Forthcoming in Technological Forecasting and Social Change

Innovation System Foresight

Allan Dahl Andersena* and Per Dannemand Andersenb

a) University of Oslo, Centre for Technology, Innovation and Culture, Boks 1072 Blindern, 0316 Oslo, Norway

b) Technical University of Denmark, Department of Management Engineering, Technology & Innovation Management Division. Produktionstorvet 424, DK-2800 Kgs. Lyngby, Denmark

* Corresponding author.

AbstractThe practice and concept of foresight has developed over several decades. However, the academic literature that addresses foresight is mainly descriptive, and it is generally acknowledged that there is gap between practice and theory in foresight. This article contributes to building the theoretical underpinnings of foresight. The article explores the co-evolution of the academic field of innovation studies and the practice and concept of foresight. The article illustrates that foresight is importing the dominant understanding of innovation from innovation studies with a time lag. Currently, foresight is in a catching-up process vis-à-vis innovation studies by gradually incorporating the implications of a systemic understanding of innovation. The latter is reflected in the current research gaps in foresight. This paper concludes that the concept of innovation system foresight (ISF) constitutes an improved integration of the contemporary understanding of innovation into foresight. Furthermore, the article explores four preliminary implications of ISF on the conceptual design of foresight, including the goal of foresight, system definition and boundary setting, participation of actors, and finally, methods for mapping the present. These four implications also constitute avenues for further research.

KeywordsForesight; innovation systems; theoretical underpinnings; systemic approach; conceptual review

1 Introduction The practice of foresight has developed over several decades, and foresight and similar approaches, such as technology forecasting, futures studies, la prospective and future-oriented technology analysis, have been used in both the private and public sectors since the mid-1940s. Foresight is also about to emerge as an academic discipline, with attributes such as dedicated international peer-reviewed journals, academic conferences and teaching at the college and university levels. However, the academic literature that addressed foresight is largely descriptive [1], and it is generally acknowledged that there is gap between practice and theory in foresight [2], [3]. In recent years, this has led to a discussion of the possible ‘theoretical underpinnings’ of foresight [4], [5]. Such theoretical underpinnings are a requirement for further theory testing, theory building, and the development of a coherent framework to guide foresight practice. A murky theoretical underpinning makes it difficult for foresight to be carried out systematically—i.e., coherently—with a theoretical framework to support decisions made before, during and after the foresight process. For example, there does

1

not seem to be any theoretically driven reasoning behind the delimitation of areas of interest, selection of participants, and criteria/design for understanding and analysing the present.

Moreover, both as a field of practice and as an academic field, foresight seems to be experiencing a ‘systems turn’, as it is gradually, albeit not always explicitly, implementing the systemic, contextual and evolutionary understanding of innovation that is dominant in the academic field of innovation studies [6]. The latter is reflected by two current research gaps in foresight. First, it has been increasingly recognised that foresight is highly context dependent [7], [8]. Therefore, foresight must be able to systematically and coherently include context to conclude anything sensible about innovation. Despite its importance, such work is currently limited [9]. Additionally, foresight exercises often do not take sufficient notice of the demand for knowledge, existing competences, and reality and wishes of firms and policy makers [10]. The lack of a (demonstrable) impact of foresight has led to an increased focus on the demand-side of the innovation process within foresight [11]. The argument is that more seriously including demand will increase the impact [12].

We suggest that the innovation system (IS) framework can address these two research gaps. Our key proposition is that explicitly taking the IS framework as the theoretical underpinning of foresight can simultaneously accommodate the need for underpinnings and associated problems and the research challenges related to context and demand for knowledge. This specific type of foresight we denote as innovation system foresight (ISF). Our proposition complements and contributes to several recent developments in foresight research. First, the nexus between foresight and ISs have been tentatively explored by others, but they have mainly focused on how foresight can contribute to IS analysis [13], [14], [15], [16]; there seems to be very limited—or no—studies on how the IS framework can contribute to foresight.

Second, others have explored the practical applications of an integration of IS analysis and foresight at a sector level [17] [16], but these studies were descriptive, application oriented, and not occupied with the conceptual integration between IS research and foresight.

Third, a systems approach to foresight is not unique. Recently, an explicit systems approach to foresight seems to have emerged [18], [19], [20]; however, the work of Saritas [18], although profound, is mainly about complex systems in general and is not focused on innovation dynamics. Our work adds an innovation focus to this line of work. Despite obvious overlaps, according to Smits, Kuhlmann, and Shapira [10], there is little communication between IS research and foresight, and the linkages between them remain embryonic.

Our contribution is an attempt to address these three recent developments in foresight research. The aim is to contribute to theory-building in foresight in the direction of developing a systemic approach that has innovation dynamics at its core. Consequently, this paper is conceptually explorative.

In section 2, we qualify our choice of systems approach by reviewing key parts of the literature within innovation studies and foresight. This review indicates that foresight has coevolved with innovation studies. The review explicates the conceptual linkages between the two fields, allowing us to systematise and explore them further. The latter reveals that foresight is ‘misaligned’ with innovation studies in its conceptualisation of innovation, preparing the ground for ISF. In section 3, we present and elaborate on the content of ISF. In section 4, we demonstrate how ISF addresses the research frontiers mentioned above, and we

2

explore the potential implications for foresight (mainly in the pre-foresight phase) of accepting the IS framework as an explicit theoretical basis. Section 5 contains concluding remarks and future perspectives.

2 The co-evolution of foresight and innovation studiesDescribing the historical development of the understanding and use of foresight is complicated by the fact that foresight is often considered to be an area of practice based on three more established traditions: technology forecasting, futures studies and technology assessment. Technology forecasting emerged in the aftermath of World War II, when the American military needed a systematic method for making informed judgments of rapid technological developments [21]. Foresight is also rooted in a European tradition of futures studies that was established in the 1960s and 1970s [22] [23]. Futures studies is sometimes characterised as an art involving creative and imaginative thinking and acting [24]. Early futures studies had a pessimistic and critical view of the future and technology. Compared to early technology forecasting, futures studies were more focused on stimulating public debate, whereas forecasting was an instrument for decision making [22]. According to Ian Miles, the technology criticism by futures studies seems to have formed the basis for a third tradition: technology assessment [22] [25], which involves a systematic assessment of the consequences for society of introducing and using new technology. Technology assessment has especially contributed to foresight with participatory methods in the form of the broad inclusion of citizens in the assessment process. In the early 1980s, Irvine and Martin introduced the term ‘foresight’ as a strategic, forward-looking technology analysis to be used as a public policy tool for priority setting in science and technology [26]. It was defined in opposition to ‘hindsight’—which was understood as the analysis of the historical origins of technological innovations. Since the 1980s, foresight has established itself as a field of practice in both public policy making and corporate strategic planning and, more recently, as a scientific discipline [27].

