Innovation policy for system-wide transformation: The case of strategic innovation programmes (SIPs) in Sweden

Research Policy xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Research Policy journal homepage: www.elsevier.com/locate/respol

Innovation policy for system-wide transformation: The case of strategic innovation programmes (SIPs) in Sweden Markus Grillitscha,b, Teis Hansena,b,c, Lars Coenenb,d, , Johan Miörnera,b, Jerker Moodyssone ⁎

a

Department of Human Geography, Sölvegatan 10, 223 62, Lund, Sweden CIRCLE, Center for Innovation, Research, and Competence in the Learning Economy, Lund University, P.O. Box 117 S-22100, Lund, Sweden c NIFU, Nordisk institutt for studier av innovasjon, forskning og utdanning, Postboks 2815, Tøyen, 0608, Oslo, Norway d Melbourne Sustainable Society Institute, University of Melbourne, Parkville, VIC, 3010, Australia e Jönköping International Business School, Jönköping University, Gjuterigatan 5, Box 1026, 551 11, Jönköping, Sweden b

ARTICLE INFO

ABSTRACT

JEL classifications: O38 B52

The orientation towards grand societal challenges can be seen as a new wave or paradigm for innovation policy. Such policy aims at system-wide transformation and is often referred to as system innovation policy. While insights from transition studies have provided novel and useful rationales for innovation policy targeting systemwide transformation, it remains unclear how to design, implement and evaluate such policies. The contribution of this paper is to translate and concretize the challenges of system innovation policy towards scope for policy action and analysis. Building on insights from transition studies we group the challenges into four domains: directionality, experimentation, demand articulation, and policy coordination and learning. We relate challenges within the four domains to three generic features of innovation systems: interests and capabilities of actors, networks, and institutions. The derived framework is applied in a case study on the strategic innovation programmes, a recent policy initiative by Vinnova, Sweden’s Innovation Agency, targeting system innovation.

Keywords: System innovation Innovation policy Transitions Transformation challenges Innovation systems Structural change

1. Introduction The past 50 years have witnessed significant changes in the principles and practices of innovation policy. Schot and Steinmueller (2018) have recently suggested three historical framings. Innovation policy 1.0 has been primarily directed to research and development (R&D) based innovation, drawing on a linear model of innovation that privileges the technological discovery process. Innovation policy 2.0, underpinned by the systems of innovation approach, acknowledges a broader knowledge base for innovation, supports commercialization of knowledge and seeks to strengthen the link between discovery and application of knowledge. Innovation policy 3.0 involves the explicit mobilization of science, technology and innovation for meeting societal needs. It addresses the issues of sustainable and inclusive societies at a more fundamental level than previous framings or their associated ideologies and practices. Drawing on socio-technical transition theory it explicitly calls attention to the directionality of innovation. For Schot and Steinmueller (2016) an urgent shift to innovation policy 3.0 is needed to future-proof our societies and economies in a time of rising inequality, climate change, and growing unemployment.

Similarly, the Organisation for Economic Co-operation and Development observes that “by and large, most innovation policies aim to foster incremental change; fostering wider system change is a new challenge for innovation policy makers, especially as many of the actions will fall in areas outside the direct remit of research ministries or innovation agencies but where their input, coordination and implementation actions will remain critical” (OECD, 2015, p.9). However, OECD’s response, advocating so-called system innovation policy understood as “a horizontal policy approach that mobilises technology, market mechanisms, regulations and social innovations” (OECD 2015, p.7), falls short in recognizing the importance of reconnecting horizontal and vertical policies in order to achieve system-wide transformation (Weber and Rohracher, 2012) and in identifying tangible ways in which to implement system innovation policy or innovation policy 3.0 (Kuhlmann and Rip, 2014; Schot and Steinmueller, 2016). The aim of this paper is to address both gaps by (i) developing a framework for empirical analysis of concrete challenges to design and implement innovation policy 3.0, and (ii) illustrating the use of the analytical framework in the context of the strategic innovation programmes (SIPs), a recent policy initiative by Vinnova, Sweden’s

Corresponding author at: Melbourne Sustainable Society Institute, University of Melbourne, Parkville, VIC, 3010, Australia E-mail addresses: [email protected] (M. Grillitsch), [email protected] (T. Hansen), [email protected] (L. Coenen), [email protected] (J. Miörner), [email protected] (J. Moodysson). ⁎

https://doi.org/10.1016/j.respol.2018.10.004

0048-7333/ © 2018 Published by Elsevier B.V.

Please cite this article as: Grillitsch, M., Research Policy, https://doi.org/10.1016/j.respol.2018.10.004

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

Innovation Agency, targeting system-wide transformations. The contribution of this paper is to move from largely theoretical and conceptual insights prevailing in our understanding of innovation policy 3.0 to generate empirical insights about the concrete and tangible challenges in designing and implementing policies that aim at systemwide transformations. To do so, the suggested framework relates key challenges identified in transition studies, namely directionality, experimentation, demand articulation, and policy learning and coordination, to three generic features of innovation systems: (1) interests and capabilities of actors, (2) networks, and (3) institutions. There are several reasons for combining insights from transition studies with the generic features of innovation systems. Firstly, we expect that the three features of innovation systems will assist in making the rather abstract notions of directionality, experimentation, demand articulation, and policy learning and coordination more accessible for empirically oriented policy analysis. Secondly, following the suggestion by Weber and Rohracher (2012), rigorous empirical studies should investigate and scrutinize whether the four challenges really represent a problem or not. Thirdly, a merit of the innovation system approach is its standing and ready application to the policy domain. By discussing what the transformation challenges imply for actors, networks and institutions, the distinct features of designing and implementing innovation policy 3.0 can be unveiled. The proposed analytical framework will be empirically illustrated with reference to two initiatives in the strategic innovation programme, BioInnovation and Re:Source, aiming for transformative change. This paper proceeds as follows. The next section develops the framework for analysing the design and implementation of innovation policy 3.0. Section 3 describes the research context and methodology. Section 4 discusses the results of the empirical case and Section 5 ends with conclusions.

articulation, and policy learning and coordination. Directionality here highlights the need for moving beyond a focus on “generat[ing] innovations as effectively and efficiently as possible, but also to contribute to a particular direction of transformative change” (Weber and Rohracher, 2012 p. 1042). Experimentation points to the role of trying out new technologies and associated practices with a focus on learning about the possibilities for overcoming structures that inhibit their diffusion (see also Sengers et al., 2016). Demand articulation addresses the importance of market uptake and underlines the necessity of learning about user needs. Finally, policy learning and coordination draws attention to policy coherence and consistency across fields and levels, as well as the necessity of reflexivity in and associated adjustments of policy processes. In the following sub-sections, a literature review is presented with the aim of identifying concrete challenges that are associated with the four domains. Furthermore, we operationalize these challenges for each of the three generic features of innovation systems: actor interests and capabilities, networks, and institutions. These generic features follow logically from the following generic definition of an innovation system, understood as networks of actors and institutions that develop, diffuse and use innovations (Markard and Truffer, 2008). Initially, the innovation system approach highlighted actors’ innovation capabilities encompassing the diverse set of competences and resources at the firm and organizational level that is used to generate, access and exploit knowledge (Edquist, 2005). Particularly in a context of innovation policy 3.0 it is, however, important to consider not only capabilities but also the strategic and political interests that guide actors in their innovation activities (Bergek et al., 2008). Acknowledging the interactive nature of innovation, the innovation system approach suggests that actors develop networks through trade, cooperation, lobbying, and other forms of interactions that allow for exchange of knowledge and related resources (Powell and Grodal, 2005). Finally, the approach incorporates institutions as a critical system component because the collective behavior of actors is enabled and constrained by formal and informal rules, often specified as regulations, norms and routines (Edquist and Johnson, 1997; Scott, 1995). The resulting analytical framework (Table 1) identifies which challenges are to be expected in the design and implementation of innovation policy for system-wide transitions based on the theoretical arguments and empirical evidence provided in the literature. We code the challenges and use the codes in the analysis of the empirical material in Section 3.

2. Conceptual foundations Conceptually, innovation policy targeting system-wide change has been heavily influenced by the burgeoning literature on socio-technical system transitions (Geels, 2005). Socio-technical systems are configurations of institutions, competences, resources, industry structures, infrastructures, and technologies that define patterns of production and consumption in relation to societal functions (e.g. transport). Owing to the alignment of the various elements to each other, such configurations are difficult to change and constitute dominant regimes. What Schot and Steinmueller (2016) frame as innovation policy 3.0 targets the radical change of socio-technical systems and thus the establishment of new patterns of production and consumption. The challenges associated with such policies deviate largely from traditional innovation policy approaches. Traditional innovation policy was legitimized by either market or system failures (Laranja et al., 2008), which are associated with innovation policy 1.0 and 2.0 respectively. Various advances have been made that suggest how the existing innovation system failure framework can be extended to address not only system improvement but also system-wide change (Kivimaa and Kern, 2016; Weber and Rohracher, 2012). The emphasis on “failures” may, however, be questioned as it discounts more pro-active, purposive and developmental approaches to innovation policies (Asheim et al., 2011; Mazzucato, 2015), which are foregrounded in innovation policy 3.0 (Schot and Steinmueller, 2016). To address this critique, we replace system failures with a focus on the challenges to system transitions. Synthesizing the frameworks provided by Weber and Rohracher (2012), Stilgoe et al. (2013), and Schot and Steinmueller (2016), we group these challenges in four domains: challenges related to directionality, experimentation, demand

2.1. Directionality Shifts in socio-technical regimes cannot be understood without actors who initiate and engage in change processes. Garud and Karnøe (2003) remind us that the driving force for new development paths involves distributed and embedded agency. Fundamental regime shifts rest on the engagement of many actors who propel institutional change (Dawley, 2014). Institutional entrepreneurs understood as actors who initiate changes that lead to a divergence from existing institutions (Battilana et al., 2009) are thus essential for providing directionality. Defining characteristics of institutional entrepreneurs are intentionality, the pursuit of a specific interest, and the capability to mobilize required resources (DiMaggio, 1988). Intentions and interests, however, evolve over time as institutional entrepreneurs, understood as reflective change agents, promote change while, at the same time, adapt to changes in their environment (Sotarauta and Pulkkinen, 2011). A key challenge is thus to promote institutional entrepreneurship directed towards the transformation and shift of socio-technical regimes (DiA1).

