This study has sought to use empirical evidence to provide some ideas on how innovations processes occur under the so-called ‘green transitions’. In this section, we analyse our empirical evidence on the basis of the theoretical debate presented in PART I. Hence, this chapter is devoted to analyse the role of innovative entrepreneurs, institutional entrepreneurship, and place-based leadership in ‘green innovations’ and in setting directionality (see Figures 1, 2).

The relevance of singling out ‘green innovations’ from other, more general, ‘innovations’ is because the former are politically willed and specifically aimed at addressing the environmental crisis. This is what Grillitsch et al. (2019), among other authors, argue is characteristic of the new wave of innovation policies – Transformative Innovation Policy (TIP) - explicitly mobilising science, resources and innovation efforts to meet societal challenges. Different cases of technological or sectoral transformative innovation occur from fundamentally different starting points, where innovations (and new development paths) are triggered either by a policy push, a technology push (by private businesses) or a demand pull (by clients/consumers). Empirical evidence gathered from our case studies in distinct industries, countries and regions shows that there is no black-and-white distinction of which particular actor is responsible for instigating change. The fact that there are policies that specifically target green innovations and transitions says as much about pull-push forces as the innovations that occur without the need for a policy push. However, regulations and policies are generally not the starting point of a transition, as they tend to be the result of a broader process of societal problem recognition and a myriad of scientific and product developments. These are often ignored when tracing the chronological order of events, as there is often no precise moment, actor or action that can be attributed as the starting point of everything. Instead, it is the sum of related and unrelated actions by different agents that lays the groundwork. This can suddenly take on new meaning once a window of opportunity opens, effected by a regulatory change, for instance, although other changes may also play a part. 

Green innovations also require a demand pull unless legislation demands the use of a specific technology or resource (e.g., via quota schemes) or by providing no alternative (e.g. by introducing bans or penalties that would push other options out of the market). In the case studies presented above, regulation and policy goals play an important role in boosting green innovation but do not, for the moment, rule out other options. However, in the wood construction case, we found evidence of more indirect ways of excluding alternatives: not necessarily conspicuously favouring wood buildings in new development programmes, but by municipalities setting ambitious carbon emissions requirements and/or by hinting what desired design features in newly developed areas could include. Similarly, in the protein shift case, conventional protein products are not being brushed aside, but there are examples of municipally or publicly owned schools that are trying to reduce their overall meat and dairy consumption by introducing more varied diets in their meal offers, while at the same time encouraging local companies to provide innovative food products. This underscores the soft power and influence that many municipalities have in the Nordic countries, where there is much room for manoeuvre despite the parameters set out by national and supra-national authorities. 

These direct and indirect, intentional and unintentional ways of pushing for change suggest that there is a need to adapt our traditional structure-agency debate understanding and adopt a more systemic understanding of innovation processes. It is evident that the current policy agenda represents a major change in the ‘structure’ – the rules of the game. From international agreements, such as the Paris Agreement and Agenda 2030, the EU Green Deal and Horizon Europe at EU level, to national legislation and sectoral policy and sub-national development strategies comes a substantial and clear message from the authorities: the economy needs to go ‘green’. However, examining in detail the micro-processes leading to individual innovations or sectoral transformations reveals a much more complex and chaotic picture of who the drivers are, the roles the different actors play and their interactions. While changes in legislative and regulatory efforts entail systems change, exploring new economic pathways or industrial path creation is a multi-level and multi-actor process.

The Schumpeterian understanding of entrepreneurship as the exercise of creative labour, opportunity recognition and idea exploitation rings true in our case studies. We recognise the ability of entrepreneurs, be they individuals or businesses, to identify windows of opportunity and take clear steps to avail of them, including solving technological bottlenecks, exploring the commercial viability of new products and solutions, developing a business model, etc. This does not necessarily mean, however, that entrepreneurs are gazing fixedly into a crystal ball, but proves their agility in applying existing knowledge and ideas to emerging opportunities occasioned by contextual changes for instance, in novel ways. Our empirical evidence supports this conclusion. Wood construction is not new, but changing contexts, both in terms of legislative and societal demands, brought new opportunities, which businesses were quick to identify and exploit. Many innovations have taken place, but mostly in incremental ways. Thus, in our first case study, the well-established companies assumed the pioneering role in developing new solutions for multi-storey wood building, as they had sufficient resources and built capacity to embark on costly investments. Start-ups, on the other hand, have played a greater role at a later stage by addressing technological and market gaps in relation to innovative materials – complementary to the industry. 

