Expert knowledge and social innovation: analysing policy debates in Japan

Abstract

Expert knowledge is considered fundamental in the policymaking process, especially when considering multidimensional phenomena such as social innovation. However, the positioning of experts in the policy debate still deserves a thorough investigation. By using Japan as case study and applying Discourse Network Analysis to the debates occurred in the Japanese National Diet (2017–2021), this study investigates the presence of discourse factions and the positioning of experts in these factions. Results show that factions exist but there is not a strong polarisation in the arena; moreover, experts do not just support neoliberal policies, they also encourage bottom-up initiatives.

Keywords:

• Expert knowledge
• social innovation
• Discourse Network Analysis
• Japan

Introduction

Expert knowledge is the type of knowledge shared by skilful individuals in a specific area or discipline (Sandberg, Persson, and Garpenby 2019, 431), and it has always been considered fundamental for policy decisions (Boswell 2009; Grundmann 2007; Radaelli 1995). The idea that scientific research and evidence can be the foundation of rational choices has become stronger in the last decades (Sandberg, Persson, and Garpenby 2019); at the same time, expert knowledge can be used for strengthening pre-determined ideas on a certain topic (e.g. Radaelli 1995). Moreover, expert knowledge can be politically used to dismantling other people’s opinions, especially when the political debate is focussing on resource allocation (Boswell 2009).

In recent years, social innovation has gained importance amongst Asian, American, and European policymakers (Kerlin 2009; The Hope Institute 2017), because of the re-design of the concept of welfare state and the role of the public sector in the production and offer of social services (Massey and Johnston-Miller 2016; Notarnicola, Berloto, and Perobelli 2020). According to Howaldt and Kopp (2012), expert knowledge plays a key role in the definition and implementation of social innovation policies. At the same time, this view is contested by other authors who claim that social innovation mainly emerges from bottom-up approaches implemented by the civil society (Adam and Hess 2010; Montgomery 2016. Scholars are divided between those highlighting the importance of a top-down decision-making system based on expert knowledge (e.g. Benneworth and Cunha 2015; Nordberg, Mariussen, and Virkkala 2020) and those considering bottom-up initiatives as the main drivers for social innovation (e.g. Membretti 2007). This has led to the identification of expert knowledge as the foundation of a neoliberal/technocratic view (Montgomery 2016); however, the role of experts in the policymaking process is not clear, and there is a lack of research focussing on their positioning in this process. Therefore, the objective of this work is to understand experts’ positioning in the political arena and how expert knowledge is used in support of specific initiatives. In particular, we aim to address the following research question: what is the positioning of experts in the social innovation policymaking process?

By using Japan as case study, this paper focuses on the social innovation debates in the Japanese National Diet from 2017 to 2021. Japan is facing relevant socio-demographic challenges, such as ageing population and population loss (Onitsuka 2019; Świtek 2014). These challenges have fostered the implementation of policy initiatives supporting social innovation (Aoo, Abe, and Kano 2019; Moore et al. 2012). Discourse Network Analysis (Leifeld 2017, 2020) is used to investigate the discourse factions in the Japanese National Diet, focussing on the positioning of experts. The next section illustrates the literature on expert knowledge and social innovation policymaking. Section three provides a brief overview of the policy initiatives related to social innovation in Japan. The fourth section is dedicated to describing the methodological approach—what is Discourse Network Analysis and how discourse network data were collected and analysed. Section five presents and discusses the main results of the analysis, while the last section concludes and proposes a research agenda for further studies on the topic.