2

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

Balance attention to supply- and • DaI1: demand-side policy instruments Promote social acceptance for emerging • DaI2: technologies

Stimulate interaction between • DaN1: producers and lead users

Develop shared vision among multiple • DiI1: actor groups Set objectives that provide direction in a • DiI2: concrete and actionable way Institutions (I)

and globally

Networks (N)

Actor interests and capabilities (A)

Table 1 Analytical framework.

Directionality (Di)

Promote institutional entrepreneurs • DiA1: Resolve conflicting interests due to • DiA2: skewed distribution of power and resources Develop capabilities in new forms of • DiA3: governance Connect and integrate directionality • DiN1: exercised by multiple types of actors, locally

Experimentation (Ex)

Stimulate entrepreneurship • ExA1: Support development of new capabilities • ExA2: in incumbents Promote an interest in experimentation • ExA3: among non-firm actors Encourage collaboration between • ExN1: heterogeneous actors Assist new actors in entering • ExN2: collaboration networks Support test and demonstration projects • ExI1: Gradually increase exposure of • ExI2: experiments to selection pressures Promote risk-taking behavior and • ExI3: acceptance of failure

Demand articulation (Da)

Support identification of lead users • DaA1: Develop innovation procurement • DaA2: capabilities in public bodies

Policy learning and coordination (Pl)

Exercise leadership across policy • PlA1: domains Overcome conflicting interests • PlA2: Develop governance learning • PlA3: capabilities Widen the scope and diversity of • PlN1: policy networks • PlN2: Challenge established hierarchies Break with existing policy rationales • PlI1: • PlI2: Incentivize diverging policy trajectories

Institutional entrepreneurship is conditioned by the different types of power agents have and use in the pursuit of their interests. Sotarauta (2009) distinguishes between formal powers to change and create new institutions or commit resources as well as network and interpretative power. Network power enables actors to draw on distributed resources, to control and facilitate the flow of information, and to build trust among partners. Interpretative power relates to the ability to alter or create new meanings and interpretations, as well as to articulate potential future visions. Smith et al. (2005) argue that the ability and power to intervene in socio-technical regimes depends on regime membership and the distribution of resources. Regime membership captures broadly all actors that contribute to reproducing regime functions. Some members will be more central than others and have consequently more power to promote or prevent change. In the context of socio-technical regimes, institutional entrepreneurs should thus be able to deal with a skewed distribution of power and resources between different members in a regime, between strong incumbent and weak emergent players, and between different transition pathways (DiA2). Actors also need to acquire new competences because system innovation policies require different forms of governance as compared to traditional innovation policies (DiA3). Loorbach (2010) emphasizes the ability to cope with a high level of uncertainty, to balance short-term actions with long-term visions, and to involve, interact, experiment and learn with a variety of actors. Furthermore, new competences are also required owing to the specific functions of actors in promoting system innovations, which differ from or complement traditional functions as becomes apparent in the case of states and higher education institutes. The state provides directionality through its capacity to establish niches (Dawley, 2014), facilitate collective learning processes (Rotmans et al., 2001) and fund research, development and education (Mazzucato, 2015; Tanner, 2014). Higher education institutes contribute as potential change agents by, among other things, modelling sustainable practices for society, by introducing students to system-thinking and educating them to solve complex problems, and by promoting exchange and interactive learning within and outside the higher education sector (Stephens et al., 2008). As regards network, a key challenge is to connect and integrate directionality exercised by multiple types of actors locally and globally because socio-technical regimes are embedded in highly complex networks that are erected at different spatial scales and stretch across different institutional domains (Geels, 2002) (DiN1). Comparing the development of wind turbine technology in Denmark and the US, Garud and Karnøe (2003) illustrate the importance of networks in the process of new path creation. The authors attribute the Danish success to interactive and cumulative learning processes between producers, users, evaluators, and regulators. This promoted quick feedback, step-by-step improvements of a simplistic design from the 1950s, and the co-evolution of a supportive regulatory environment in relation to R&D funding and certification. Globalization of production and the global nature of grand challenges imply that global actor networks play an important role in providing directionality. Global actor networks consist of, among others, intergovernmental organizations, states, transnational cooperations and global civil society organizations (Dicken, 2011). Negotiations of global climate objectives and treaties, orchestrated by intergovernmental organizations, involve nation states but also, indirectly, interest groups including fundamental environmentalists as well as climate change deniers (Raskin et al., 2002). At a more operational level, Rock et al. (2006) show how directionality can work in concert between global civil society organizations, transnational cooperations, and local actors, and how the diffusion of environmental standards in global production networks depends on interactions between these different types of actors. As regards institutions, shared visions are considered essential for providing directionality (Weber and Rohracher, 2012) and feature prominently in transition management approaches (Loorbach, 2010; 3

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

Rotmans et al., 2001) (DiI1). A shared vision contributes to system innovation by identifying plausible future scenarios, by formulating the technical, institutional, and behavioural problems that are to be solved, by providing a stable point of reference for target setting and monitoring, and by providing a metaphor that can unite different actor groups and focus capital and resources (Smith et al., 2005). A shared vision forms when core elements of the visions in different sectors and social groups align, i.e. when certain values, norms and strategic objectives about future pathways converge (Raven et al., 2010). The challenge thus is to align the interests of broad stakeholder groups to agree on a shared vision and objectives that provide direction in a concrete and actionable way (DiI2). Hodson and Marvin (2010) remind us that this is usually a messy task due to multi-level governance arrangements. Shared visions do not necessarily (and perhaps hardly ever do) emerge from a broad consensus about how a transition towards a sustainable future should take form but may be the negotiated result between a group of self-interested parties. Accordingly, shared visions need to be evaluated carefully, asking questions about whose or which social interests get promoted on the ground of which expectations (ibid).

importance in experimentation activities, Schot and Geels (2008) also acknowledge that too much diversity may make it difficult to reach consensus and lead to resource fragmentation. In terms of formal institutions, government policies may provide room for experimentation in multiple ways, directly in the form of funding and other forms of support for test and demonstration projects (ExI1) (Frishammar et al., 2015; Klitkou et al., 2013; Nill and Kemp, 2009; van der Laak et al., 2007) and indirectly through support for entrepreneurship and firm diversification (Feldman et al., 2011; Neffke and Henning, 2013). The strategic niche management literature stresses the importance of issues associated with insufficient policy support for experimentation, but also that too large or too long policy protection for specific experiments will lead to lack of incentives for eliminating negative side-effects and, potentially, expensive failures (Kemp et al., 1998; Schot and Geels, 2008). To exemplify, Nill and Kemp (2009) mention the case of the Dutch wind turbine industry where producers were not incentivized to develop internationally competitive designs because of generous policy support (see also Kamp, 2002). Thus, while policy support for experimentation is important for transformative change, experiments also need gradual exposure to selection pressures (ExI2). Informal institutions may also lead to challenges related to experimentation. Engagement in experimentation activities is influenced by cultural aspects such as risk-taking behaviour, acceptance of failure and esteem for entrepreneurialism (ExI3), which have been found to vary significantly between innovation systems (Efrat, 2014; Shane, 1993; Turró et al., 2014). To exemplify, Näyhä and Pesonen (2014) demonstrate that transformative change in the forestry industry in Scandinavia and North America is significantly hampered by a conservative organizational culture, which constrains experimentation with new technologies and business models.

2.2. Experimentation Regarding actors, an underlying explanation for challenges related to experimentation may be a shortage of entrepreneurship (ExA1), since experimentation with new technologies, markets and organizational forms is central to the activities of entrepreneurs (Stern, 2006). Historically, the experiments of entrepreneurs have played significant roles for transformative change, e.g. Edison for electricity distribution (Hargadon and Douglas, 2001). Thus, new entrants are arguably the most important sources of disruptive innovation since incumbents are bound by existing organizational routines and values, and profit from current technologies (Chandy and Tellis, 2000; Leonard-Barton, 1992). However, some incumbents do still engage significantly in experimentation efforts (Rosenbloom, 2000; Roy and Sarkar, 2016), and incumbents may potentially play an important role for transformative change owing to their access to capital and technical capabilities (Chandy and Tellis, 2000; Hockerts and Wüstenhagen, 2010). In particular, incumbents with loosely coupled, stand-alone divisions focused on experimentation activities may make important contributions to transformation processes if these new divisions have sufficient resources and authority (Chang et al., 2012; Hansen and Coenen, 2017; Hill and Rothaermel, 2003). Thus, the absence of intrapreneurs, i.e. individuals working in an entrepreneurial manner within existing organizations, and new divisions in incumbents (ExA2) may also lead to insufficient experimentation. Finally, recent work on urban sustainability experiments highlights the core role played by, in particular, local government actors, but also community-based organizations in leading these efforts (Bulkeley and Broto, 2013; Bulkeley et al., 2011). Consequently, challenges related to experimentation may also be caused by a lack of engagement of non-firm actors (ExA3). On the network-side, predominance of relations between homogenous actors (ExN1 and ExN2) is a main underlying explanation for insufficient experimentation. Experimentation requires not only interaction and collaboration across knowledge fields (Lyall et al., 2013; Rekers and Hansen, 2015), but also between firms, users, policymakers and interest groups (Coenen et al., 2015; Geels and Raven, 2006). In line with this, the strategic niche management literature stresses the importance of networks that are broad (covering multiple types of stakeholder, including traditional outsiders) and deep (involving actors who can mobilize resources) (Schot and Geels, 2008). However, while the core argument is that diversity in actor-constellations is of primary