In other cases, however, exploiting market opportunities does involve the application of unique trade secrets, patents and breakthrough innovations. This is more clearly visible in the protein shift where the holders of such “secrets” or new knowledge are often scientists, some of whom chose to become entrepreneurs themselves or work in tandem with existing businesses to exploit the knowledge they hold. We could, therefore, identify two distinct paths in which companies exploit knowledge when conducting product development. Either they apply analytical knowledge (basic science) or synthetic knowledge, as in existing knowledge used to solve immediate problems. One pertinent observation in the protein case study is that larger and well-established companies tend to use synthetic knowledge, as in known processes and inputs used for imitating meat, including soy, peas and beans, whereas smaller start-ups and scale-ups more often work with basic science in laboratories, with niche products and radical or disruptive technologies such as precision fermentation and cellular agriculture. This can also be seen reflected in the backgrounds of those engaged in protein development. In companies such as BioMass Protein, ORF Genetics, Gelatex, Solar Foods and Mycrorena, the founders and their staff seem to primarily consist of natural scientists, whereas in contrast, business entrepreneurs are behind companies such as the Swedish fish-substitute start-up Hooked Foods and the pea-based company Bärta. That being said, these companies do not work in a vacuum and all interconnect with relevant knowledge brokers to gain the skills and insights needed for the development and marketing of their products. 

Therefore, at least in the early phases, large and small businesses appear to contribute to the protein shift in different ways: One contributes new knowledge through more radical products and inventions, and the other instigates endogenous change, improving and adapting existing products and commercialising and scaling-up production. However, this reality appears to be changing, as transnational companies, many of which have a dominant stake in the traditional food sectors’ value chains, are now establishing or absorbing much of the R&D activities related to novel protein food products, thus securing their stake in emerging market niches. This suggests that once new ingredients and products become part of the standard offering on supermarket shelves, new products and innovations become systematised and controlled by larger companies. 

In relation to change agents, both smaller and larger companies have a role to play in creating the conditions and contexts for the protein transition. As part of exploiting niche markets, smaller companies face a prerogative to diversify their path creation and, as such, contribute to creating a more dynamic technological, sectoral or regional market. Smaller companies in the protein sector are creating new pathways or extending existing ones. Beyond product development processes, which is a key role for entrepreneurs, it would be wrong to assume that business entrepreneurs are the sole instigators of systems transformation. In the cases under review, innovative entrepreneurship cannot be seen as wholly separate from institutional and place-based developments, as well as the more network-based processes of innovation in which businesses engage in co-operative development with multiple actors. 

It is clear from both case studies that entrepreneurs are currently capitalising on the favourable political context within which they find themselves, as well as their ability to be part of the solution to significant societal challenges (institutional demand). The political winds in Denmark in favour of finding alternative protein sources to diversify their pork production have clearly created a positive context that works in favour of protein businesses and start-ups, such as BioMass Protein. Likewise, the explicit regulatory developments pushing for a substantial cut in carbon emissions in the construction sector send a clear signal to businesses in favour of wood-based solutions. Moreover, the roles of different actors cannot be explained in a simple structure-agency relationship whereby the changes in structures enable businesses to conduct the creative labour on their own, but rather by public authorities, scientists, as well as other intermediate actors such as science parks or cluster organisations becoming involved in co-development and driving forward processes of system innovation and industrial path creation together. 

What is the role of the state and public authorities when it comes to green innovation? We identify a wealth of examples of institutional innovation in our case studies, not only from the Schumpeterian understanding of public institutions as simply ‘adapting’ as part of the evolutionary change taking place in society at large but also in line with organisational theory and institutional entrepreneurship in which the public sector is understood to have a more proactive and intentional role in building competitive advantage. By shaping the baseline structures and pushing for politically-willed developments, public institutions are not merely fixing market failures but actively engaging with innovation processes and setting a direction towards economic prosperity or environmental sustainability, for example (path creation). In the wood construction case, the removal of bans or restrictions on multi-storey building in wood are clear examples of fixing market failures. Nor did the state authorities stop with these first-step measures, but continued by setting out a clear development path: from nursing innovation and the creation of a market niche by drafting ambitious strategies for development to more directly influencing paths by establishing R&D programmes and funding networks to support development (e.g. Wood City Sweden). At the sub-national level, certain municipalities and regions took an even more prominent role in fuelling innovation systems by mobilising local actors and enabling the flow of knowledge and learning, as well as boosting market expansion by using their role as a major client to purchase public buildings made of wood. In some cases, municipal engagement goes even further. Skellefteå municipality has directly spearheaded building new schools and other municipally funded projects utilising public procurement as one tool for supporting the wood construction industry in the region. 