Theoretical Framework: Expert Knowledge and Social Innovation Policymaking

The use of expert knowledge has been investigated in depth in the policy literature. While the first studies suggested that this type of knowledge can support policymakers in addressing political issues (e.g. Weiss 1979), following studies highlighted that it can also be used for different purposes—such as legitimising pre-determined positions (Radaelli 1995). Sabatier (1988) and Christensen (2021) pointed out that research-informed advice can support the adoption of rational decisions (Christensen 2021); at the same time, experts can be motivated by a system of beliefs similarly to all other stakeholders. Drawing on the work of Boswell (2009), Sandberg, Persson, and Garpenby (2019) suggested that the very nature of political debates is key for the use of expert knowledge; when there are conflicts between policymakers for the allocation of resources, knowledge is likely to be used for re-enforcing personal positions rather than developing evidence-informed policies.

Multiple perspectives and opinions emerge in the policymaking arena, and expert knowledge can be used for supporting one position rather than another. In the last years, social innovation has gained the attention of policymakers (Howaldt and Kopp 2012; Marques, Morgan, and Richardson 2018; Ney 2014), and divisions on how to implement social innovation policies have emerged among these actors (Montgomery 2016). The same definition of social innovation is still a matter of debate in the scientific community (Edwards-Schachter and Wallace 2017; Grilo and Moreira 2022; van der Have and Rubalcaba 2016). According to Mulgan (2006, 146), social innovation is a broad concept that refers to all ‘innovative activities and services that are motivated by the goal of meeting a social need’. Pol and Ville (2009) deepened this concept, stressing the importance of improving quality and quantity of life: for them, social innovation is something that leads to specific improvements in education and environmental quality, and increase life expectancy. In their view, not all social innovations are desirable: only those that are meeting the above targets. However, still in recent years social innovation continued to be perceived as a buzzword or quasi concept. Literature reviews, empirical and theoretical studies have tried to clarify what are the main characteristics of social innovation: the importance of civil society actors; the existence of cross-collaborations between civil society, public and private sectors; the focus on changing social practices; different modes of providing good and services; achieving social targets; addressing social and environmental issues, in addition to the economic ones (Edwards-Schachter and Wallace 2017; Jenson 2015; Ziegler 2017).

A key element, highlighted in several studies, is the relationship between public and private stakeholders. Despite policymakers perceive social innovation as a magic concept (Bragaglia 2021), its vagueness create coordination issues between members of the civil society, politicians, and representatives of the private sector—when it comes to implement strategic decisions for supporting social innovation. According to Membretti (2007) and Adam and Hess (2010), local initiatives from citizens and communities of individuals are at the core of the innovation process, and they do not necessarily need inputs from national or international experts, neither their validation. This view is in favour of a bottom-up approach where initiatives are led by individuals and organisations with a local bond. Others (Benneworth and Cunha 2015; Howaldt and Kopp 2012; Nordberg, Mariussen, and Virkkala 2020) support a view where expert knowledge is considered a prominent element in this process: recently, Leitheiser and Follmann (2020) demonstrated that urban policymakers consider expert knowledge more important than citizen knowledge, and technical advice from scientists and experts is generally included in the policy initiatives implemented at local level. Nicholls and Murdock (2012) pointed out that social innovation is a political construct which is strongly influenced by social and individual expectations. Indeed, social innovation has often been used to justify policy choices supported by specific political views. Moulaert, MacCallum, and Hillier (2013) described how neoliberal policies are sometimes presented as social innovation policies; together with Jessop et al. (2013), they criticised the approach of governments that delegate the delivery of social services to the market and call it social innovation. Moreover, Montgomery (2016) argued that there are two schools of social innovation: one supporting a neoliberal (or technocratic) view, where the role of experts is predominant, while the other empowers the participation of communities in the decision-making process. In recent year, the first view has prevailed; mainly in European and US policy debates (Fougère, Segercrantz, and H. Seeck 2017), as a consequence of the 2008 financial crisis, but also in Asian countries such as Japan (Aoo 2018).