2.3. Demand articulation On the actor-side, an underlying explanation for challenges related to demand articulation may be a lack of lead users or users that modify existing usage practices (DaA1) (Schot et al., 2016). By definition, lead users have needs that are radically different from existing offerings on the marketplace (Von Hippel, 1988) and they may destabilize existing socio-technical arrangements by inventing, co-developing and testing radical innovations. Thus, while the actions of regular users will generally stabilize socio-technical regimes (see e.g. Shove et al., 2014), the presence of lead users may have the opposite effect. For instance, the initiation and development of the car-sharing niche in Switzerland was fully led by users (Truffer, 2003). Lead users can commercialize radical ideas on their own or expedite transformative change if firms are able to benefit from lead users’ needs. However, this requires significant changes in the capabilities of firms’ marketing departments (Von Hippel, 1986). An additional possible underlying explanation for challenges related to demand articulation on the actor-side is the lack of competence in innovation procurement among public sector actors (DaA2). Innovative public procurement can be understood as “the purchasing activities carried out by public agencies that may lead to innovation” (Guerzoni and Raiteri, 2015, p. 729). While the main rationale of public sector procurement has traditionally been lowering of prices, procurement focused on stimulating innovation for higher-order public-policy goals such as sustainability necessitates significant changes in the needed competences of procurement officers (Edler and Georghiou, 2007), who for instance must learn to procure using functional requirements (Edquist and Zabala-Iturriagagoitia, 2012). On the network-side, a main underlying explanation for the lack of

4

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

uptake of radical innovations in markets is a lack of user–producer interaction (DaN1) (Gertler, 1995; Lundvall, 1988). The absence of such relations between niche firms and their customers may hinder transformative change. Conversely, intense user–producer interaction enables bidirectional flows of information and feedback leading to market establishment for emerging technologies such as solar photovoltaics (Dewald and Truffer, 2012). Regarding institutions, a variety of formal institutions influence demand articulation by, for instance, subsidizing private demand for specific goods and services (e.g. through feed-in tariffs). However, overemphasis on technology-push policies may hinder innovation diffusion and their impact on society (DaI1). Demand articulation challenges may also stem from characteristics of informal institutions. While proven technologies have built up legitimacy among users over decades, emerging technologies need to attain such social acceptance (DaI2) (Bergek et al., 2008). However, informal institutional differences may lead to significant divergence in the possibilities for emerging technologies to appear as desirable and convincing substitutes for existing solutions. To exemplify, variations in professional cultures among farmers have been found to explain differences in the extent and character of biogas technology diffusion (Wirth et al., 2013).

governance learning reflects major reflexive and institutional capacity going beyond actor networks or groups of actors. While the first two dimensions have played an important role already in the innovation system perspective, the third dimension adds a new level of learning that targets reframing the role of innovation and existing systems of production and consumption. On a related note, with regard to network dynamics, the abilities for learning are partly dependent on the scope of the networks and partly on their composition in terms of internal hierarchies (Battilana, 2011; Burt, 2005). The more closed and localized the networks, the higher the risk for lock-in and weak learning capabilities because such networks are not exposed to alternative knowledge, norms and routines. Furthermore, in such closed networks actors with a strong power position remain largely unchallenged, which make vested interests a more serious concern than in open and diverse networks (PlN1). Divergent trajectories, which are basic preconditions for system transformation, are thus in many cases suppressed before they even start because of such vested interests and skewed distribution of power in the networks (PlN2). Conformism and weak policy learning capabilities may be challenged or further underpinned by the institutional structure in which these actors and networks are embedded. One of the major institutional challenges is to break with existing policy rationales (PlI1). Morgan (2016) reminds us that innovation policy rationales are engrained in the political discourse and stabilized in policy repertoires. Regulations, norms, standards and routines that influence economic development are based on established logics, which indicates that those institutions that evolve in economies tend to foster stability and continuity rather than change and disruption (Glückler and Bathelt, 2017; Martin and Sunley, 2006). This implies that actors, who may be initially inclined to contribute to the transformation of existing systems of production and consumption, are pushed back into conventional routes by institutional pressures. This refers to institutional hysteresis (Setterfield, 1993) where policy learning capabilities are hampered both by blocking alternative policy rationales and by providing incentives for maintaining status quo through conformist strategies. For this reason, it is a challenge to incentivize diverging policy trajectories (PlI2).

2.4. Policy learning and coordination It is well documented that failure to learn and adapt in the support structure of an innovation system may render a situation of institutional and political lock-in (Grabher, 1993; Hassink, 2010), which in turn hampers transformation and emergence of new growth paths (Lundvall, 2010). Weber and Rohracher (2012) underline the importance of both vertical and horizontal coordination in order to ensure “coherent policy impulses from different policy areas in order to make sure that indeed the necessary goal-oriented transformative changes for tackling major societal challenges can be achieved” (p.1043). Previous studies on regional economic transformations have shown that this requires a broad involvement of actors beyond the level of those that traditionally have been perceived as policymakers (Cooke and Morgan, 1994). While adaptive policy may handle lock-in situations in a reactive manner, the proponents of policy learning suggests a proactive approach in which policy learning actually drives transformation through changes in the collective perception of change agents (Bennett and Howlett, 1992; Borrás, 2011). Among the main challenges to such learning and coordination are lack of or weak leadership (PlA1), as well as conflicting interests among actors (PlA2) and, consequentially, weak coordination capabilities. Drawing on a three-fold typology proposed by Bennett and Howlett (1992), Borrás (2011) makes a link between policy learning and organizational capacity by suggesting three intertwined levels of policy learning (PlA3). The first, government learning, relates to learning and coordination within and across public government organizations. The second, network learning, refers to collective learning taking place in the support structure of an innovation system through broadening of the competence base and new combinations of experiences and practices, including stakeholders such as firms, universities and non-governmental organizations. The third dimension, governance learning, encompasses state–economy–civil society relations broadly understood. This dimension refers to the emergence of new policy paradigms. All three types of policy learning and coordination play essential roles for system innovation, although in complementary ways. Government learning reflects administrative capacity in the support structure and policy domain (e.g. effectiveness, professionalization), network learning reflects analytical capacity among actors throughout the system (e.g. absorption and adaptation to changed preconditions), and

3. Research context and methodology 3.1. The case of strategic innovation programmes (SIPs) in Sweden Sweden’s Innovation Agency, Vinnova, has during the past decade increasingly endorsed a societal challenge driven logic for policy, not only in discourse but notably also in terms of resource allocation (OECD, 2013). The Swedish strategic innovation programme (SIP) – coordinated by Vinnova in collaboration with the Swedish Energy Agency and the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning – represents a suitable case to identify and analyse the abilities and challenges connected with combining vertical and horizontal policies in order to achieve system-wide transformation (Weber and Rohracher, 2012). This is because it explicitly targets system-wide transformation and subscribes to the ambition of departing from a holistic approach, with the coordination of actors, networks and institutions presented as the very core of its policy intervention, aiming not only to increase Sweden’s international competitiveness but also addressing the grand challenges (Vinnova, 2017). Two such SIPs are used as illustrative cases for this study. Targeting different challenges and departing from different sectors, they both represent specific policy measures aimed at stimulating system transformation, and are closely linked to Sweden’s horizontal so-called innovation partnership programmes, a broad-based platform approach

5

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

contemporary social phenomenon (Yin, 2013). To reduce potential selection biases (Eisenhardt, 1989) and provide contrastable illustrations to our conceptual arguments, we do a comparative analysis of two cases. These were selected based on what could be expected regarding the potential information and knowledge that can be drawn from their study, in order to “maximise the utility of information from small sample and single cases” (Flyvbjerg, 2006 p. 230). More concretely, the cases were identified through discussions with Vinnova, who had previous knowledge about the characteristics and potential challenges faced by different SIPs, and based on an initial systematic review of existing material relating to the SIPs granted by Vinnova. A systematic review of existing material was initially conducted in order to provide material for the selection of cases, and to provide a first indication of the challenges that were explicitly perceived and acknowledged in the SIPs. Material for the review was obtained from three main sources: (i) documentary and digital material published by the SIPs through various channels, such as their websites and press releases, and information published by participating firms about their involvement in the SIPs; (ii) newspaper articles and press releases from other sources, such as funding organizations (Vinnova and the Swedish Energy Authority), participating universities and other public actors; and (iii) material provided by Vinnova and by the programme offices of the SIPs themselves, primarily in the form of application texts, earlier versions of suggested innovation agendas, internal policy documents guiding the operative work of the programmes, and internal working papers. The review allowed us to identify key individuals, not only those at the “core” of the current activities of the SIPs, but also those that were involved early but having more peripheral positions. The material was coded based on the categories suggested in the analytical framework and thus provided a first basis of the empirical analysis of the two cases. Semi-structured interviews were conducted with key actors involved in the SIPs. After the initial selection procedure, interview partners were identified using a “snowball sampling” technique (May, 2011). Interview partners were asked to give recommendations about additional interview partners that they perceived as having information that was relevant to the topic. Such recommendations were evaluated by relying on the findings from the initial document review, and on the ongoing process of analysing the interview data, in order to avoid a selection of interview partners representing one particular view. Interviews were conducted until, in combination with data from the document review, data saturation (Glaser, 2017) was reached. In total, nine interviews were conducted with key actors involved in the SIPs. The interviews were conducted during April and May 2016. Four interviews were with actors connected to Re:Source, and five interviews with representatives of BioInnovation. The interview partners represent a mix of programme managers, board members, expert team members and members of the strategic committee, for both programmes. The interviews are summarized in Table 2 showing the order of interviews, the type of actor, the SIP being represented, and the duration of the interview. The interview guides (see Appendix A) were designed as a set of theoretically informed questions centred around the analytical framework presented in Section 2. For each interview partner, we collected information about personal background and experience with innovation policy. We then asked the interview partners about the challenges they faced in the process of designing and implementing the SIPs. Initially, we asked the questions in an open manner and then we prompted the interview partners on the dimensions identified in the analytical framework. Finally, we enquired about the policy responses to the experienced challenges. All interviews were recorded and transcribed by the authors.