However, we observed that municipalities' ability to act varies significantly based on key contextual differences. While both Skellefteå and Växjö municipalities lie within the same country (Sweden) and are thus subject to identical legal frameworks and baseline conditions, Skellefteå has a more dominant position in controlling development, thanks to its substantial financial resources, land ownership and a local housing company under its control. Växjö, on the other hand, finds itself in a less privileged position, although still possibly better off than other municipalities at home and abroad. Nevertheless, both municipalities represent successful examples of pushing forward industry development. What is interesting here is that local authorities, as institutional entrepreneurs, have had to adopt different strategies to foster change by adapting to their own specific regional conditions. In some cases, institutional entrepreneurship can take the form of a single person, a visionary with strong leadership capacity. For instance, the chair of the municipal council of Skellefteå and governor of Västerbotten Region at the time pushed for wood construction at a very early phase, even before a national strategy was in place. Achieving sustainability or innovation was not the main goal per se, but rather shaping the economic development path of the region after years of decline within the manufacturing industries. The city’s success in advancing wood construction can be attributed to its capacity to leverage unique strengths, resources and historical roots. In this sense, the public official took the entrepreneurial discovery process upon himself. 

Institutional entrepreneurship at local level has also revolved around circumventing national regulation, i.e. the material neutrality principle established in Sweden’s building codes, to support desired developments. This has been made possible through the help of other softer tools, such as presenting the type of architecture envisaged in a certain area, in combination with more blunt instruments such as public procurement (without directly demanding specific material use). 

Institutional entrepreneurship also takes place in the form of new organisations, hybrid structures or networks aimed at pushing certain developments. The creation of Wood City Sweden (which the former governor of Västerbotten has chaired) provides an example, first in the form of a project and then an association of municipalities and regions formed specifically to address the bottlenecks in the market of wood buildings. Another example is the creation of the Wood Innovation cluster led by the municipality of Skellefteå to bring together market, science and public sector actors to co-ordinate innovation actions. The relevance of such hybrid structures corroborates Grillitsch and Sotarauta’s (2020) conclusion that multi-actor involvement, as well as the right institutional drivers to renew, create or extend existing innovation paths are key to pooling powers, resources and skills. What is interesting to note is the role individuals assume in multi-actor partnerships is not necessarily dependent on the institutions they represent but more on the character of the person. For instance, individuals can act as leaders, visionaries, mentors, critics or ‘vision brokers’, regardless of their formal position or employer. Individuals, in effect, can change employer yet remain connected to, or influential for, a specific development by ‘wearing a different hat’ or representing a different ‘actor’.

In the protein shift case study, it is clear that there is a general societal megatrend pushing urgently in relation to climate change. This is reflected in political policies at supranational, national and local levels. The EU Green Deal and Horizon, the Farm to Fork strategy and the CAP are all increasingly adopting mission-oriented objectives to steer development and innovation towards increasingly sustainable production systems. On the one hand, tighter regulation aims to reduce the negative externalities of pollution, emissions and loss of biodiversity by setting limits on land use and the use of fertilisers and, on the other, through boosting innovative solutions by allocating substantial funding for R&D and innovation. VINNOVA, Sweden’s innovation agency has put these ambitions into practice at a national level by funding R&D on novel foods and plant proteins, for example. While this mission approach in some specific funding programmes may lead to ‘picking winners’, our overall view is that many of these investments are actually ‘picking the losers’, as in choosing whom not to fund older industries and technologies based on their negative impact on society and the environment. However, neither are these alternatives (and competing developments), such as conventional concrete and steel construction systems or livestock farms being directly banned or restricted: often, they do not need additional funding, as they are already well-positioned within the market. Indeed, in the latter case, we find EU-level agricultural subsidies that benefit both sustainable and unsustainable farming (although the CAP is increasingly setting requirements for subsidies in line with the sustainability goals). 