In the following, we will not further distinguish between forecasting, futures studies and foresight. We will consider these traditions to be elements in the same continuum under the label of foresight. Moreover, we will focus on foresight in the context of public policy for science, technology and innovation (STI). Hence, the essential rationale and motivation for (public policy) foresight is ultimately to link STI policy more effectively to social and economic development, with innovation as the main lever [13].

Despite early and distinguished contributions, innovation studies1 did not emerge as a research field until approximately 1960 [28]. Since then, the field has grown and matured, entailing the accumulation of a large body of knowledge about innovation and its impacts [29]. Within the richness of the field, we limit our interest to the evolution of the dominant conceptualisation of innovation, which currently ends with the systemic understanding epitomised in the notion of a system of innovation. The core topic of IS research is to understand the interaction between technological change and economic performance [30].

1 It is not obvious where the boundaries of innovation studies should be located. A useful definition for this paper, which has recently been used by Ben Martin [61], is given by the journal ‘Research Policy’. Here, it is defined as studies “devoted to analysing, understanding and effectively responding to the economic, policy, management, organizational, environmental and other challenges posed by innovation, technology, R&D and science. This includes a number of related activities concerned with the creation of knowledge (through research), the diffusion and acquisition of knowledge (e.g. through organizational learning), and its exploitation in the form of new or improved products, processes or services”.

3

There is a wealth of definitions of IS. We will, however, follow a broad and basic definition given by Lundvall, who defines an IS as the elements and relationships that interact in the production, diffusion and use of new and economically useful knowledge [31]. The IS approach first emerged as a national IS concerned with how nations build knowledge infrastructures for economic development [31], [32], [33]. Since then, the IS framework has developed in different directions that include different levels and dimensions, such as regional IS [34], [35], sectoral IS [36], [37] and technological IS [38], [39]. This reflects that an IS may be delimited spatially, sectorally or technologically. In parallel, foresight has been developed in the context of national, regional, sectoral and technology/science-specific foresight exercises.

2.1 Co-evolution of conceptualisations of innovation, innovation policy and foresight

Changes in the understanding and content of foresight since the 1940s have been motivated by changing rationales for engaging in foresight and broadening the scope of its areas of application. This, in turn, has generated a vast diversity in the methods applied in foresight exercises. Several studies have discussed this development and have identified different generations of foresight [1], [40], [41], [42], [43], [44].

During the same period, the understanding and the “mode” of innovation has also changed, and several studies have identified successive generations of “innovation models” [45], [46].2 The generations reflect changes in the sources, nature, and complexity of the innovation process and also a change in how well we have understood these changes.

The rationale for innovation policy has changed in response to succeeding innovation models. Lundvall and Borrás [47] identify three ideal types of innovation policy in the post-World War II period: science policy, technology policy, and innovation policy.3 Science policy essentially relies on a linear model of innovation. Science is observed as being capable of solving society’s problems; therefore, the most important activity is basic science, and the most relevant actors are universities and research organisations. Technology policy emphasises the links between science and industry. Attention moves towards engineering and from the internal organisation of universities towards how they link to industry. The ideal type of innovation policy is derived from the insight that innovation is a systemic process. Here, innovation policy basically concerns all relevant aspects of society that influence the process of innovation, which makes policy systemic. In the following, we will to bring together the main points of the literature that describe the evolution of the “models of innovation”, the practice of innovation policy and foresight.

2.1.1 1st Generation—The science-push innovation model

Prior to World War II, science was not generally observed as relevant for production and economic wealth, but this changed immediately after the war due to, inter alia, the role of science in winning the war [47]. The economic recovery in the USA and Europe was partly based on the emergence of new industries with close links to science (for example, semiconductors and electronic computing). Given these circumstances, it is hardly surprising that innovation was mainly perceived as a linear process from scientific discovery through 2 The reviews referred to in the following have a clear geographical focus on research in USA and Europe and are thus not globally representative. Still, the latter is a consequence of the fact that innovation studies as a discipline has been concentrated in these regions. The field of innovation studies has only very recently diversified geographically [28], [29]. 3 These ideal types are difficult to precisely pinpoint in time, but they are rooted in different policy rationales that are related to the changing understanding of innovation.

4

technological development in firms and finally, to the market place. The transformation of scientific output to technology and innovation was largely ignored [45]. This led to an innovation policy with a strong focus on basic science (science policy). With this understanding, foresight was merely understood to be the forecasting of technological development and the domain of technological experts with a focus on natural sciences and engineering. Forecasting exercises were concerned with the accuracy of predictions as an essential part of a science and technology transfer (to society) policy [41]. Thus, it is suitable to label this foresight practice as ‘science foresight’.

2.1.2 2nd generation—The demand-pull innovation model

In the mid-1960s, the focus was on growth, productivity and large-scale industry. Technology and product innovation slowed and employment growth rates stagnated, and in this contractionary environment, competition between firms intensified. The latter created a stronger focus on marketing, which enhanced focus on demand-side factors in innovation. The market was observed as a cradle for ideas that could orient R&D efforts, which was consequently given a ‘reactive role’. On this basis, a ‘market pull’ innovation model was proposed [45]. Foresight in this period also increasingly recognised the importance of the demand side in the successful transfer of science-based knowledge to industry. Representatives from industry were now included among the key actors in the foresight process [1], [41]. We label this practice ‘technology foresight’. The innovation policy ideal moved from focusing on science policy towards technology policy.

2.1.3 3rd generation—The linkage model

The 1970s were a period of economic crisis where firms focused more on market consolidation and rationalisation. According to Rothwell [45] and Dodgson et al. [46], this context stimulated a more profound interest in discovering the sources of innovation, giving rise to several original empirical studies.4 These studies basically found that both the science-push and the market-pull models were extreme and atypical descriptions of innovation processes. Instead, innovation is most often a process consisting of distinct but interacting and interdependent stages—e.g., the continuous processes of feedback between design, product development, and marketing—with varying links to research communities. Hence, the focus was on interactions between phases and actors [45], [48] that combine supply-push and demand-pull perspectives.

Foresight in this period was characterised by a ‘broadening’ of the market perspective by the inclusion of a wider set of actors in foresight exercises. This period was inspired by foresight activities in Japan, where the use of bottom-up inputs and process benefits were especially noted [24], [49], [22]. Insights from innovation studies implied that foresight practitioners started to adopt a more complex and non-linear view of innovation. It was recognised that there were insufficient ‘bridging organisations’ in the socio-economic system, and foresight began to be observed as an arena for making the necessary network connections [1]. The latter was an early sign of a systems approach, which implied a broadening of the actors that could be considered relevant to include. Hence, 3rd-generation foresight adds social stakeholders, such as NGOs, consumer groups, and various other interest groupings, with a less techno-optimistic view. It reflects the emergence of futures studies and technology assessment in this period. Socio-economic problem solving was increasingly emphasised as an organising principle rather than the pursuit of scientific opportunities [41]. These changes imply that the policy ideal moved towards a systemic ideal and closer to what could be described as ‘innovation policy’.