Table 2 List of interviews. No. Role in the SIP

Type of organization

SIP

Duration

1 2 3 4 5 6 7 8 9

SIP SIP Industry Industry Academia Academia Industry Industry Industry

Re:Source Re:Source BioInnovation Re:Source Re:Source BioInnovation BioInnovation BioInnovation BioInnovation

52 61 55 59 55 39 56 62 67

Program manager Innovation manager Board member Board member Strategic committee member Member of expert team Board member Member of expert team Strategic committee member

min min min min min min min min min

launched by the Swedish government (Government Offices of Sweden, 2016). The partnership programmes have been made a priority by the prime minister, and Vinnova has been specially tasked with assisting the work on these programmes in 2016–2018. The SIPs should thus be seen in the light of this horizontal platform policy approach, encompassing numerous actors, perspectives and measures, thereby cutting across thematically different but increasingly interdependent policy domains. As a consequence of the government’s mission for Vinnova to connect vertical and horizontal policies, the SIP targets diversification more than concentration, as opposed to the previous strong focus on place-based specialization strategies (Asheim et al., 2011; Cook and Eriksson, 2011). This reorientation of strategic focus to thematic platforms transcending sectoral, geographical and organizational domains also reflects a recent trend in European innovation policy where increased focus is geared towards addressing the grand challenges through exploiting variety and new combinations in knowledge domains (Cagnin et al., 2012; Coenen et al., 2015). Currently, 16 SIPs have been established. The remainder of this paper provides examples of system innovation challenges, as well as responses to those challenges, focusing on two SIPs: BioInnovation and Re:Source (Vinnova, 2017). BioInnovation aims at supporting a complete transition to a bio-based economy in Sweden by 2050. Core sectors involved in the initiative are the forestry, chemicals and textile industries and related stakeholders. Re:Source aims at making Sweden a world-leading circular economy minimizing and reusing waste, with particular focus on materials supply, a sustainable energy system, as well as more efficient use of resources in business and society. Due to the all-embracing scope of Re:Source, this initiative is not immediately associated with specific industry sectors in the same way as BioInnovation. The implementation of these programmes is carried through by consortia with organizations from industry as well as academia and the public sector. The explicit ambition is to promote cross-fertilization of competences and experiences by stimulating interaction and collaboration across sectoral boundaries. Both programmes are directly linked with the government’s innovation partnership programme, targeting transition to a circular and bio-based economy (Government Offices of Sweden, 2016). 3.2. Research method A case study methodology (Flyvbjerg, 2006; Yin, 2013) is used to gain in-depth insights into the challenges related to the design and implementation of innovation policies 3.0. Case studies are suitable when the research is aiming for an in-depth description of a

6

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

The interview transcripts and document materials were coded. The coding procedure took place in three steps. First, a coding guide was developed based on the initial identification of challenges in the literature review, comprising 3 × 4 parent nodes (actors-networks-institutions and the four types of challenges identified). Second, the material coded at the intersection of two parent nodes (e.g. actors and directionality) was again coded in 26 ‘child nodes’ corresponding to the challenges identified in the theoretical discussion (the list of codes is reported in Appendix B). The analysis led to the identification of empirical illustrations of challenges covering all dimensions of our analytical framework. Using the analytical framework, we identified not only the challenges perceived by the interview partners but also challenges that were not perceived or not dealt with in the SIPs. In the following section we present an integrated analysis of the two SIPs. Here we focus on the encountered challenges and the responses for the two programmes, as well as challenges that were left unaddressed by the SIPs.

to conflicts (DiA2). In the case of Re:Source, the waste recycling industry and the municipality-owned waste-handling companies compete for the same waste. The public waste-handling companies in Sweden enjoy by law a monopoly on handling waste produced by households, and thus are in a stronger position compared to private firms. Rather than confronting this, SIP representatives tend to take a passive stance and formulate objectives and actions that receive sufficient buy-in and are achievable. One interview partner expressed this as follows: “I don’t know if it is necessary for them [conflicts] to be reflected in the agenda though. I see it as a development process [..] At this moment, I don’t see any need to create watersheds” (Interview: Programme board representative).1 Such a passive stance, circumventing institutional change, is incentivized as the performance of programme managers is based on the output of the SIP. Rather than not absorbing funds and implementing fewer activities, from a programme manager’s perspective it may be preferable to promote collaborative activities that only weakly contribute to the strategic goals. Therefore, it is important to develop incentive mechanisms that promote addressing the root of the problem. In the concrete case of waste handling mentioned above, this would mean changing institutions that cause the conflicting interests (DiA1). Other challenges at the actor level include a lack of governance capabilities and forms of governance in the context of a SIP (DiA3). There is no template for how a SIP should be managed and leading actors engage in activities intended to develop best practice and enable institutional entrepreneurship to emerge bottom-up. For example, there are activities promoting learning processes across different actor groups in the consortiums, and also between different SIPs. It was stated by our interview partners that they keep informed of developments in other SIPs, for example, through workshops and meetings organized by Vinnova, and adopt practices that are proven to work. Apart from the process of formulating the agendas, activities targeting the network- and institutional dimensions are plentiful and mainly consist of various forms of data collection, both to get an overview of the current state of the innovation field in terms of existing collaborations (DiN1), and to identify common ground for enabling institutional change processes (DiI1, DiI2). For example, in one of the SIPs, the role and influence of civil society organizations have been investigated and key actors are developing plans for how to mobilize the power and influence of such organizations both to push and provide direction for some of the diverging interests within the programme and also to align existing strategies and visions with broader societal trends. It was highlighted by our interview partners that civil society organizations were often in the lead when it comes to influencing public opinion in questions affecting the programme participants. The rational for the involvement of these organizations reached far beyond legitimation activities, but rather provided vision and a way forward for the programme as a whole (DiI2). While our conceptual frame started from the assumption that institutional entrepreneurs are those taking the lead in transformation projects through deliberately challenging established rules of the game and power structures (DiA1), we found little evidence or indications of this in our study. Although there were competing agendas developed in parallel, and subsequently combined as a result of the construction and rules of the SIPs, the change processes that arose from such combinations are better described as unintentional compromises in which diverging interests were mediated (DiA2, DiN1). They consisted of practical compromises to allow concrete action within the frame of this type of state-governed innovation programme (DiI2).

4. Results: analysis of discourse about encountered challenges and programme responses 4.1. Directionality Many of the initial efforts undertaken by key actors in the case of the SIPs under investigation were related to overcoming the challenge of directionality. By design, a SIP needs to be based on a “strategic innovation agenda”, which provides the basis of a funding application submitted to Vinnova (Vinnova, 2017). The formation of a programme consortium and the formulation of innovation agendas involves challenges related to all three levels of the analytical framework. A wide range of different actors needed to be identified and mobilized to take part in the process, which led to challenges related to connecting and integrating directionality exercised by multiple types of actors (DiN1). In both cases, rather than taking the form of open discussions including a variety of stakeholders, several competing agendas were developed in parallel by actor groups, reflecting differences in vision and interests. Key individuals from research organizations, associations of public organizations and private industry brought these agendas together, with significant experience in drafting proposals for funding. They acted as “boundary spanners” and used their interpretative power in order to mediate between different interests, dealing with challenges related to conflicting interests (DiA2) and a lack of governance capabilities in developing innovation programmes addressing grand challenges (DiA3). By synthesizing the individual agendas, formulating comprehensive visions and agreeing on objectives fulfilling those visions, challenges related to the institutional level were also targeted (DiI1, DiI2). However, this happened in an ad hoc manner, and we have no evidence for a “strategic” process that ensured that the SIPs were set up in an adequate manner to ensure socio-technical system transition. However, for both cases, our analysis indicates that while the innovation agendas provide an important first step in developing directionality for the SIPs, the attempt to achieve shared visions among many stakeholders led to broad, catch-all agendas embracing the various priorities mentioned in the agendas (DiN1). Consequently, the investigated SIPs were designed as an umbrella of several agendas representing different actor interests rather than as a collectively deliberated, well-aligned and integrated programme. This is problematic because it weakens the power of the SIPs to provide clear objectives and concrete direction for the participating actors (DiI1, DiI2). Furthermore, the interpretation of interview material in combination with results from our document analysis suggests that challenges related to conflicting interests are not perceived and dealt with owing to a lack of governance capabilities and network power within the SIP consortiums. There are examples of situations where the distribution of power and resources between members of different regimes gives raise

1

7

All quotes are translated from Swedish by the authors.

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

4.2. Experimentation

networking activities, and actors can get customized support for writing joint applications and collaboration agreements. This is true particularly for BioInnovation, but similar activities are mentioned also by Re:Source interview partners. However, the focus is generally on collaboration, with an implicit belief that collaborations will lead to increased experimentation by default (ExN1). Furthermore, whilst diversity in actor-constellations is indeed highlighted as an important feature, several interview partners have mentioned the challenge of involving actors other than “the usual suspects”; there seems to be a link between already well-developed risk-taking behaviour and an interest in taking part in diverse actor-constellations (ExI3, ExN2). While industry incumbents have access to capital and technological capabilities, many are somewhat inexperienced when it comes to working in large actor-constellations in which experimentation is promoted (ExA2).