The substantial impact of public policy and more active institutional entrepreneurship challenges the notion that the private sector is inherently innovative, takes risks and pushes the boundary of existing markets. While a number of pioneers have taken steps to challenge the status quo, such as investing in multi-storey wood construction and developing novel food products, a large portion of private entities not only remain within their comfort zones but also actively resist change. In the construction sector, resistance is evident across the entire supply chain, from contractors to banks and insurance companies. Similarly, in the food sector, traditional corporations and food-producing companies have, for the most part, resisted change. Nonetheless, once it has been demonstrated that a promising market for wood buildings or alternative protein products genuinely exists, many companies, including those who initially resisted change, rapidly pivot and follow the trends. However, as we have also seen, public actors are also in place, promoting and pushing new agendas. Therefore, the concept of the 'entrepreneurial state' is perfectly applicable in these case studies. Without the active involvement of the state and sub-national authorities, private actors might have struggled to make significant progress.

Moreover, institutional entrepreneurship also occurs beyond the realm of public institutions in the form of normative institutional and societal behaviour changes. Altering consumer patterns and demand is another form of institutional entrepreneurship that challenges the baseline conditions, or rules of the game from the societal side - thus creating a market for new businesses and new products. Innovation is, as are most things, discourse and language dependent: what we say and how we frame it matters. Indeed, the clearly evident change in values, as well as the meta-narratives identified for and against certain building materials or food products, have been important determining factors in the development of new market niches. We will address this more informal dimension of institutional innovation in the next section. 

As innovations and entrepreneurial discovery processes exist within a social context, social change determines the scope of the opportunities that facilitate the emergence of these innovations. However, it is not always clear whether social change is the cause or the effect of innovations. Our empirical evidence shows that the reality is more chaotic. Given that many actors are focusing on ‘green innovation’ specifically, as opposed to other innovations, it suggests that the political and social context is prepared for innovations which are specifically situated within the ‘green’ sphere. This may suggest further that policy initiatives follow societal change and can, therefore, be seen as a policy response to climate concerns. This does not necessarily imply that society has changed its opinion on any given matter, but at least that a sufficient critical mass pushing for new policy initiatives has been reached. Society is not homogenous. Whilst some voices may become louder many others remain silent either due to lack of awareness, indifference or because the issue at stake is too complex for simplistic opinions.

In the protein shift case study, we have seen that there are at least three meta-narratives comprising a myriad of different actors with related views associated with the protein shift or sustainability within the food industry. Disagreements around the problem definition exist, which in turn impacts the choice and form of solutions. Moreover, the manifestation of different dominating discourses can be identified in various communities and although this may bode well for mutual understanding and common problem-solving approaches, it can also act as a brake on the development of alternative solutions (as we saw in the case of excessive cognitive proximity). The potential for innovation, be it technological advancements, institutional-regulatory changes or shifts in behavioural-societal dynamics, hinges on the impact of these overarching narratives on societal and political discourse. The translation of these narratives into action by decision-makers, businesses and society shapes innovation opportunities. For example, product developers innovating in the food space target different consumer groups based on these groups’ perceived values. Yet it remains unclear whether changing values will lead to radical regime change, with one technology/product category dominating the market, or a case of multiple technological ‘regimes’ with diversified market offerings that meet the demands of social groups with different values. What is clear is that dominating meta-narratives have an impact on the push-and-pull forces manifested in consumer choices, policy decisions and what businesses offer. 

The wood construction case witnessed a similar situation, where broader social support for sustainability has led to an increase in the use of wood materials, even though it remains highly unlikely that other materials will be simultaneously phased out. Although the role of social innovation, behaviour and perceptions seems to have a more direct impact on the protein shift, where individual consumer choice plays a central role, these factors are also observable in the multi-storey wood construction path creation. Changing social values appear to have impacted the perception of modern wooden structures where they come to represent both a renewed elevated status and revitalisation of urban areas. However, fear of fire and mould also act as deterrents to development, more so for investors, contractors and other players in the value chain, such as banks and insurance companies, than the final consumer. Uncertainty remains an important threat to change. Similarly, as in the protein case study, we see that competing narratives continue to vie for attention, e.g. the juxtaposition of concerns regarding wooden structures and deforestation with narratives emphasising sustainability and health benefits. Pilot projects have an important role to play, especially at a time when a dominant narrative is being challenged, as they create a space for an alternative narrative to unfold. In the wood construction case, one such pilot project is the architecturally ambitious high-rise SARA Kulturhus in northern Sweden. 