4 See Rothwell [45] page 9.

5

2.1.4 4th generation—The integrated model

The early 1980s was a period of economic recovery where firms became increasingly aware of the strategic importance of developing and using new technologies. In this context, the rise of information and communication technologies (ICTs) was central to innovation. Meanwhile, the global outlook of firms extended and the number of international strategic alliances grew. Organisational innovations and the rise of Japanese firms in the 1980s were central to changes in innovation studies. ‘Western’ firms started to look at the Japanese product development system, which was characterised by short product life cycles and low production costs, for inspiration. It was based on the simultaneous integration of firm-external and -internal actors in product development processes [45]. Here, the stages of the innovation process are not observed as sequential but rather as mutually integrated, which implies the intensive communication of information and knowledge between actors. It involves more complex information flows within the firm and with multiple external sources of innovation [46]. There is no obvious counterpart in foresight models to the integrated innovation model, cf. the next section.

2.1.5 5th generation—A systems model

The fifth generation innovation model is characterised by the impact of ICT. Rothwell [45] argues that because ICT has the capacity to speed up parallel and integrated processes of innovation, the technology of technological change was itself changing. The strategic environment for firms became further globalised and increasingly took the shape of a network economy where firms are mutually interdependent [46]. Innovation thus became more of a networking process involving vertical linkages, the use of ICT, and horizontal linkages (joint R&D and strategic alliances). These changes are associated with the emergence of the ‘learning economy’ [50], where knowledge is the most important factor for economic performance and where learning is the most important process. The rising complexity of technology and innovation limits the value of a firm's internal knowledge. This forces firms to collaborate because the innovation-relevant knowledge is distributed across a range of actors and knowledge bases in society [51]. Hence, innovation is best understood as a distributed and systemic process in which firm-external knowledge and learning are crucial factors. The growing importance of firm-external linkages suggests a systems approach to understanding innovation that focuses not only on the performance of individual firms but also on how they are embedded into complex social and economic relationships in their environments. The IS framework was suggested to conceptualise these changes [52].

The broad fault line between a systemic and non-systemic perception of innovation is located between the 2nd and 3rd generations of foresight. Hence, in foresight, there are no obvious conceptual counterpart models to the 4th and 5th generation understandings of innovation. Still, Miles et al. [1] have further identified 4th and 5th generations of foresight that unfolded in the 2000s. In the 4th generation, foresight begins to have a distributed and coordinating role, which implies that more diverse types of organisations are conducting foresight and that more diverse stakeholders are identified as relevant participants. The 5th generation covers a diversity of foresight levels, locations, dimensions, methods, design and rationales. Often, the focus is on moulding structures within STI and addressing (the S&T dimension of) grand societal challenges. Here, there is no observable approximation to an IS understanding of innovation or to an increasing focus on innovation. What is observable is increasing diversity arising, on the one hand, from experimentation and application to new domains and, on the other hand, from the coexistence of diverse rationales seldom related to innovation. Current foresight practice has thus gone well beyond what is conceptualised as technology foresight [53]. In conclusion, with respect to the understanding of innovation, the 3rd, 4th and 5th

6

generations of foresight do not contain radical differences and may, for our purposes, be considered the same. For this reason, we label them jointly with the broader term ‘foresight’; see Table 1.

Although the notion of generations in innovation studies and foresight in both time and space is crude, it allows us to illustrate that the trajectories of the respective fields have co-evolved around the dominant understanding of innovation. It also enables us to see that, more recently, there has been a decoupling of these trajectories.

The decoupling has two complementary explanations. First, due to rising diversity and volume in foresight, the trend in innovation-oriented foresight is difficult to observe because it has diminished in the field relative to other forms of insight. Second, foresight as a field is currently ‘misaligned’ with innovation studies in its conceptualisation of innovation as a systemic and evolutionary phenomenon [3], [16]. Because innovation models and related policy rationales are not equally valuable for understanding innovation, diagnosing problems and prescribing solutions, there is reason to explicitly re-couple the conceptualisations of innovation in innovation studies and foresight.

3 Innovation System Foresight The recent decoupling described above constitutes a motivation for suggesting deeper integration with the IS framework. In addition, the review of foresight research reflects a growing diversity of disciplines, rationales, methodologies and approaches contained within the term 'foresight'. Such diversity may be positive,5 but little of it is related to innovation and still less to the IS framework. Hence, from our perspective, there seems to be a need for conceptual clarification and an explicit formulation and positioning of innovation-oriented foresight.

The term ‘innovation foresight’ has been used previously [16], [54]. Porter describes innovation foresight as being different from science-oriented foresight and technology-oriented foresight because it demands more attention to socio-economic contextual forces interacting with emerging technical capabilities to affect commercial product and services [54]. Porter’s point is that the innovation domain is different from and broader than the domains of science and technology and that users play a more pronounced role in the innovation domain. Still, it is not emphasised that the science, technology and innovation dimensions of an IS are interdependent in a systemic way, reflecting that not all understandings of innovation are systemic. Thus, there is reason to be conceptually more precise.

On this basis, we propose the term ‘innovation system foresight’ (ISF) to denote foresight that explicitly takes the IS framework as its theoretical underpinning and thus operates with a systemic, contextual and evolutionary understanding of innovation. Moreover, ISF is more accurate than ‘innovation foresight’ because it stresses the systemic and process-like character of both foresight and innovation. Foresight is often defined as systematic, but it is important to note that being systematic does not imply a systemic view. We define ISF as a systemic, systematic, participatory, future-intelligence-gathering and medium-to-long-term vision-building process aimed at present-day decisions and mobilising joint actions to improve

5 Obviously, the diversity found in foresight can be a strength for grasping the diversity of the real world, but the balance between diversity and generality is delicate. The IS approach contains significant diversity within its framework that leaves some flexibility for adapting the analysis to the object of interest.

7

innovation system performance with the ultimate goal of improving desirable socio-economic performance.

The key message of ISF is that we must take into account that new knowledge is generated and diffused in a ‘system of innovation’ that has actors, institutions and dynamics that must be analysed and included in any meaningful foresight exercise. Hence, the systemic nature of innovation processes must be central to foresight thinking to generate better strategies and a stronger impact. That foresight practitioners tend to overestimate the power of foresight as a result of underestimating the complexity of innovation [55] is a testament to the latter point. Without a systemic understanding of innovation, foresight is bound to be have limited impact on the grounds of weak conceptual understanding.