Challenges related to insufficient experimentation were found for all three levels of our analytical framework, but they differed substantially between the two SIPs investigated in our study. This reflects differences in the actors involved, in the relative importance of incumbents, small and medium-sized enterprises (SMEs) and public actors, and in the existing innovation practices in the industry. However, representatives of both SIPs highlighted challenges associated with developing new capabilities among incumbents (ExA2) and to promote an interest in experimentation among non-firm actors (ExA3). These are perceived as serious challenges by programme managers as well as participating actors, as they are crucial for the long-term success of the SIPs. “We have a thesis saying that the large dragons [..] will not be able to transform themselves, but it will be small actors that can take risks. Therefore, it is important to expose small firms or entrepreneurs so that they can grow or be acquired by the large firms. In that way, change is introduced in the large firms as well” (Interview: Programme strategist).

“The industry is not used to work in innovation projects of this type. [..] It is very different from previous things… It is very good. But it is important that it looks attractive. It is a clash of worlds” (Interview: Industry representative).

In the case of BioInnovation, interview partners suggested that experimentation was constrained because of a shortage of SMEs focused on experimenting with new technologies (ExA1), and closed collaboration networks (ExN2), which prevent the development of such firms. As a response, BioInnovation formed so-called expert teams consisting of representatives with a background in R&D from firms, industry organizations, universities and research institutes. The expert teams were established for a number of thematic areas such as “materials” and “chemicals and energy” and are intended to intervene in the formation of project consortia. Concretely, when proposals for projects are submitted to BioInnovation, the relevant expert teams may for instance suggest combining project proposals or including new actors in order to ensure that existing collaboration patterns between incumbent actors are not reproduced over time. Thus, expert teams broker to facilitate experimentation. In addition to introducing expert teams, BioInnovation also provides training to members on how to work in innovation projects involving diverse partners, since many lack such experience, and develop template agreements that are intended to guide the process of establishing such consortia. The case of RE:Source illustrates a gradual exposure to selection pressures (ExI2) by the design of the project application process. Few restrictions applied to the first call for projects, which consequently resulted in many smaller pre-projects that covered a large variety of topics being funded. Among these were several test and demonstration projects of highly experimental character, reflecting the programme management’s intention to support such projects (ExI1). Yet, according to interviewees, upcoming calls will be narrower and guided by the results of the pre-projects. Consequently, the expectation is that many fewer – but larger – projects will be funded. This will allow RE:Source to set a clearer direction in its activities, but will necessarily also limit the breadth of experimental activities. This highlights a tension between developing a strong directionality, as discussed in the previous section, and promoting experimentation among diverse programme participants (ExN1). The following quotation illustrates the change from a very open and experimental approach to a more narrow focus:

Finally, some activities also target institutional change, particularly when it comes to incentivizing risk-taking behaviour in the public sector (ExI3). It is acknowledged that, in particular in the case of Re:Source, regulations are continually changing. Activities aim to intervene in these change processes, mainly by providing information to public actors in order for them to be able to make sound, evidencebased decisions. It is explicitly stated in the programme communication that the SIP is no lobbying actor, but should produce knowledge about how to incentivize risk-taking and experimentation (ExI3). “Laws are constantly changing. It changes the playing field in this area. That is why it is very important to be part of creating the large picture, and to provide policymakers with a good basis for decisions. It should not be a lobby organization, but bring forward objective facts” (Interview: Member of programme management). While interest and ability to develop new capabilities of incumbents (ExA2) is a well-recognized challenge, the programmes under study display few efforts to deal with this challenge, which made us question to what extent this is a real or mostly a rhetorical problem. It is also worth noting, and exploring in further studies, the measures needed to promote risk-taking behaviour and entrepreneurial experimentation through inclusion of new actors in collaboration networks (ExN2). The present study did not find many concrete examples of such changes in the programme. 4.3. Demand articulation In terms of demand articulation, mainly challenges on the actor and network level have been identified and acknowledged by our interview partners. On the actor level, challenges related to developing innovation procurement capabilities in public bodies (DaA2) were addressed by both programmes, albeit in very different ways. BioInnovation has launched a call for projects specifically on bio-based innovation and public procurement, informed by analyses of the market and political conditions for bio-based products. These analyses provided an evidence-based foundation for directing resources towards the issue of procurement capabilities, with the aim of stimulating future demand for bio-based solutions. Thus, BioInnovation is directly financing the establishment of innovation procurement capabilities in public sector actors. Seven projects have been funded that are concerned with development of biobased products to, for example, the health-care sector and the construction industry. As exemplified by the following

“In the beginning, it is important for the board to make sure that the direction is broad, that we don’t dwell on things that we think are most important, but instead to allow all innovative forces in society to come forward. [..] Then, later, to narrow it down” (Interview: Programme representative). Several interview partners highlight various workshop and

8

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

quotation, it was expressed by our interview partners that these projects were a means of supporting the public sector in learning about innovation procurement in the bio-based sector (DaI1).

indirect attention in the two programmes. The acknowledgement of innovation procurement as an important area for action reflects an attempt to balance attention to supply- and demand-side policy instruments (DaI1). There are, however, no activities explicitly targeting the question of how such a balance could be reached or what would constitute an optimal balance in the context of the two programmes. Furthermore, there is no explicit acknowledgement of the importance of social acceptance for emerging technologies (DaI2). While keeping in mind the limited scale and scope of the present study, this observation calls for more attention in future studies, especially given the strong connection with institutional theory and demand-side attention in transition studies literature.

“This is a way of supporting the public sector [..], departing from a needbased perspective and, maybe not take shortcuts, but at least learn things about the complicated process around innovation procurement” (Interview: Expert team leader, BioInnovation). The iterative processes of first learning about the specific challenges of biobased procurement and then formulating tools to cope with these challenges was also highlighted in interviews (DaI1). Similarly to responses to other actor-level challenges, Re:Source targeted the lack of innovation procurement capabilities in the public sector by stimulating networking, establishing an innovation procurement platform for interaction between policymakers, industry actors and academia. While this does not explicitly target procurement capabilities per se, it has the benefit of identifying and possibly targeting other challenges, such as the identification of lead users (DaA1), and to stimulate interaction between producers and lead users (DaN1). There is, however, a risk that focus on networking and interaction tend to downplay concrete issues that would have surfaced in more targeted activities. For example, one of our interview partners highlighted concrete challenges related to procurement, but our analysis did not indicate that efforts targeting this area acknowledged the specific challenges perceived by some of the programme participants, but rather targeted the challenge more broadly. This highlights the need for coordinating activities targeting different types of transformation challenges, as the development of capabilities in public procurement may be hampered by challenges related to conflicting interests (see the following quotation) in terms of a skewed distribution of power and resources (DiA2) and conflicting interests between policy domains or spatial levels of governance (PlA2).

4.4. Policy learning and coordination The initiation of SIP as such can be seen as a recognition of policy learning and coordination challenges. The Swedish government, by lessons learned from previous experiences and experimentation, adapted its policy to strengthen leadership and bridge conflicting interests (PlA2) instead of underpinning them through further specialization into sectors and/or regions, which used to be the preferred approach in Sweden during the 1990s and early 2000s (Cooke and Eriksson, 2011). Following almost two decades in which such sectoraland territory-focused policy programmes had been gradually refined, Vinnova, through SIP, turned attention towards sector- and geographytranscending programmes unified by thematic focus, spanning sectoral and spatial domains, with an explicit challenge-based rationale for policy intervention. This shift was justified by the acknowledgement of the multi-faceted nature of the challenges in combination with the identified difficulties in achieving the integration and coordination that are necessary for addressing complex transformation (PlA1). Through this new approach, the government thus catered for both horizontal and vertical policy coordination (Weber and Rohracher, 2012). The maintained requirement of a triple helix composition of initiatives within the SIPs aims to stimulate policy learning and coordination by design. A problem, however, is that the variety of stakeholders included in the programme and project applications substantially decreases in the implementation process. This was both observed and acknowledged as a prioritized challenge by our interview partners. Again, by design, the SIPs involved a broad range of actors from different policy domains (PlN1). As a consequence, activities targeting policy learning and coordination challenges at the actor and institutional levels were plentiful, while the network level was somewhat neglected in the post-agenda-setting phase. In particular, challenges related to the existence of established hierarchies were not acknowledged by our interview partners (PlN2). On the actor level, both programmes experienced challenges with developing governance learning capabilities (PlA3), related to issues of involving industry actors in the early stages of implementation. For example, the first round of activities was largely driven by, and targeting the activities of, academic actors and institutes. This reflected the fact that while industry actors indeed had been part of formulating the agendas, they perceived the role as being of a more supportive character. As one interview partner expressed it:

“In Stockholm, in the latest procurement of domestic waste, they consider frying oil from e.g. McDonald’s as domestic waste. Then [they] have to leave it to the recycling station procured by the municipal. They can’t sell it to a company doing recycling of fats, for example” (Interview: Industry representative). Insufficient user-producer interaction was particularly identified as an important challenge in the case of BioInnovation (DaN1). Users lack knowledge about bio-based products, which disincentivizes upstream firms from including firms downstream in the value chain in innovation projects. Consequently, the market relevance of new bio-based products is often uncertain (DaI2). This issue has been addressed by establishing user–producer interaction as a highly prioritized topic of the programme. Firstly, this implies that collaboration between actors from different parts of the value chain, and between different industries’ value chains, is given particular attention (DaN1). Such collaborations may allow, for example, pulp and paper firms to learn from, for example, textile firms, which have greater knowledge about consumer patterns in emerging markets for bio-based products. Secondly, BioInnovation has stipulated that inclusion of users as work package leaders is an important evaluation criterion for potential new projects (DaA1). Consequently, innovation projects that do not involve users are unlikely to obtain funding through BioInnovation. According to interviewees, this approach has been effective for product areas where lead users with specified demands are available, yet it was also highlighted that this is not always the case. Thus, an unfortunate consequence of this approach is that more radical innovation projects where no lead users exist will be less likely to receive funding through BioInnovation (DaA1). Challenges related to the institutional level have received only