Social innovation can be applied to multiple activities and sectors, and according to Westley (2013, cited in Milley et al. 2018) it is rooted into at least six different schools of thought: social entrepreneurship, innovation theory, institutional entrepreneurship, sociotechnical transitions, resilience and social-ecological transformation, and social economy. However, while the outcome of social innovation has been widely explored in the literature, the policymaking process has not (Borzaga and Bodini 2014; Nicholls and Edmiston 2018). This is because of the complexity of the topic, the lack of agreement on the meaning of social innovation, and the different themes that can be covered by this phenomenon. Recent studies (Galego et al. 2022; Macke et al. 2018; Phillips et al. 2015; van der Have and Rubalcaba 2016), have pointed out that social innovation focuses on the following themes: intra-organizational well-being, business ethics, and corporate social responsibility; the importance of the local aspect, with a specific focus on community development; participation, inclusion, and empowerment of citizens; sustainability of climate, environment, and health; cross-sector alliances and networks; social entrepreneurship; active role of institutions.

Galego et al. (2022, 5) argued that social innovation initiatives are characterised by collective actions from citizens and organisations (to satisfy the needs of a local community), social relationships (i.e. networks), and ‘opportunities for public participation in decision-making’. In their review of the literature on social entrepreneurship, Macke et al. (2018) stressed the importance of the relationship between social entrepreneurship and the concept of social innovation. At the same time, they found that this concept is associated to the idea of networking, because networks enable social entrepreneurs to share ideas, advice, and building trust in their community. Similarly, Phillips et al. (2015) found that the two concepts of social innovation and social entrepreneurship are strongly interconnected, and social networks are fundamental for determining the success of social innovation initiatives. Spanning boundaries and moving across sectors lead to the acquisition of new knowledge and experience, which are drivers for innovation. Starting from such assumptions, these authors developed the idea of social innovation systems as communities made by multiple stakeholders, whose aim is to address social issues while being innovative from a social perspective. Moreover, their literature review emphasised the importance of institutions in supporting social organisations for addressing social needs. Their importance has also been discussed by Ndou and Schiuma (2020), which included this theme as well as citizens participation and networking among the key themes emerging from the literature. Finally, van der Have and Rubalcaba (2016) detected four main thematic clusters in the social innovation literature: community psychology, which includes eco- and natural resource management; creativity, which includes intrapreneurship, business ethics, and corporate social responsibility; social and societal challenges, which includes sustainability of climate, environment and health; local development, which includes the role of institutions and the participation of citizens to the decision-making process. A visual representation of all the themes identified in the above studies is presented in Figure 1.

Figure 1. Social innovation themes. Source: created by the authors.

These themes are not mutually exclusive; however, policymakers often concentrate their energies on strategic actions linked to just one or a few of these themes. In this respect, while expert knowledge is considered relevant in policymaking, scholars do not have a shared view for describing the role of experts in the social innovation policymaking process—and how their view is supportive of certain actions rather than others.

Social Innovation in Japan

The increased interest towards social innovation in the last decades has been influenced by two main factors: (a) the reduced trust in the for-profit enterprise model, due to the recent global economic crises; (b) the retreating of the public sector, which has led to the transition from a welfare state to a welfare mix system (Notarnicola, Berloto, and Perobelli 2020; Phillips et al. 2015). These factors have contributed to address the public debate in several countries, paving the way to different policy strategies (Boelman and Heales 2015).

Japan is a country where the public interest on social innovation is rising. Because of its ageing population, socio-demographic and economic challenges are putting pressure on public finances (National Institute of Population and Social Security Research 2020). This has pushed the Government to search for novel approaches for supporting social change, similarly to other developed countries. At the same time, natural disasters—such as the great earthquake in 2011 and its consequences on the country’s nuclear energy production systems (Kenens et al. 2020)—have influenced the behaviour of citizens and organisations and their participation to the production, management, and offer of public goods. This has been a radical change compared to the past, when Japanese citizens used to delegate the design and management of social activities to public authorities—and therefore showed little interest in social welfare issues (Tanimoto 2008). Nowadays, people are more concerned about the issues faced by the Japanese society, and they show a higher propensity to discuss about solutions and strategies for addressing these issues. Such situation has a direct effect on the national political debate, making social innovation more appealing to the policymakers (Bragaglia 2021).