Accepting ISF does not necessarily imply that science-oriented and technology-oriented foresights are irrelevant. Instead, they are complementary and valuable elements to ISF insofar as they are designed to pay attention to interactions between science, technology and innovation. For example, priority-setting in science policy is still relevant, but the design and process behind this concept should be derived from ISF. If it does not, science policy can be everything from unproductive to harmful for innovation and for achieving desired impact.The previous literature review and current proposal allow us to combine different (generations of) conceptualisations of innovation, generations of foresight practice and design, and generations of innovation policy. In Table 1, we ascribe to each generation a label and place ISF in its historical and conceptual context. The conceptualisations and models should be understood as complementary and coexisting ideal types that, most likely, do not exist in a pure form. Still, the table illustrates a trajectory of development within foresight that reflects a gradual shift from a linear perception of innovation towards a systemic shift. Retrospectively, our exercise explicates theoretical underpinnings of all foresight generations on the basis of their respective conceptualisations of innovation as phenomenon.

The ISF concept adds value to foresight research by explicitly emphasising the necessity of a systemic understanding of innovation and by defining an innovation-oriented style of foresight within the vast diversity of foresight practices and rationales. ISF further holds the potential for addressing the lack of theoretical underpinnings in foresight and other research gaps in foresight. We will explore these in the next section.

8

Conceptual understanding of innovation in foresight

Type of innovation policy

Description of foresight generations

Structure/design of foresight

Label6

1st The science-push model of innovation

Science policy

Mainly forecasting Mainly expert group driven and elitist

Science foresight

2nd The demand-pull and couplings model

Technology policy

Emphasises the matching of technological opportunities with market and nonmarket (environment and social issues).

Increasingly involving firms and policy makers

Technology foresight

3rd The demand-pull and couplings modelThe integrated model of innovation

Technology policyInnovation policy

Signifies an enhancement, or broadening, of foresight’s market perspective by inclusion of a broader social dimension that involves concerns and inputs from a broad range of social actors.

Increasingly involving socio-economic actors; more inter-disciplinarily

Foresight (diversity)

4th Mix of conceptualisations with relatively diminishing emphasis on innovation

n.a. Foresight becomes distributed and broadened in scope. Intensifies characteristics of the 3rd generation.

Increasing diversity in terms of actors, levels, goals and designs

Foresight (diversity)

5th Mix of all conceptualisations co-existing with an increasing focus on systems and network approaches, but with relatively diminishing emphasis on innovation

n.a. Foresight becomes concerned with the science and technology systems perspective and (or because of) increasing orientation towards solving societal challenges

Increasing diversity in terms of actors, levels, goals and designs

Foresight (diversity)

Prop

osal

IS approach Systemic Innovation policy

Focus on innovation and transforming IS with increased attention to the demand for knowledge, context factors and diversity in innovation dynamics.

System delimitation and nature structures design; must be inclusive.

Innovation system foresight

Table 1: Conceptual linkages between generations of the understanding of innovation, innovation policy and foresight (developed by the authors).7

4 Implications of innovation system foresight

4.1 Addressing research gaps

The gradual shift from a linear perception of innovation towards a systemic framework in foresight is reflected by two related research gaps that can be addressed by ISF. First, it is argued that foresight should increasingly move from being about priority setting towards being more focused on implementing insights and realising structural change [56]. A lack of 6 The label is derived from underlying conceptualisations of innovation rather than the object and purpose of foresight. Hence, e.g., science foresight relies on the linear model of innovation, etc.7 With respect to the understanding of innovation, the 3rd, 4th and 5th generations of foresight are not radically different and may, for our purpose, be considered the same.

9

(demonstrable) impact of foresight has led to an increased focus on the demand-side of the innovation process within foresight [11]. The argument is that including demand more seriously will increase impact [12].

The role of demand for knowledge and users of innovation is theoretically central to the IS framework. As the importance of firm-external knowledge for innovation increases, interactive learning becomes the most important type of learning [57]. This point was first introduced in the form of user-producer interaction [31], [58] in which innovation is observed as emerging from a confrontation of user needs with technological opportunities. This situation entails interdependence in innovation endeavours between users and producers via the exchange of information and interactive learning. It implies that the ability to identify, articulate and communicate problems and possibilities on behalf of both users and producers is very important and that the competence of users is as important as the competence of producers [58], [59]. Consequently, the quantity and quality of interactive learning linkages are likely to improve the performance of the IS [58], [60]. These theoretical considerations can support ISF in conceptualising the demand for and users of innovation in foresight design.

Second, it is increasingly recognised that foresight is embedded in a wider context that influences both the foresight process and its potential impact on innovation activity [7], [8]. Foresight must be able to systematically and coherently account for context diversities and how this relates to innovation. According Schoen et al., the contextual nature of innovation is being recognised in foresight, but further work on the issue is lacking [9]. Foresight can learn important lessons about how and why context influences innovation from IS research.

Due to its foundation in evolutionary theorising on socio-technical change [61], [62], an IS is understood as open, evolving and recursive. One implication is that innovations follow certain and different trajectories across time and space in a path-dependent manner [63], [64]. Innovation dynamics differ across contexts because (i) industries depend on different knowledge bases and because technological opportunities differ across knowledge bases as a consequence of existing strongholds in production, science and technology; (ii) technological and innovation competencies, embodied in people and firms, are unequally distributed across space and time as a consequence of the specialisation of the industrial structure and the education system; and (iii) the quality and volume of demand for output differs across industries, resulting in diverse demand-pull effects [6]. Moreover, institutions relating to appropriability conditions, competition and market structure differ in importance across industries [32], [65].

In IS research, context is most often conceptually decoded by focusing on how institutions influence innovation dynamics of interest, which mainly occurs via its influence on interactive learning and knowledge flows in the economy.8 Other important context characteristics that influence innovation are industry structure, distributed knowledge bases [66], regulation and policies, infrastructure, and the external environment [67]. Hence, IS research can give foresight a systematic tool for decoding how context influences innovation.

8 Institutions are understood to be sets of common habits, routines, established practices, rules, or laws that regulate the relationships and interactions between individuals and groups. Moreover, organisations and institutions are seen as being distinct despite their interacting with and affecting one another. Organisations are actors such as firms, universities and states. Institutions, however, influence how actors behave [79].

10

4.2 Design and conceptual implications

In addition to accommodating research gaps, a principal reason for suggesting ISF is the need for theoretical underpinnings. The absence of such underpinnings makes it difficult for foresight to be carried out systematically, i.e., coherently with a theoretical framework to support decisions made before, during and after the foresight process. For example, one could argue that horizon scanning not founded in a theoretical framework would essentially be blind. Additionally, when one defines opportunities, threats, or bottlenecks, one needs a base from which to coherently distinguish between relevant and irrelevant factors.

Figure 1: Generic phases and steps in a foresight exercise—adopted and modified from Andersen and Rasmussen [78].

