, which created a favourable environment for spin-offs. The two cases illustrate how new growth paths emerge in peripheral regions, which over time change the character of the RIS, i.e. thicken it in terms of the knowledge and support infrastructure and industrial dynamics.

As regards the arrival of individual actors, scholarly work largely ignores thin RISs. According to Florida (2003), highly-qualified individuals tend to cluster in metropolitan areas. However, creative class does not only exist in thick and diversified RISs (Petrov, 2007). Literature on transnational diaspora entrepreneurs illustrates that mobility of individuals can lead to new industrial path development in emerging markets (Riddle et al., 2010; Saxenian and Sabel, 2008). Such entrepreneurs bring not only knowledge to the region but are also embedded in global networks that facilitate, for instance, access to venture capital or access to global markets.

As regards informal networks, Saxenian and Sabel (2008) highlight the importance of social ties to emigrants. While traditionally often regretted as braindrain, the authors attribute the development of Taiwan into a global leader in technological entrepreneurship focused on semi-conductor and computer-related firms largely to the sourcing of knowledge from expatriates, in particular the Taiwanese diaspora in the US. Accordingly, knowledge exchange through these informal linkages was instrumental for informing the responsible government officials about the venture capital industry and supporting the establishment of such an industry. This created a favourable environment for technological entrepreneurship, also attracting finance, FDI and return migration of qualified individuals. This example unveils how informal and formal networks and the arrival of external actors, together with strong government intervention, were fundamental for transforming a thin RIS into a thicker one.

In summary, organizationally thin RISs in general are characterized by the highest need and the lowest attractiveness and absorptive capacity of all three RIS types. However, the stylized facts ignore the heterogeneity of regions and lack a dynamic perspective that highlights the upgrading of the knowledge base in peripheral regions. Such a dynamic perspective allows for the possibility that thin RISs become thicker over time and transform into dynamic places for particular industries. The seeds for the creation of new industrial paths in thin RISs are often found within one or few strong organizations with global knowledge linkages. Furthermore, the arrival of external actors can be the nucleus for a new path. However, new path development also requires building regional absorption capacity over time. This encompasses upgrading the local knowledge and support infrastructure with strong policy support.