Historically, the main actors of social innovation in Japan have been non-profit organisations (Fujisawa et al. 2015) and local and national authorities (IDB. 2016). However, in the last two decades other stakeholders have become particularly important in addressing social problems. According to two reports from METI (2008) and the Japan Finance Corporation (2014), 20% of social businesses are registered as for-profit organisations; while this is a minor quota compared to non-profit organisations, it is still the second largest group of organisations operating in practice as social businesses, followed by sole proprietors (11%) and unions (7%). Most of the social businesses have a small or even micro dimension, and they depend on donations and grants for supporting their activities. The characteristics of social entrepreneurship in Japan has been thoroughly examined by Kaneko (2013). In his work, he used an historical perspective for dividing social entrepreneurial activities into three generations: the first generation refers to the neighbourhood associations and cooperatives established in the first half of the twentieth century; the second generation includes community businesses and social businesses established at the end of the 1970s, in response to the inaction of the national government to address local issues due to natural disasters; and the third generation is the one consisting of non-profit and for-profit organisations emerged at the beginning of the 2000s, the only generation which focuses on all dimensions of social entrepreneurship—social mission, market orientation, and social innovation. Hence, according to this author, it is at the beginning of the new century, with the so-called last generation of social entrepreneurial activities, that social innovation has become reality in the Japanese context.

In addition to social entrepreneurship, Japanese universities and other higher education organisations are playing a relevant role in enabling social changes. In his empirical study on the development of social innovation, Tanimoto (2012) found that universities are central actors in social innovation clusters, i.e. clusters including multiple stakeholders (such as social enterprises, research institutions, support organisations, and funding agencies) that establish cooperative relationships for creating and offering innovative social solutions to the territory. At the same time, Yonezawa (2018) argued that Japanese higher education institutions should support national innovation policies and enabling social changes—stressing again the importance of connections with industry players and the national government. Finally, citizens’ associations are consistently involved in social change projects, especially when focussing on themes related to sustainability: in different Japanese areas, it has been possible to observe—in the last 20 years—an increasing number of associations focussing on the development of local projects, for example establishing a renewable energy production system (e.g. Maruyama, Nishikido, and Iida 2007) or designing smart cities (e.g. Chatfield and Reddick 2016).

Discourse Network Analysis

Our analysis concentrates on the Japanese legislative body: the National Diet. Legislative bodies are fundamental in the policymaking process (Gerston 2014), because they enable politicians to discuss policy solutions to political problems. Multiple perspectives and opinions emerge in these debates, and they are directly linked to the outcome of the policy process. At the same time, legislative bodies give experts the opportunity to share their knowledge and express opinions, which can be used as the foundation of evidence-based policymaking (Christensen 2021; Geddes 2018).