A foresight process can be described as consisting of three main phases: pre-foresight, foresight and post-foresight [68]. Each contains a number of steps; see Figure 1. ISF directly impacts pre-foresight and mapping and indirectly impacts post-foresight. Considering the structure of foresight in Figure 1, we can observe that, in foresight research, there is no theoretically founded reasoning behind (1) the goal of foresight, (2) the delimitation of the area of interest (system definition), (3) the selection of relevant participants, or (4) the criteria/design for understanding and analysing the present. We argue that the IS framework can supply a foundation for such decision making and can make it analytically coherent. Hence, the focus is not on the methods or techniques of foresight but on the implications of accepting ISF for the conceptual design of foresight.

4.2.1 Goal of foresight

According to Barré and Keenan, the following are the most common goals of foresight: (1) exploring future opportunities to set priorities for investment in science and innovation

11

activities; (2) reorienting the IS (this goal is related to priority setting but extends further—i.e., there may have been a preliminary diagnosis that the IS does not match the needs of the country); (3) demonstrating the vitality of the IS (in this context, foresight becomes a shop window in which to demonstrate the technological opportunities that are available); (4) bringing new actors into the strategic debate (a growing tendency is the use of foresight as an instrument to broaden the range of actors engaged in science and innovation policy); and (5) building new networks and linkages across fields, sectors and markets or around problems. Barré and Keenan [3] argue that foresight can have one or several of these goals, reflecting the diversity of its theoretical foundations.

From the perspective of ISF, it is possible to establish a hierarchy between the goals mentioned and then to establish channels of causality between them. The most crucial objective of ISF is to ‘strengthen’ the IS, which involves building, transforming or reorienting the system by removing barriers and promoting desirable innovation activities. Building new networks and linkages, bringing new actors into the strategic debate, mapping (demonstrating) the’ vitality’ of the IS, and exploring future opportunities to set priorities for investment in science and innovation are all instruments for achieving the improvement of an IS; see Figure 2. This implies that, to be meaningful, these instruments must be understood and designed as being embedded in an IS.

Figure 2: Hierarchy of and relationships between goals of foresight (developed by the authors).

As observed in the figure, we consider ISF to be a suitable tool for both strengthening and transforming an IS. This distinction echoes that of ‘innovation (in) system’ (shorter term optimisation of a relatively fixed structure) versus ‘system innovation’ (structural transformation over longer periods of time) [69]. The main difference between the two is the temporal perspective. Short-term ISF focuses on IS strengthening, whereas ISF for the medium to long term focuses on IS transformation. It is possible to connect these concepts because IS strengthening cannot be fundamentally separated from IS transformation. From an evolutionary perspective, it is misguided to think of an IS as a static structure that ‘produces’ innovation. An IS is never static but constantly evolving. There are feedbacks between innovation and system structures that necessarily co-evolve.

4.2.2 System definition and boundary setting

According to our knowledge, in foresight, there is no agreed-upon method for setting system boundaries and thus deciding which factors are external and internal. This is problematic for developing a systematic foresight method based in theory and for making comparisons between and generalisation of foresight exercises. The setting of initial boundaries influences

12

Means of foresightExploring future opportunities so as to set priorities for investment in science and innovation activities. Demonstrating vitality of the ISBringing new actors into the strategic debate. Building new networks and linkages across fields, sectors and markets or around problems.

Immediate goal

Strengthening and/or transforming the IS

Ultimate goal

Economic growth, international competi-tiveness, development and social well-being

the choice of methodology, data collection and stakeholder involvement in subsequent steps in the foresight process and is thus a crucial step.

As mentioned, an IS may be delimited spatially, sectorally or according to technological field. These determinants of limits are mutually complementary and may be applied individually or in a mixed manner depending on the objective of the research. Regardless of the IS type of interest, the organising principle for setting boundaries is that an IS should include the elements and relationships that interact in the production, diffusion and use of new and useful knowledge [31]. The main point is to have basic principles (based in theory) for how to set the boundaries and to explicitly argue for why they were decided as they were. In parallel, national, regional, sectoral and technology foresights exist, but their boundaries are not decided according a theory-founded principle. IS research does not deliver a finished tool. There is flexibility in and debate about how such boundaries are determined, which is reflected by the diversity of IS types. Still, embedded in these system types and the discussions around them is the accumulated knowledge about how to best understand innovation in certain contexts, which is valuable to foresight. Moreover, because ISs are open, interdependency and interaction between system levels and dimensions must be considered, cf. section 4.1.

4.2.3 Participation of actors

Which actors are considered relevant is closely related to where the system boundaries are set. IS strengthening and transformation requires a distributed innovation policy, which in turn requires (meaningful) participation by all key stakeholders. In general, it can be said that one should make sure that those who are responsible for making the decisions necessary for achieving the desired change are involved throughout the foresight process such that they feel a sense of ownership over its results [70]. It is therefore important to enrol actors that are representative of all elements and relationships that affect the innovation processes in a given system. Identifying these elements and actors is the first step in IS analysis. If the goal is IS transformation, an expert-based foresight will be insufficient. The main point is that because actors are observed as the primary agents of change, ISF must be ‘inclusive’ [71] to be transformational. Hence, ISF is necessarily systemic, distributed and inclusive. Moreover, the task of system definition decides which actors in industry, science, the government, the public and other public authorities that ought to participate in the foresight processes. System definition thus strongly influences the post-foresight phase and the likelihood of successful implementation. With a systems approach, the crucial systemic interdependency between pre-foresight and post-foresight activities emerges. Such systemic interdependencies between phases of foresight must be accounted for in the theoretical framework.

4.2.4 Mapping the present

To the best of our knowledge, in foresight, there is no agreed-upon method for analysing the present and for providing a deep understanding of the system for which foresight will be carried out. SWOT and STEEPV analyses are often used, but they are not based in a coherent theoretical framework nor do they give explicit guidelines for analysis.9 This hinders the

9 It has been argued that the SWOT framework pays insufficient attention to the (learning) capabilities of individual firms and to the fact that economic competition is increasingly innovation based [80]. This indicates that the framework is not developed to study the dynamics of the learning economy. It has also been argued that the SWOT framework is theoretically superficial and consequently not informative with respect to the content of the analytical steps involved. The latter has contributed to a SWOT practice that does not go beyond description in the most general terms. Such an analysis lacks rigor and analytical content [81]. Others have argued that SWOT is derived from

13

consistency and, in turn, the reliability of the framework. Moreover, these frameworks were not developed to grasp innovation processes and are thus weak in this respect. Mapping is an important step because the quality of any foresight will depend on the quality of the initial mapping exercise. ISF requires the analysis of the system’s current state with a focus on barriers and opportunities for innovation. The outcomes of the analysis can be used as organising topics for the further foresight process.