Narratives are connected to people’s perceptions of different vying alternatives and, ultimately, to choices that people and institutions make (or which are available to them). It is particularly interesting that the debate is often couched in black-and-white terms when the answers to the concerns about sustainability and ethics often lie in an extended grey zone.

When analysing the role of a place in innovation systems, a fundamental question arises: how do we define the system? Defining the system depends on what represents the most appropriate level of aggregating the relevant actors, resources and series of conditions that define a system of innovation. The literature suggests different alternatives, including national, regional, technological and sectoral innovation systems. The key differences are whether the emphasis is placed on the geographical dimension and physical boundary that encloses the system or on the networks and connections between relevant actors, regardless of location. In practice, the most relevant way to aggregate the system often depends on the industry, sector and technology that we are investigating, as well as the ways actors and value chains are organised. To disentangle this riddle, we will avail of Bochma’s (2005) understanding of proximities to analyse our empirical findings. For Boschma, ‘proximity’ refers to the ability to understand, adopt and adapt, to identify novelty, interpret and exploit new knowledge. This occurs across five dimensions, namely geographical, cognitive, organisational, social and institutional proximities. These dimensions of proximity may or may not exist simultaneously and may become more or less relevant depending on the type of innovation and phase of development. Moreover, proximities can either produce positive effects by aiding the flow of information and path creation or negative effects by hindering the influx of new ideas, causing technological, institutional and behavioural lock-ins that resist change and innovation.

Geographical proximity, the colloquial understanding of proximity, is relevant to national and especially regional innovation systems. Geography generally determines regions’ industrial legacy, the resources available, the knowledge and skills, the established networks, the institutional structures, as well as ‘tacit knowledge’ or more implicit societal norms or ‘ways-to-do-things’. This is especially prominent in the case of wood construction, where the industrial legacy and proximity to resources have produced a positive lock-in, allowing traditional forestry regions to lead the innovation processes in wood construction. Localised knowledge spillovers became useful when structural change, in the form of changed legislation, enabled new developments, as in companies developing new solutions for multi-story wood construction. Under new dynamic networks, innovation and transformational changes emerge and grow as there is a short ‘distance’ between key players across public institutions, companies, municipalities and academia. This short distance can refer to geographical proximity in place developments, such as in Skellefteå or Växjö, but also cognitive proximity across longer distances via intensive collaboration.

In the case study on protein transitions, we see that the role of a “place” also matters less so in connection to specific typologies such as rural or urban qualities but more in relation to different forms of proximity. Based on the type of company, product and location, it appears that the proximity to specific qualities in a place is important, such as access to university facilities or a highly specialised labour force. In these cases, the existing knowledge bases available to the companies in these areas shape that company’s trajectory and, therefore, their ability to innovate, change and learn (Boschma, 2005; Hansen, 2015; Zukauskaite and Moodysson, 2016). However, there is a dual situation. While start-ups tend to stay close to universities or large university cities, the suppliers of input factors (raw materials) are internationally based, possibly due to the spatially blind nature of some of the proteins (e.g., larvae, yeast, fungi and commodity crops). However, in some cases, geographical proximity to biomass suppliers is highly beneficial. In one observed case, the company is situated between a university town and a rural area. 

Furthermore, case studies revealed many examples of ‘too little’ geographical proximity, where actors were compelled to bring skills, resources and markets closer to each other in order to succeed in furthering their ventures. However, other cases also revealed examples of ‘too much’ proximity, leading to a myopic, inward-looking environment where cross-fertilisation is rare and inefficient. For example, some companies and scientific networks focusing on a specific protein food category or industrial process may largely work alone and ignore the opportunities for cross-learning. In the construction sector, innovation via the use of wood materials was limited during the century-long moratorium on tall wooden buildings, thus enabling the consolidation of concrete and steel construction systems and value chains. Strong cognitive and organisational proximity produced a lock-in situation making it extremely difficult for newcomers offering disruptive alternatives to enter the market. The extent of this difficulty led ‘disruptors’, at least in the early stages of development, to circumvent existing actors and value chains and create parallel avenues to enable market penetration.