“There are quite a few actors who are slower than others. It’s always a longer process with big companies. They took part by writing letters of intent, but now they should become active members and then it is slower and harder” (Interview: Programme management representative). To give a concrete example, while the first round of activities within BioInnovation was largely about financing research designed and implemented by academic actors, with little involvement by industry, the

9

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

second round evolved towards better-defined problem-oriented projects clearly targeting solutions and with broader involvement from industry (PlN1). Part of the explanation to this development was the coordination efforts carried out in the project consortium. While industry representatives were indeed part of the programme, the programme configuration did not favour their direct involvement, due the difference in size and financial resources and the divergent time horizons of industry and academia. One way of targeting the development of governance learning capabilities was the implementation of tools for dialogues between the programme members and the programme management in Re:Source. Acknowledging the dominance of academic actors in the programme management, they sent out surveys to all non-academic members of the programme, in order to get a sense of the interests and perceived challenges in a broader sense.

transition and innovation systems. What follows is a discussion and summary of the key issues from the empirical case study, the resulting policy and theoretical implications, as well as suggestions for further research. 5.1. Key issues and lessons learned from the case study As regards directionality, we find that conflicting interests are not directly dealt with. The causes for such conflicting interests are often institutional, path-dependent and compounded by an unequal distribution of power. However, institutional entrepreneurship and change are not actively addressed in the policy design of the SIPs. Furthermore, shared visions are broad and general with a lack of concrete and actionable objectives. Being that broad, the visions provide limited guidance for directionality. This may relate to the inability to resolve conflicting interests or objectives in the first place. It remains an open question to what extent actionable objectives should be compromised for a breadth of involvement of heterogeneous stakeholders with different and sometimes conflicting interests. This in turn relates to the impact claim that the policy programme seeks to assert. Here, it is important to note that Vinnova, for most of its programmes, has adopted the “standards of evidence” guidelines of Nesta, a UK innovation agency, to acknowledge that (public) investments in innovation require a variegated but ex ante specified approach to demonstrating and monitoring impact. Conversely, the investigated SIPs set explicit measures for opening up closed collaboration networks and busting policy silos, which is a key challenge for experimentation. A concrete measure in this regard is to form expert teams that actively engage in project team formation as boundary spanners by suggesting additional partners to be included or by merging of projects. Concerning demand articulation, both SIPs work on the development of innovation procurement capabilities in public bodies, which was considered an important challenge on the actor side. In that regard, we identified very targeted and direct approaches (for example funding projects with the main focus on building innovation procurement capabilities in public bodies) as well as more indirect approaches (for instance, the establishment of a platform for interaction between different actors). The first approach may lead to a higher chance of fulfilling the core aim. The overall design of the SIPs also reflects an explicit aim for policy learning and coordination, in the sense that the previously dominant territorial approach with sectoral specialization has shifted to a platform policy approach integrating and coordinating across territorial as well as sectoral domains. This also related to the above-mentioned outcome of silo-busting, which can be considered to be a key element of governance experimentation (Bulkeley et al., 2016). However, a lack of ability to transcend the community boundaries of academia and industry remains. This is primarily due to “institutional mismatch” such as different time horizons and different expectations of results among actors in those different communities (and relates to one of the first points mentioned above). In addition, we identified some issues that concern interdependencies between the transformation challenges. In some cases, addressing one transformational challenge may lead to negative effects for another. For instance, clear directionality poses limits to experimentation activities as it narrows down the scope for open-endedness and serendipity. Also, a focus on including end users in projects hampers more radical experiments where lead users do not yet exist. This can be conceived of as a lead-user dilemma. Involving users may dampen disruptiveness and radicality but generate greater legitimacy, while not involving them allows for more blue sky innovation at the

”We wanted to take the pulse of the stakeholders. We agreed in the programme management, which actually consists of mainly academia and institutes, what is the most interesting. We know that academia and institutes are interested of today, but now we wanted to have other stakeholders’ perspectives” (Interview: Programme management representative) Challenges related to exercising leadership across domains (PlA1) and incentivizing diverging policy trajectories (PlI2) were experienced in both programmes. For example, in Re:Source, activities targeted the development and provision of material to decision makers, informing them about the current state of the field and ongoing developments. These explicitly target the coordination of actors, to avoid contradictions and conflicts to be reflected in policies and incentives (PlA1, PlA2). The idea is to provide a unified body of knowledge on which policymakers, both in the industry and in the public sector, but with a tendency to emphasize the latter, can base their decisions. ”We have one activity which will only target the identification and shaping of decision making processes in public authorities, so that they do not make contradictory decisions. To avoid that one authority decide something based on how they think it should be, but then […] some other authority has already taken steps towards another contradictory decision. We want to help authorities to coordinate their processes” (Interview: programme representative). In BioInnovation, a so-called “innovation race” was launched: a 60hour-long workshop aimed at identifying ideas for promoting the bioeconomy in Sweden and – importantly – communicating these to decision makers such as the minister for enterprise, who participated. The ambition for this was to generate learning through awareness of the potential and relevance of alternative applications of biomaterials and biotechnologies, and to gain legitimacy. The underlying assumption is that such learning will trigger institutional change towards breaking with existing policy rationales and trajectories (PlI1, PlI2). 5. Conclusion The starting point of this paper was the recent shift in innovation policy, which increasingly prioritizes system-wide changes. This shift has raised fundamentally different challenges for policymakers compared to traditional innovation policy approaches. There has been a gap in adequately operationalizing these new challenges to guide the design, implementation, and evaluation of innovation policies targeting system-wide transitions. This paper contributes to filling this gap by proposing an analytical framework that identifies concrete and tangible challenges that are to be expected given the theoretical arguments and empirical evidence found in the literature on socio-technical system

10

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

risk of remaining a “hopeful monstrosity”. Conversely, in other cases, addressing one specific transformational challenge may also have positive side effects on another, which highlights the need for coordinating activities targeting different types of transformation challenges

are dimensions that cut across the various transformation failures. Our framework highlights that literature on system innovation policy needs to engage with both actor-, network- and institution-focused theories. Otherwise, the result is going to be an overly abstract, non-actionable policy agenda that gives an incomplete or naïve picture of the challenges at the operational level in need of attention and the possibilities for addressing them.

5.2. Policy implications A key policy implication from the empirical findings is that more emphasis needs to be placed on institutional change and institutional entrepreneurship. Concretely, this can be done by acknowledging conflicting interests and identifying the underlying institutional incentives structures. This requires a high amount of inter-organizational trust and dialogue (Marques and Morgan, 2018). Dealing with institutional change will often go beyond the competences of programme managers and participants. This implies that incentives for initiatives that tackle institutional change need to be accompanied with support from Vinnova leadership in coordination with other related state agencies and regulatory bodies (e.g. the Swedish Energy Agency or the agency responsible for public procurement legislation). The investigated SIPs have shown rather limited involvement of business and industry actors. Furthermore, the SIPs have hardly engaged with issues such as social acceptance for emerging technologies, consumer attitudes and social practices. Hence, even though the SIPs have to some extent addressed policy silos, the desired broad engagement and involvement has only partially been realized. Having said this, breadth of involvement of stakeholder groups (with potentially conflicting interest) may compromise actionability. Programmes (and evaluations) should be reflective in terms of how the involvement of stakeholders contributes to solving grand challenges, what potential barriers for ambitious joint projects may be, and how these barriers can be overcome. Considering the complex nature of grand challenges, broad engagement and involvement should follow the model of an embedded state. Evans (1995) argued that rather than debating “how much” states intervene in society, it would be preferable to discuss different ways in which they intervene and the implications of these modes of engagement for society at large. The embedded state stands in contrast to the concept of the neoliberal state, which stresses an arm’s length relationship between the state and other socio-economic actors. Following the example of current innovation policies at the EU level (Morgan, 2016) an embedded state is capable of creating and animating networks, helping to decide on strategic investments and finding a balance between centralized decision-making and decentralized action. Moreover, when designing and evaluating pro-active innovation programmes that aim at addressing system-wide challenges, it is important to go beyond general statements of purpose and the broad involvement of stakeholders. One way to do this is to concretely reflect about how the transformation challenges are addressed at the level of actors, networks and institutions at the operational level of supported innovation projects. Explicit treatment of how a proposed programme and corresponding projects set in motion processes of system-wide change will help to evaluate potential impacts.

5.4. Future research There is still a dearth of knowledge about the link from theory to policy practice. Theory suggests that socio-technical systems encompass a large variety of actors and institutional domains that produce and reproduce dominant patterns of production and consumptions. Following this insight, transition-focused policy instruments are thought to require the involvement of broad actor groups. However, in this empirical case, this has led to rather watered-down, catch-all agendas and difficulties in identifying and implementing ambitious joint projects. To what extent the involvement of broad actor groups in strategic programmes leads to better results than, for instance, a variety of potentially competing experiments of more narrowly defined stakeholder groups still has to be confirmed empirically. In collaboration with a donor, an experimental policy approach could be envisaged that allows comparing such fundamentally different policy approaches. Related to this, it may be asked whether and how the need for broad involvement differs at different stages of the policy design and implementation process, and how an active involvement of different stakeholder groups can be secured and sustained. More generally this calls for a greater engagement of innovation studies scholarship with political science and policy studies. Theory focuses a lot on institutional entrepreneurship, strategic agency and deliberate negotiations, and institutional change. However, we find very little evidence of this in the context of the two strategic innovation programmes. The Swedish example seems to indicate that although the programme aims to stimulate/promote new constellations of actors collaborating in new ways, we find very little support that such policy-based initiatives actually lead to new trajectories or that they stimulate entrepreneurial experimentation among incumbents. Our case study thus raises the questions: To what extent and how can instruments such as the strategic innovation programmes stimulate actors to enact and demonstrate institutional entrepreneurship? Or, should policy focus on mediation and brokering in networks, and how can this mediation – if at all – influence the development and direction of a shared vision and institutional change on a more aggregate level? This also points to an important critical limitation of our study, which concerns the time-period that we have considered. Institutional analysis would require a more longitudinal engagement with these questions that are respectful of the temporal aspects of institutional change. Institutions do not change overnight (or in a couple of years). Acknowledgements This work was funded under the Vinnova project “Policy for System Transition: The Case of Strategic Innovation Programmes”. Supporting funding through the “Swedish Transformative Innovation Policy Platform” (Vinnova), the Marianne and Marcus Wallenberg Foundation, and from the City of Melbourne is also gratefully acknowledged. We are thankful for insightful comments received at several workshops with Vinnova and the OECD as well as at conferences in Brighton (SPRU 50th Anniversary Conference), Lund (2016 EU-SPRI Conference) and Montreal (16th Conference of the International Schumpeter Society, 2016). Previous drafts of this paper have improved substantially because of the thoughtful criticism and suggestions by guest editors Johan Schot and Ed Steinmueller as well as two anonymous referees.