As a point for further exploration, we suggest the ‘functions approach’ to IS10 [38], [72], [73], [74], [75]. As opposed to earlier work in IS research, the functional approach is more focused on innovation processes rather than the more static structural and institutional features of systems. The latter is accommodated by an analytical distinction between components (actors, networks, and institutions) and key functions in the IS. Bergek et al. suggest using seven key functions central to IS performance to assess the overall dynamics of the system [76]. Such an analysis can illustrate weak points in the system that, in turn, can inform the foresight process. The analysis consists of six steps. First, define the IS of interest on the basis of, inter alia, research objective, breadth and depth, and spatial domain. Second, identify the structural components of the system. Third, map the key functions of the system. Fourth and fifth, assess the functioning of the system and identify the inducement and blocking mechanism for further IS strengthening and transformation. This information can, in turn, be used for formulating policy to improve the system (step six) [76]. This mapping will give a picture of current innovation dynamics and can inform the following steps of the foresight process. How the details of the analysis should be defined is beyond the scope of this paper, but guidance can be found in the literature.

5 Conclusion and perspectives The literature review of innovation studies and foresight research has illustrated that changes in the two areas have co-evolved over the years, with foresight importing the dominant understanding of innovation from innovation studies with a time lag. Currently, foresight is in a catching-up process vis-à-vis innovation studies by gradually incorporating the implications of a systemic understanding of innovation. The latter is reflected in current gaps in the foresight research that we addressed with the concept of innovation system foresight (ISF). We explored four preliminary implications of ISF that constitute avenues for further research. Our considerations are of a theoretical nature and concern the conceptualisation and design of foresight. Still, our analysis merely constitutes a first tentative step towards initiating a constructive dialogue between foresight and the IS approach.

ISF must cover the interdependent dimensions of science, technology and innovation and identify the synergies and barriers to their interaction. This must be conducted with equal attention to users and producers of knowledge and context parameters. At its core, ISF thus carries a critique of both science foresight and technology foresight (even if systemic) because these setups will find it difficult to identify actors in the demand side with whom to interact in the real economy. The absence of concrete recipients or users of generic technological developments (such as nanotechnology or robotics) undermines the effectiveness of foresight by (a) removing the possibility of interactive learning and feedback loops and (b) not enrolling down-stream agents of change that are needed for achieving (desirable) IS transformation.

theoretical underpinnings and therefore often produces shallow and misleading results [82].10 A functions approach has mainly been applied to technological IS analysis, but in principle, it is applicable to other types of IS.

14

Furthermore, the focus on demand suggested by us is different from the demand-focus related to the notion of grand social challenges. Consider two foresight designs. One consists of an expert group analysing which technologies will be critical for international competitiveness in 20 years. Another consists of a group of experts analysing which social needs will dominate in 20 years and, subsequently, which technologies will be in demand due to the latter. Ontologically speaking, the two foresights are not so different. Both processes are exclusive and non-transformational with respect to the IS. They analyse the demand side but do not include it. In ISF, neither scientific progress nor technological R&D can generate ready-made ‘optimal’ innovations regardless of how detailed an analysis of future demand is made. Instead, the focus should be on how technological opportunities, user competence and preferences, and context idiosyncrasies can find common ground going forward.11 The latter requires inclusive processes of interactive learning and vision building.

We have not discussed any implications for foresight methods, but the centrality of interactive learning may suggest that face-to-face interaction is a crucial element in achieving shared visions and, in turn, effective distributed policy. Another implication is that the type of foresight suggested requires additional skills from foresight managers. Conducting the IS analysis demands skilled efforts, and foresighting on the basis of its outcome will be a complex and costly task. The paper points toward several areas of future research. One is to develop in more detail how to actually perform the IS analysis and how to integrate the results with the foresighting activities. Additionally, implications for foresight methods can be explored further. The development of systemic methods [18] seems a fruitful path to follow.

When Martin and Johnston in 1999 framed technology foresight as a tool for wiring up ISs, their main point was that it could increase the quantity of interactions in an IS [14]. ISF adds to this perspective. The latter wiring-up suffers from the conceptual design problems, typical to technology-oriented foresight discussed in this paper. It does not explicitly contain a systemic and evolutionary understanding of innovation, and it is not thought out along the lines of intentionally transforming a whole IS. The latter implies that ISF has the potential to influence the direction of innovation and IS transformation rather than ‘blindly’ maximising linkages around random technologies. Moreover, by being inclusive, ISF goes beyond quantity and adds a focus on the quality of interactions (the learning in them). These points echo Georghiou’s [77] argument that ‘foresighting for development’ (and ISF) may not only make successful innovation more likely but also shape the direction of innovation towards solutions to important social problems.

6 References

[1] I. Miles, J. C. Harper, L. Georghiou, M. Keenan, and R. Popper, “The Many Faces of Foresight,” in in The Handbook Of Technology Foresight: Concepts and Practice, L. Georghiou, J. Harper, M. Keenan, I. Miles, and R. Popper, Eds. 2008, pp. 3–23.

[2] É. Hideg, “Theory and practice in the field of foresight,” Foresight, vol. 9, no. 6, pp. 36–46, 2007.

[3] R. Barré and M. Keenan, “Revisiting Foresight Rationales: What Lessons from the Social Sciences and Humanities?,” in in Future-Oriented Technology Analysis -

11 One may argue that the focus on grand social challenges stems from the increased pressure on public science to yield higher social returns by addressing social problems. It does not stem from the insight that demand-side inclusion in long-term policy-making processes is needed to make it effective.

15

strategic Intelligence for an Innovative Economy, C. Cagnin, M. Keenan, R. Johnston, F. Scapolo, and R. Barré, Eds. Springer, 2008.

[4] T. Fuller and K. Loogma, “Constructing futures: A social constructionist perspective on foresight methodology,” Futures, vol. 41, no. 2, pp. 71–79, Mar. 2009.

[5] M. A. Öner, “On theory building in Foresight and Futures Studies: A discussion note,” Futures, vol. 42, no. 9, pp. 1019–1030, Nov. 2010.

[6] G. Dosi, “Sources, procedures, and microeconomic effects of innovation,” Journal of economic literature, vol. 26, no. 3, pp. 1120–1171, 1988.

[7] M. Cariola and S. Rolfo, “Evolution in the rationales of foresight in Europe,” Futures, vol. 36, no. 10, pp. 1063–1075, Dec. 2004.

[8] R. Barré, “Foresight and their themes: analysis, typology and perspectives,” in Conference Proceedings: The role of foresight in the selection of research policy priorities, Seville, Spain, 2002.

[9] A. Schoen, T. Könnölä, P. Warnke, R. Barré, and S. Kuhlmann, “Tailoring Foresight to field specificities,” Futures, vol. 43, no. 3, pp. 232–242, Apr. 2011.

[10] R. Smits, S. Kuhlmann, and P. Shapira, The Theory and Practice of Innovation Policy: An International Research Handbook. Edward Elgar, 2010.