Cognitive proximity, which is more associated with like-minded people with common understanding and ideas and less with physical distance, is therefore key to enabling technological innovation to emerge, evolve and spread. The important condition here is that these actors are willing and able to stay in close communication, share knowledge and advance science, even if dispersed across large geographical distances. One example is the partnership between Polarbröd and the fungi company Mycorena, located in the north and south-western part of Sweden, respectively. Drawing on the potential to close the resource loop and encourage circular economy practices, the two companies work together to minimise waste by feeding waste streams to fungi despite being located far apart. However, capitalising on fungi is currently halted by the definition of waste in EU regulations. This is a good example of co-operation toward a common goal but also demonstrates the potential of geographical proximity. 

While the industrial legacy rooted in very specific places has played an important role in the wood construction case, it is also clear that intensive cross-fertilisation between companies working with the development of new building systems and products has taken place mostly across the Nordic countries and central Europe, (although also more recently in the Baltic countries, North America and to a much lesser extent, globally). The substantial work carried out in Germany and Austria in developing mass timber products proved useful for companies and scientists in the Nordic countries, where the forestry industry and wood construction markets are highly developed. It is interesting to note that while industrial development requires close geographical proximity amongst different types of actors (i.e., private, public and academia), technological breakthroughs, independent of industrial development, are less determined by place and more by like-minded companies and professionals learning from each other across borders and industrial sectors. Moreover, significant knowledge exchange has occurred between municipalities, e.g., across Swedish regions that are particularly keen to boost sustainable construction, including in more urban regions where the forestry industry has less or no direct presence. Here, cognitive proximity precedes geographical proximity, although the national level remains an important dimension, as municipalities across Sweden operate under the same national rules.

Excessive cognitive proximity may also limit the opportunity for alternative or hybrid options (e.g., construction systems and materials), a situation which undercuts systems innovation. Many argue that the protein debate is largely one-sided and in the wood construction conversation, some actors have been described as ‘wood Talibans’, a title inferring a certain degree of partiality. 

The case studies have also revealed evidence of the role of organisational proximity, understood as the extent to which relations are shared in an organisational arrangement within and between organisations (Boschma, 2005). One obvious example is the substantial extent to which the Skellefteå municipality administration can drive action by owning its own energy company, by co-ordinating the multi-actor Wood Innovation Cluster and through access to funding via the investment bank Svenska Kommuninvest, which is jointly owned by a large number of Swedish municipalities. This demonstrates strong relations within and between organisations and the power this affords the municipality to exert influence on other actors. There is also evidence of organisational proximity emerging from the recently established consortium following the Vinnova call for ‘recipes for the future’, in which the project consortium has been tasked with establishing an innovation platform for where actors can connect. 

Organisational proximity is also, and often, closely related to social proximity, which refers to the extent to which economic structures are embedded in social contexts (Boschma, 2005). Relations between agents at the very lowest level (microlevel) are important here, as they bring trust, the “Nordic gold”, to the table. At local level and particularly in regions with small populations, the ‘human factor’ is of utmost relevance, as individuals representing different organisations know or are at least acquainted with each other and have built trust relationships that can make supporting new innovative action more efficient and faster. Skellefteå municipality and the broader region is a clear example of this phenomenon where the (social) distance between stakeholders is said to be ‘one phone call’ away. Interestingly, cultural proximities, including values and ethics – as we have seen in the protein case – do not belong in this category but rather to the category of institutional proximity. Also of note here is the role of agents and trust in the development of new products and services in the protein space: innovation in start-ups and scale-ups is dependent on social proximity to enable the exchange of tacit knowledge (Boschma, 2005). This is closely connected to the intersection of place, time and identity. 

Finally, the fifth dimension of proximity is Institutional proximity: a general condition relevant, in one form or another, to all the observed processes. Institutional proximity concerns the ways relationships between actors and innovation are determined by the institutional environment. 