5.3. Theoretical implications Whereas literature on system innovation and transformation failures typically establishes a contrast to system failures, we suggest that transformation failures can be looked at through a systemic lens. Discussing and relating transformation failures to the basic features of innovation systems (actors’ interests and capabilities, networks and institutions) leads to more analytical depth and allows operationalization of challenges for innovation policies that address system-wide change. Furthermore, the systemic perspective provides a suitable conceptual framework to assess the interdependencies between different transformation failures, because actors, networks and institutions 11

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

Appendix A. Interview guide

Appendix B. Code list Intersecting nodes Actors – A Networks – N Institutions – I Parent nodes

Child nodes

Directionality

Di

Experimentation

Ex

12

Promote institutional entrepreneurs Resolve conflicting interests Develop governance capabilities Coordinate directionality Develop shared vision Set actionable objectives

Di Di Di Di Di Di

A1 A2 A3 N1 I1 I2

Stimulate entrepreneurship Support new capabilities

Ex A 1 Ex A 2

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

Demand articulation

Da

Policy learning and coordination

Pl

Promote experimentation non-firm Encourage collaboration Collaboration with new actors Support test and demonstration Expose experimentation to selection Promote risk-taking

Ex Ex Ex Ex Ex Ex

A3 N1 N2 I1 I2 I3

Identify lead users Innovation procurement Stimulate interaction with lead users Balance supply- & demand side policy Promote social acceptance

Da Da Da Da Da

A1 A2 N1 I1 I2

Exercise policy leadership Overcome conflicting interests Develop policy learning Widen policy networks Challenge hierarchies Break with policy rationales Incentivize diverging policy

Pl Pl Pl Pl Pl Pl Pl

A1 A2 A3 N1 N2 I1 I2

mission-oriented innovation policy. Res. Policy 41, 1757–1769. Efrat, K., 2014. The direct and indirect impact of culture on innovation. Technovation 34, 12–20. Eisenhardt, K.M., 1989. Agency theory: an assessment and review. Acad. Manag. Rev. 14, 57–74. Evans, P., 1995. Embedded Autonomy: States and Industrial Transformation. Princeton University Press, Princeton, NJ. Feldman, M.P., Lanahan, L., Miller, J.M., 2011. Inadvertent infrastructure and regional entrepreneurship policy. In: Fritsch, M. (Ed.), Handbook of Research on Entrepreneurship and Regional Development: National and Regional Perspectives. Edward Elgar, Cheltenham, pp. 216–251. Flyvbjerg, B., 2006. Five misunderstandings about case-study research. Qual. Inq. 12, 219–245. Frishammar, J., Söderholm, P., Bäckström, K., Hellsmark, H., Ylinenpää, H., 2015. The role of pilot and demonstration plants in technological development: synthesis and directions for future research. Technol. Anal. Strateg. Manag. 27, 1–18. Garud, R., Karnøe, P., 2003. Bricolage versus breakthrough: distributed and embedded agency in technology entrepreneurship. Res. Policy 32, 277–300. Geels, F.W., 2002. Technological transitions as evolutionary reconfiguration processes: a multi-level perspective and a case-study. Res. Policy 31, 1257–1274. Geels, F.W., 2005. Technological Transitions and System Innovations. Edward Elgar Publishing, Cheltenham. Geels, F.W., Raven, R., 2006. Non-linearity and expectations in niche-development trajectories: ups and downs in Dutch biogas development (1973–2003). Technol. Anal. Strateg. Manag. 18, 375–392. Gertler, M.S., 1995. Being there: proximity, organization, and culture in the development and adoption of advanced manufacturing technologies. Econ. Geogr. 71, 1–26. Glaser, B., 2017. Discovery of Grounded Theory: Strategies for Qualitative Research. Routledge. Glückler, J., Bathelt, H., 2017. Institutional Context and Innovation. Edward Elgar, Cheltenham. Government Offices of Sweden, 2016. Innovation Partnership Programmes: Mobilising New Ways to Meet Societal Challenges. Government Offices of Sweden, Stockholm. Grabher, G., 1993. The weakness of strong ties; The lock-in of regional development in The Ruhr area. In: Grabher, G. (Ed.), The Embedded Firm: On the Socioeconomics of Industrial Networks. Routledge, London/New York, pp. 255–277. Guerzoni, M., Raiteri, E., 2015. Demand-side vs. Supply-side technology policies: hidden treatment and new empirical evidence on the policy mix. Res. Policy 44, 726–747. Hansen, T., Coenen, L., 2017. Unpacking resource mobilisation by incumbents for biorefineries: the role of micro-level factors for technological innovation system weaknesses. Technol. Anal. Strateg. Manag. 29, 500–513. Hargadon, A.B., Douglas, Y., 2001. When innovations meet institutions: edison and the design of the electric light. Adm. Sci. Q. 46, 476–501. Hassink, R., 2010. Locked in decline? On the role of regional lock-ins in old industrial areas. In: Boschma, R., Martin, R. (Eds.), The Handbook of Evolutionary Economic Geography. Edward Elgar, Cheltenham, pp. 450–468. Hill, C.W.L., Rothaermel, F.T., 2003. The performance of incumbent firms in the face of radical technological innovation. Acad. Manag. Rev. 28, 257–274. Hockerts, K., Wüstenhagen, R., 2010. Greening Goliaths versus emerging Davids: theorizing about the role of incumbents and new entrants in sustainable entrepreneurship. J. Bus. Ventur. 25, 481–492. Hodson, M., Marvin, S., 2010. Can cities shape socio-technical transitions and how would we know if they were? Res. Policy 39, 477–485. Kamp, L.M., 2002. Learning in Wind Turbine Development: A Comparison between the Netherlands and Denmark. Utrecht University, Utrecht.

References Asheim, B., Boschma, R., Cooke, P., 2011. Constructing regional advantage: platform policies based on related variety and differentiated knowledge bases. Reg. Stud. 45, 893–904. Battilana, J., 2011. The enabling role of social position in diverging from the institutional status quo: evidence from the UK National Health Service. Organ. Sci. 22, 817–834. Battilana, J., Leca, B., Boxenbaum, E., 2009. How actors change institutions: towards a theory of institutional entrepreneurship. Acad. Manag. Ann. 3, 65–107. Bennett, C.J., Howlett, M., 1992. The lessons of learning: reconciling theories of policy learning and policy change. Policy Sci. 25, 275–294. Bergek, A., Jacobsson, S., Sandén, B.A., 2008. “Legitimation” and “development of positive externalities”: two key processes in the formation phase of technological innovation systems. Technol. Anal. Strateg. Manag. 20, 575–592. Borrás, S., 2011. Policy learning and organizational capacities in innovation policies. Sci. Public Policy 38, 725–734. Bulkeley, H., Broto, V.C., 2013. Government by experiment? Global cities and the governing of climate change. Trans. Inst. Br. Geogr. 38, 361–375. Bulkeley, H., Broto, V.C., Hodson, M., Marvin, S., 2011. Cities and Low Carbon Transitions. Routledge, Abingdon, UK. Bulkeley, H., Coenen, L., Frantzeskaki, N., Hartmann, C., Kronsell, A., Mai, L., Marvin, S., McCormick, K., van Steenbergen, F., Palgan, Y.V., 2016. Urban living labs: Governing urban sustainability transitions. Curr. Opin. Environ. Sustain. 22, 13–17. Burt, R.S., 2005. Brokerage and Closure. Oxford University Press, Oxford. Cagnin, C., Amanatidou, E., Keenan, M., 2012. Orienting European innovation systems towards grand challenges and the roles that FTA can play. Sci. Public Policy 39, 140–152. Chandy, R.K., Tellis, G.J., 2000. The incumbent’s curse? Incumbency, size, and radical product innovation. J. Mark. 64, 1–17. Chang, Y.-C., Chang, H.-T., Chi, H.-R., Chen, M.-H., Deng, L.-L., 2012. How do established firms improve radical innovation performance? The organizational capabilities view. Technovation 32, 441–451. Coenen, L., Hansen, T., Rekers, J.V., 2015. Innovation policy for grand challenges. An economic geography perspective. Geogr. Compass 9, 483–496. Cooke, P., Eriksson, A., 2011. White Spaces Innovation in Sweden: Innovation Policy for Exploring the Adjacent Possible, Vinnova Report. Vinnova, Stockholm. Cooke, P., Morgan, K., 1994. The regional innovation system in Baden-Wurttemberg. Int. J. Technol. Manag. 9, 394–429. Dawley, S., 2014. Creating New Paths? Offshore wind, policy activism, and peripheral region development. Econ. Geogr. 90, 91–112. Dewald, U., Truffer, B., 2012. The local sources of market formation: explaining regional growth differentials in German photovoltaic markets. Eur. Plan. Stud. 20, 397–420. Dicken, P., 2011. Global shift. Mapping the Changing Contours of the World Economy, 6th ed. Guilford Press, New York. DiMaggio, P.J., 1988. Interest and agency in institutional theory: institutional patterns and organizations. Cult. Environ. 1, 3–22. Edler, J., Georghiou, L., 2007. Public procurement and innovation: resurrecting the demand side. Res. Policy 36, 949–963. Edquist, C., 2005. Systems of innovation: perspectives and challenges. In: Fagenberg, J., Mowery, D.C. (Eds.), Oxford Handbook of Innovation. Oxford University Press, pp. 181–208. Edquist, C., Johnson, B., 1997. Institutions and organizations in systems of innovation. In: Edquist, C. (Ed.), Systems of Innovation. Frances Pinter, London. Edquist, C., Zabala-Iturriagagoitia, J.M., 2012. Public procurement for innovation as