[11] R. Smits and S. Kuhlmann, “The rise of systemic instruments in innovation policy,” International Journal of Foresight and Innovation Policy, vol. 1, no. 1/2, p. 4, 2004.

[12] L. Georghiou and J. Cassingena Harper, “From priority-setting to articulation of demand: Foresight for research and innovation policy and strategy,” Futures, vol. 43, no. 3, pp. 243–251, Apr. 2011.

[13] C. Cagnin, E. Amanatidou, and M. Keenan, “Orienting European innovation systems towards grand challenges and the roles that FTA can play,” Science and Public Policy, vol. 39, no. 2, pp. 140–152, Apr. 2012.

[14] B. R. Martin and R. Johnston, “Technology foresight for wiring up the national innovation system: experiences in Britain, Australia, and New Zealand,” Technological Forecasting and Social Change, vol. 60, no. 1, pp. 37–54, Jan. 1999.

[15] F. Alkemade, C. Kleinschmidt, and M. Hekkert, “Analysing emerging innovation systems: a functions approach to foresight,” International Journal of Foresight and Innovation Policy, vol. 3, no. 2, pp. 139–168, 2007.

[16] M. Weber, P. Schaper-Rinkel, and M. Butter, “Sectoral Innovation Foresight: Introduction,” 2009.

[17] B. Dachs and M. Weber, “Sectoral Innovation Foresight Textiles and Clothing Sector,” Europe INNOVA, www.europe-innova.eu/web/guest/sectoral-innovation-watch/publications-tools, 2010.

16

[18] O. Saritas, “Systemic Foresight Methodology,” in 4th International conference on Future-Oriented Technology Analysis and Grand Societal Challenges – Shaping and Driving Structural and Systemic Transformations, Seville, Spain, 2011.

[19] M. Atilla Oner and O. Saritas, “A systems approach to policy analysis and development planning,” Technological Forecasting and Social Change, vol. 72, no. 7, pp. 886–911, Sep. 2005.

[20] O. Saritas and M. A. Oner, “Systemic analysis of UK foresight results,” Technological Forecasting and Social Change, vol. 71, no. 1–2, pp. 27–65, Jan. 2004.

[21] Jantsch and Erich, Technology forecasting in perspective. Paris: OECD, 1967, p. 401.

[22] I. Miles, “The development of technology foresight: A review,” Technological Forecasting and Social Change, vol. 77, no. 9, pp. 1448–1456, Nov. 2010.

[23] W. Bell, Foundations of Futures Studies: History, Purposes, and Knowledge, New Brunswick (USA): Transaction Publishers, 2003, p. 390.

[24] B. R. Martin, “Foresight in science and technology,” Technology Analysis & Strategic Management, 1995.

[25] I. Miles, “From Futures to Foresight,” in in The Handbook Of Technology Foresight: Concepts and Practice, L. Georghiou, J. C. Harper, M. Keenan, I. Miles, and R. Popper, Eds. Edward Elgar Pub, 2008.

[26] J. Irvine and B. R. Martin, Foresight in science: picking the winners. Pinter London, 1984.

[27] Norwegian Research Council, “Foresight i Norge 2009. Mot et nytt kunnskapsfelt,” Norwegian Research Council, 2010.

[28] J. Fagerberg, H. Landström, and B. R. Martin, “Exploring the emerging knowledge base of ‘the knowledge society’,” Research Policy, vol. 41, no. 7, pp. 1121–1131, Sep. 2012.

[29] J. Fagerberg and K. Sapprasert, “National innovation systems: the emergence of a new approach,” Science and Public Policy, vol. 38, no. 9, pp. 669–679, Nov. 2011.

[30] C. Freeman, “Preface,” in in Putting Africa first - The making of African innovation systems, M. Muchie, P. Gammeltoft, and B. Å. Lundvall, Eds. Aalborg University Press, 2003.

[31] B.-Å. Lundvall, Ed., National Systems of Innovation – Toward a Theory of Innovation and Interactive Learning. London: Pinter, 1992.

[32] R. Nelson and N. Rosenberg, “Technical Innovation and National Systems,” in in National Innovation Systems - A Comparative Analysis, R. R. Nelson, Ed. Oxford University Press, USA, 1993.

17

[33] C. Freeman, Technology policy and economic performance: Lessons from Japan. London: Pinter, 1987.

[34] P. Cooke, M. G. Uranga, and G. Etxebarria, “Regional innovation systems: Institutional and organisational dimensions,” Research policy, vol. 26, no. 4–5, pp. 475–491, 1997.

[35] B. T. Asheim and M. Gertler, “The geography of innovation: regional innovation systems,” in in The Oxford handbook of innovation, J. Fagerberg, D. Mowery, and R. Nelson., Eds. Oxford University Press, USA, 2005.

[36] F. Malerba, “Sectoral systems of innovation and production,” Research Policy, vol. 31, no. 2, pp. 247–264, Feb. 2002.

[37] F. Malerba, “Sectoral systems: how and why innovation differes across sectors,” in in The Oxford handbook of innovation, J. Fagerberg, D. Mowery, and R. Nelson, Eds. Oxford University Press, USA., 2005.

[38] A. Bergek, M. Hekkert, and S. Jacobsson, “Functions in innovation systems: A framework for analysing energy system dynamics and identifying goals for system-building activities by entrepreneurs and policy makers,” in in Innovation for a low carbon economy: economic, institutional and management approaches, T. J. Foxon, J. Köhler, and C. Oughton, Eds. Edward Elgar, 2008.

[39] B. Carlsson and R. Stankiewitz, Technological systems and economic performance: the case of factory automation. Dordrecht: Kluwer Academic Publishers, 1995.

[40] G. Reger, “Technology Foresight in Companies : From an Indicator,” Technology Analysis & Strategic Management, vol. 13, no. 4, 2001.

[41] L. Georghiou, “Third Generation Foresight - Integrating the Socio-economic Dimension,” NISTEP, 2001.

[42] G. Tegart and R. Johnston, “Some Advances in the Practice of Foresight,” in EU-US Seminar: new technology foresight, forecasting and assesment methods, Seville, Spain, 2004, no. May.

[43] D. Schlossstein and B. Park, “Comparing recent technology foresight studies in Korea and China: towards foresight-minded governments?,” Foresight, vol. 8, no. 6, pp. 48–70, 2006.

[44] H. A. Linstone, “Corporate planning , forecasting , and the long wave,” Futures, vol. 34, pp. 317–336, 2002.

[45] R. Rothwell, “Towards the Fifth-generation Innovation Process,” International Marketing Review, vol. 11, no. 1, pp. 7–31, 1994.

[46] M. Dodgson, D. Gann, and A. J. Salter, Think, play, do: technology, innovation, and organization. Oxford University Press, USA, 2005.

18

[47] B.-Å. Lundvall and S. Borrás, “Science, technology and innovation policy,” in in The Oxford handbook of innovation, J. Fagerberg and D. Mowery, Eds. Oxford University Press Oxford, 2005.