There is a clear connection to the national innovation system in the development of the wood construction sector, where the interlinkages between regulation, national policy and R&D programmes with existing companies throughout Sweden, as well as industry organisations, have all played a significant role in boosting innovation and market development. The more sporadic nature of the same development in Finland has been one of the main barriers to up-scaling the wood construction industry there. Although much of the necessary knowledge came from abroad, and the fact that these developments did not happen in isolation, the role of national innovation systems has been nonetheless crucial. In the protein case, however, the presence of national innovation systems is much less apparent and developments in different industrial and technological strands have emerged more sporadically. Instead, we notice a close interlinkage between scientists and companies that focus on these developments, e.g. cell agriculture or protein alternatives internationally, which have not been particularly determined by national regulatory frameworks, with the exception of certain market creations where first Singapore and later the USA have granted permits to sell cultured meat in the market. However, institutional arrangements do play a role; hence the role of geography, particularly in the form of R&D funding programmes geared towards food systems change. 

More generally, Nordic countries are well placed to advance the development of protein systems, with well-functioning innovation systems, a high ranking on human development indices, more rigorous regulations and governments recognising the importance of working towards a more sustainable future. These well-functioning and dynamic institutional environments, in combination with a society that shows a high degree of openness towards innovation, enable businesses and scientists not only to lead technological innovation processes but also gives them the confidence to pilot and launch them. Institutional arrangements at local and regional levels can be even more impactful in supporting or halting innovation. This is more clearly the case in wood construction, where municipal strategies, local spatial planning systems and tools, as well as the closer relationships between key actors, determine the capacity to pool resources and drive change. 

A final relevant theme to this analysis is transition: more specifically, the green transition. A fundamental presumption that we need to take into account is that we are studying the role of agency in green innovation in a policy context that intends to produce profound transformations in our economic and social systems. Not only innovation policies but also the overarching policy frameworks at EU level, which are then transposed to national and sub-national levels, are all designed with the explicit intent of changing the status quo. Moreover, the changes are not just any changes but are specifically aimed at addressing major societal challenges, including the environmental crisis. This context is a key starting point for this analysis, as studying innovation under a policy context-oriented towards maintaining the status quo would obviously produce entirely different results.

Transitions, or transformational innovations, require changes in the interlinked social and technical systems, understood as ‘socio-technical transitions’. Transformational policy activates different agents to drive innovation, be it technological, social, entrepreneurial or institutional innovations and be they incremental or radical, intentional or unintentional. The sum of all these actions will produce systems change. 

Geels’ (2002) argument that technological transformation results from a process of technological substitution is evidenced in the empirical case studies, although only partially. Literature on sociotechnical transitions emphasises the inherent challenges in these processes, often marked by resistance from influential entities in politics, business and social structures. Overcoming the significant inertia within sociotechnical systems demands more than just technological innovations or alterations in consumer behaviour: it calls for far-reaching and interlinked innovative actions. For this reason, sociotechnical transitions may follow diverse pathways, with the complete system transformation being just one option. It is common for transitions to commence by enhancing or expanding existing systems rather than dismantling them outright (Geels 2002; Geels 2012; Araujo 2014; Geels et al. 2018; Sorrell 2018).

The sectors studied in our case studies follow this pattern whereby new technologies, processes, products and behaviours lead to diversification and incremental change rather than full regime change. Many innovations are also not intended to replace the function of existing products but to transform the processes needed to achieve the end results. For instance, an apartment building or a sausage will fulfil their intended purpose, no matter how they are produced. Yet major change can be achieved through new building processes such as modular technologies or by means of Mass Timber structures in the case of construction, or the feed used in animal farms, animal welfare standards, crop cultivation technologies, fertilisers and pesticides used in the case of meat production. The resulting differences are significant in terms of environmental footprint, efficiency, organisation of labour, costs, etc. 

There are a number of innovations that can potentially lead to radical change, such as cultivated meat or precision agriculture. However, we have not found evidence that these can completely substitute existing products and practices, but nonetheless may lead to a more diversified market and set of practices.

When it comes to institutional structures, the green transition has required substantial transformations in governance structures, particularly the role of public authorities in supporting sectoral development, organisational change, generating new market conditions and assuming a leading role in path creation. The same can be said of social innovation, as it is primarily societal awareness that has triggered change in the political and market spheres. Yet, our case studies show that awareness of environmental issues does not necessarily lead to consumer behaviour that can be considered sustainable. For example, when it comes to consumer choice in selecting an apartment, the decision is mostly guided by convenience, location and price rather than the construction materials or energy efficiency of the building.