13

Research Policy xxx (xxxx) xxx–xxx

M. Grillitsch et al.

Rotmans, J., Kemp, R., Van Asselt, M., 2001. More evolution than revolution: transition management in public policy. Foresight 3, 15–31. Roy, R., Sarkar, M.B., 2016. Knowledge, firm boundaries, and innovation: mitigating the incumbent’s curse during radical technological change. Strateg. Manag. J. 37, 835–854. Schot, J., Geels, F.W., 2008. Strategic niche management and sustainable innovation journeys: theory, findings, research agenda, and policy. Technol. Anal. Strateg. Manag. 20, 537–554. Schot, J., Steinmueller, W.E., 2016. Framing Innovation Policy for Transformative Change: Innovation Policy 3.0. Science Policy Research Unit (SPRU), University of Sussex, Brighton. Schot, J., Steinmueller, W.E., 2018. Three frames for innovation policy: R&D, systems of innovation and transformative change. Research Policy 47, 1554–1567. Schot, J., Kanger, L., Verbong, G., 2016. The roles of users in shaping transitions to new energy systems. Nat. Energy 1, 16054. Scott, W.R., 1995. Institutions and Organizations. Foundations for Organizational Science. Sage, London. Sengers, F., Wieczorek, A.J., Raven, R., 2016. Experimenting for sustainability transitions: a systematic literature review. Technol. Forecast. Soc. Change. Setterfield, M., 1993. A model of institutional hysteresis. J. Econ. Issues 27 (3), 755–774. Shane, S., 1993. Cultural influences on national rates of innovation. J. Bus. Ventur. 8, 59–73. Shove, E., Walker, G., Brown, S., 2014. Transnational transitions: the diffusion and integration of mechanical cooling. Urban Stud. 51, 1506–1519. Smith, A., Stirling, A., Berkhout, F., 2005. The governance of sustainable socio-technical transitions. Res. Policy 34, 1491–1510. Sotarauta, M., 2009. Power and influence tactics in the promotion of regional development: an empirical analysis of the work of Finnish regional development officers. Geoforum 40, 895–905. Sotarauta, M., Pulkkinen, R., 2011. Institutional entrepreneurship for knowledge regions: in search of a fresh set of questions for regional innovation studies. Environ. Plann. C Gov. Policy 29, 96–112. Stephens, J.C., Hernandez, M.E., Román, M., Graham, A.C., Scholz, R.W., 2008. Higher education as a change agent for sustainability in different cultures and contexts. Int. J. Sustain. High. Educ. 9, 317–338. Stern, S., 2006. Economic experiments: the role of entrepreneurship in economic prosperity. Melbourne Rev. 2, 53–56. Stilgoe, J., Owen, R., Macnaghten, P., 2013. Developing a framework for responsible innovation. Res. Policy 42, 1568–1580. Tanner, A.N., 2014. Regional branching reconsidered: emergence of the fuel cell industry in European regions. Econ. Geogr. 90, 403–427. Truffer, B., 2003. User-led innovation processes: the development of professional car sharing by environmentally concerned citizens. Innov. Eur. J. Soc. Sci. Res. 16, 139–154. Turró, A., Urbano, D., Peris-Ortiz, M., 2014. Culture and innovation: the moderating effect of cultural values on corporate entrepreneurship. Technol. Forecast. Soc. Change 88, 360–369. van der Laak, W.W.M., Raven, R.P.J.M., Verbong, G.P.J., 2007. Strategic niche management for biofuels: analysing past experiments for developing new biofuel policies. Energy Policy 35, 3213–3225. Vinnova, 2017. Årsredovisning 2016. Vinnova, Stockholm. Von Hippel, E., 1986. Lead users: a source of novel product concepts. Manage. Sci. 32, 791–805. Von Hippel, E., 1988. The Sources of Innovation. Oxford University Press, New York. Weber, K.M., Rohracher, H., 2012. Legitimizing research, technology and innovation policies for transformative change: combining insights from innovation systems and multi-level perspective in a comprehensive ‘failures’ framework. Res. Policy 41, 1037–1047. Wirth, S., Markard, J., Truffer, B., Rohracher, H., 2013. Informal institutions matter: professional culture and the development of biogas technology. Environ. Innov. Soc. Transit. 8, 20–41. Yin, R.K., 2013. Case Study Research: Design and Methods. Sage, Thousand Oaks, CA.

Kemp, R., Schot, J., Hoogma, R., 1998. Regime shifts to sustainability through processes of niche formation: the approach of strategic niche management. Technol. Anal. Strateg. Manag. 10, 175–198. Kivimaa, P., Kern, F., 2016. Creative destruction or mere niche support? Innovation policy mixes for sustainability transitions. Res. Policy 45, 205–217. Klitkou, A., Coenen, L., Andersen, P.D., Fevolden, A., Hansen, T., 2013. Role of demonstration projects in innovation: transition to sustainable energy and transport. Presented at 4th International Conference on Sustainability Transitions. Kuhlmann, S., Rip, A., 2014. The Challenge of Addressing Grand Challenges: a Think Piece on How Innovation Can Be Driven Towards the “grand Challenges” As Defined Under the Prospective European Union Framework Programme Horizon 2020. European Research and Innovation Area Board (ERIAB). Laranja, M., Uyarra, E., Flanagan, K., 2008. Policies for science, technology and innovation: translating rationales into regional policies in a multi-level setting. Res. Policy 37, 823–835. Leonard-Barton, D., 1992. Core capabilities and core rigidities: a paradox in managing new product development. Strateg. Manag. J. 13, 111–125. Loorbach, D., 2010. Transition management for sustainable development: a prescriptive, complexity-based governance framework. Governance 23, 161–183. Lundvall, B.-Å., 1988. Innovation as an interactive process: from user-producer interaction to the national system of innovation. In: Dosi, G., Freeman, C., Silverberg, G., Soete, L.L. (Eds.), Technical Change and Economic Theory. Pinter, London, pp. 349–369. Lundvall, B.-Å., 2010. National Systems of Innovation: Toward a Theory of Innovation and Interactive Learning. Anthem Press, London. Lyall, C., Bruce, A., Marsden, W., Meagher, L., 2013. The role of funding agencies in creating interdisciplinary knowledge. Sci. Public Policy 40, 62–71. Markard, J., Truffer, B., 2008. Technological innovation systems and the multi-level perspective: towards an integrated framework. Res. Policy 37, 596–615. Marques, P., Morgan, K., 2018. The heroic assumptions of smart specialisation: a sympathetic critique of regional innovation policy. In: Isaksen, A., Martin, R., Trippl, M. (Eds.), New Avenues for Regional Innovation Systems: Theoretical Advances, Empirical Cases and Policy Lessons. Springer, New York, pp. 275–293. Martin, R., Sunley, P., 2006. Path dependence and regional economic evolution. J. Econ. Geogr. 6, 395–437. May, T., 2011. Social Research. McGraw-Hill Education, Maidenhead, UK. Mazzucato, M., 2015. The Entrepreneurial State: Debunking Public vs. Private Sector Myths. Anthem Press, London. Morgan, K., 2016. Nurturing novelty: regional innovation policy in the age of smart specialisation. Environ. Plann. C Gov. Policy 35, 569–583. Näyhä, A., Pesonen, H.-L., 2014. Strategic change in the forest industry towards the biorefining business. Technol. Forecast. Soc. Change 81, 259–271. Neffke, F., Henning, M., 2013. Skill relatedness and firm diversification. Strateg. Manag. J. 34, 297–316. Nill, J., Kemp, R., 2009. Evolutionary approaches for sustainable innovation policies: From niche to paradigm? Res. Policy 38, 668–680. OECD, 2013. OECD Reviews of Innovation Policy: Sweden. OECD, Paris. OECD, 2015. System Innovation: Synthesis Report. OECD, Paris. Powell, W., Grodal, S., 2005. Networks of innovators. In: Fagerberg, J., Mowery, D., Nelson, R.R. (Eds.), The Oxford Handbook of Innovation. Oxford University Press, Oxford. Raskin, P., Banuri, T., Gallopín, G., Gutman, P., Hammond, A., Kates, R., Swart, R., 2002. Great transition, the promise and lure of the times ahead. SEI PoleStar Series Report no. 10. Stockholm Environment Institute, Boston. Raven, R., Bosch, Svd., Weterings, R., 2010. Transitions and strategic niche management: towards a competence kit for practitioners. Int. J. Technol. Manag. 51, 57–74. Rekers, J.V., Hansen, T., 2015. Interdisciplinary research and geography: overcoming barriers through proximity. Sci. Public Policy 42, 242–254. Rock, M.T., Angel, D.P., Pao Li, L., 2006. Impact of firm-based environmental standards on subsidiaries and their suppliers: evidence from motorola-penang. J. Ind. Ecol. 10, 257–278. Rosenbloom, R.S., 2000. Leadership, capabilities, and technological change: the transformation of NCR in the electronic era. Strateg. Manag. J. 21, 1083–1103.

14