[48] S. J. Kline and N. Rosenberg, “An overview of innovation,” in in The Positive Sum Strategy: Harnessing Technology for Economic Growth, R. Landau and N. Rosenberg, Eds. Washington D.C: National Academy Press, 1986.

[49] T. Kuwahara, “Technology Forecasting Activities in Japan,” Technological Forecasting and Social Change, vol. 60, no. 1, pp. 5–14, Jan. 1999.

[50] B.-Å. Lundvall and B. Johnson, “The Learning Economy,” Journal of Industry Studies, vol. 1, no. 2, pp. 23–42, Nov. 1994.

[51] K. Foss and N. J. Foss, “Authority in the Context of Distributed Knowledge,” DRUID Working Papers, vol. 03–08, 2002.

[52] K. Smith, “Industrial structure, technology intensity and growth: issues for policy,” in DRUID conference on National Innovation Systems, Industrial Dynamics and Innovation Policy, Rebild, Denmark, June 9-12, 1999, 1999.

[53] M. Butter, F. Brandes, M. Keenan, and R. Popper, “Editors’ introduction to the European Foresight Monitoring Network,” foresight, vol. 10, no. 6, pp. 3–15, 2008.

[54] A. L. Porter, “Technology foresight: types and methods,” International Journal of Foresight and Innovation Policy, vol. 6, no. 1/2/3, p. 36, 2010.

[55] E. A. Eriksson and M. Weber, “Adaptive Foresight: Navigating the complex landscape of policy strategies,” Technological Forecasting and Social Change, vol. 75, no. 4, pp. 462–482, May 2008.

[56] J. Edler and L. Georghiou, “Public procurement and innovation—Resurrecting the demand side,” Research Policy, vol. 36, no. 7, pp. 949–963, Sep. 2007.

[57] B. Johnson, “Institutional Learning,” in in National Systems of Innovation – Toward a Theory of Innovation and Interactive Learning, B. A. Lundvall, Ed. London: Pinter, 1992.

[58] B.-Å. Lundvall, “Product Innovation and User-Producer Interaction,” Industrial Development Research Series, Aalborg University Press, no. 31, 1985.

[59] L. Pasinetti, Structural economic dynamics: a theory of the economic consequences of human learning. Cambridge University Press, 1993.

[60] J. Fagerberg, D. C. Mowery, and B. Verspagen, “The evolution of Norway’s national innovation system,” Science and Public Policy, vol. 36, no. 6, pp. 431–444, 2009.

[61] B. R. Martin, “The evolution of science policy and innovation studies,” Research Policy, vol. 41, no. 7, pp. 1219–1239, Sep. 2012.

19

[62] B.-Å. Lundvall, B. Johnson, E. S. Andersen, and B. Dalum, “National systems of production, innovation and competence building,” Research policy, vol. 31, no. 2, pp. 213–231, Feb. 2002.

[63] W. B. Arthur, Increasing Returns and Path Dependence in the Economy. The Univeristy of Michigan Press, 1994.

[64] G. Dosi, “Technological paradigms and technological trajectories A suggested interpretation of the determinants and directions of technical change,” Research Policy, vol. 11, no. 3, pp. 147–162, Jun. 1982.

[65] K. Pavitt, “Sectoral patterns of technical change: Towards a taxonomy and a theory,” Research Policy, vol. 13, no. 6, pp. 343–373, Dec. 1984.

[66] P. Robertson and K. Smith, “Distributed knowledge bases in low- and medium-technology industries,” in in Innovation in Low-tech Firms and Industries, H. Hirsch-Kreinsen and D. Jacobsen, Eds. Aldershot: Edward Elgar, 2008.

[67] B. Gregersen, B. Johnson, and A. Kristensen, “Comparing national systems of innovation. The case of Finland, Denmark and Sweden,” in in Explaining Technical Change in a Small Country: The Finnish National Innovation System, 1994.

[68] B. R. Martin and J. Irvine, Research Foresight: Priority-Setting in Science. London: Pinter, 1989.

[69] F. W. Geels, “Processes and patterns in transitions and system innovations: Refining the co-evolutionary multi-level perspective,” Technological Forecasting and Social Change, vol. 72, no. 6, pp. 681–696, 2005.

[70] C. Cagnin, “Aspects to consider when designing foresight processes,” in in Foresight International Seminar: from theory to practice, Brasilia: CGEE, 2011.

[71] D. Loveridge, “Inclusive foresight,” Foresight, vol. 7, no. 3, pp. 31–47, 2005.

[72] S. Jacobsson and A. Bergek, “Innovation system analyses and sustainability transitions: Contributions and suggestions for research,” Environmental Innovation and Societal Transitions, vol. 1, no. 1, pp. 41–57, 2011.

[73] A. Bergek, S. Jacobsson, and B. a. Sandén, “‘Legitimation’ and ‘development of positive externalities’: two key processes in the formation phase of technological innovation systems,” Technology Analysis & Strategic Management, vol. 20, no. 5, pp. 575–592, Sep. 2008.

[74] S. Jacobsson and A. Bergek, “A framework for guiding policy makers intervening in emerging innovation systems in ‘catching up’ countries,” European Journal of Development Research, vol. 18, no. 4, 2007.

[75] S. Jacobsson, “Transforming the energy sector: the evolution of technological systems in renewable energy technology,” Industrial and Corporate Change, vol. 13, no. 5, pp. 815–849, Oct. 2004.

20

[76] A. Bergek, S. Jacobsson, B. Carlsson, S. Lindmark, and A. Rickne, “Analyzing the functional dynamics of technological innovation systems: A scheme of analysis,” Research Policy, vol. 37, pp. 407–429, Feb. 2008.

[77] L. Georghiou, “Future of foresighting for economic development,” in UNIDO Expert Group Meeting on the Future of Technology Foresight, 2007, no. May 2007.

[78] P. D. Andersen and B. Rasmussen, Fremsyn: Metoder, praksis og erfaringer. Danish Agency for Science, Technology and Innovation, Mnistry of Science, Innovation and Higher Education, 2012.

[79] C. Edquist and B. Johnson, “Institutions and Organizations in Systems of Innovation,” in in Systems of Innovation, C. Edquist, Ed. London: Pinter, 1997.

[80] M. H. Zack, “Introduction,” in in Knowledge and Strategy, M. H. Zack, Ed. A Butterworth-Heinemann, 1999.

[81] T. Hill and R. Westbrook, “SWOT analysis: It’s time for a product recall,” Long Range Planning, vol. 30, no. 1, pp. 46–52, Feb. 1997.

[82] E. K. Valentin, “SWOT analysis from a resource-based view,” Journal of Marketing theory and Practice, vol. 9, no. 2, pp. 54–69, 2001.

21