Several methods and analytical techniques have been developed throughout the years for analysing policy debates and their impact on the policymaking process (e.g. Gastil 1992; Yanow 2007). Most of these techniques fall under the umbrella of discourse analysis, which is used for investigating form and content of discourses, and whose application in policy studies allows to understand what policy issues are important in the political debate and what are not (Pautz 2018). In the last years, Discourse Network Analysis has emerged as a promising analytical tool for concentrating on the content of the discourses, because of its capacity to analyse both actors and concepts according to a multidimensional approach (Leifeld and Haunss 2012). Discourse Network Analysis is a method developed by Philp Leifeld which integrates content analysis and Social Network Analysis (Leifeld 2017, 2020). It allows to operationalise policy debates and create discourse networks from text data, by analysing all claims expressed by a set of actors on a specific topic or a group of topics. Claims can be found in a text when actors reveal their preferences on policy-related aspects (Fisher and Leifeld 2019), and by associating actors and concepts via individuals claims it is possible to define five types of discourse networks: affiliation networks, actor congruence networks, conflict networks, concept congruence networks, and dynamic discourse networks (Leifeld and Haunss 2012). An affiliation network is the basic form of discourse network that can be created and visualised using Discourse Network Analysis, and it is the type of network analysed in this study. Affiliation networks are made by a set of actors A = {a₁, a₂, a₃, …a_n} and a set of concepts C = {c₁, c₂, c₃, …c_n}; generally, actors have different positions and ideas, and therefore they can agree or disagree with a concept. This reflects in the creation of binary ties between actors and concepts, in which a positive sign is associated to agreements and a negative sign is associated to disagreements. By using this information, it is possible to create bipartite graphs where ties between actors and concepts exist, and these ties assume different values according to the presence of agreements or disagreements. As stated by Leifeld and Haunss (2012, 391), ‘applied to the empirical data, an actor is connected to a concept in the affiliation network if she or he makes a claim in which they use this concept’; an example of affiliation network is illustrated in Figure 2, where it is possible to see that actors can express claims about more than one concept, and their agreement (or disagreement) with these concepts, identified by the line dash, make them aligned or conflicting with other actors in the arena.

Figure 2. Example of affiliation network. Source: created by the authors.

The key steps in Discourse Network Analysis are: (a) collecting data; (b) coding of claims. This method requires text data reporting claims expressed by certain actors on specific topics, and such data can be found in different sources: newspapers and media materials, policy documents, minutes, and transcribed interviews. Since the focus of this study is on legislative bodies, the analysis uses verbatim reports produced by the Japanese National Diet—国会 (House of Representatives—衆議院 and House of Councillors—参議院) from 2017 to 2021¹, the period following the 48th general national elections. This approach has been used in previous studies using Discourse Network Analysis (e.g. Fisher, Waggle, and Leifeld 2013; Fisher, Leifeld, and Iwaki 2013; Ghinoi and Steiner 2020); in particular, Fisher et al. (Fisher, Waggle, and Leifeld 2013; Fisher, Leifeld, and Iwaki 2013) have highlighted the importance of congressional hearings when policymakers are interested in obtaining information from recognised experts in specific fields—which is a fundamental aspect characterising this research. Since there are several speeches included in each report, and most of them are not focussing on social innovation, it was developed a search strategy for locating those in which political actors and experts expressed claims on social innovation. First, text parts dedicated to social innovation were extrapolated by using the keyword social innovation in English, in kanji (社会変革), and in katakana (ソーシャル・イノベーション). Once identified the text parts, all claims were coded according to the main themes related to social innovation that emerged from the literature review (see Figure 1): intra-organizational well-being, business ethics, and corporate social responsibility; the importance of the local aspect, with a specific focus on community development; participation, inclusion, and empowerment of citizens; sustainability of climate, environment, and health; cross-sector alliances and networks; social entrepreneurship; active role of institutions. During the coding process, it was checked if claims were coherent with the main themes and if the social innovation keywords were actually used because the main speech was about social innovation—and not other topics. The second author, who is more familiar with the empirical context, has been the principal coder, while the first author double-checked only those cases that were considered uncertain; while this is clearly a strategy that requires more efforts from a single researcher, it helps to reduce concern over intercoder reliability, as discussed by Malkamäki et al. (2021).

As illustrated in Figure 3, for each report it is possible to identify multiple speeches from different actors, and in the same speech an actor might be able to make claims on different themes. Since Discourse Network Analysis enables to create an affiliation network where ties between actors and themes represent if actors agree or disagree on those themes, in the coding process it was controlled for this aspect; however, it was found that no one expressed disagreement on a specific theme. Politicians and experts made claims for supporting specific strategies and solutions, rather than discrediting other people’s opinions. Therefore, the final bipartite graph is an affiliation network where policymakers and experts (i.e. network actors) are linked to social innovation themes (i.e. concepts) because they agreed on such themes². Once created such network, UCINET 6 was used for the quantitative analysis of the actors’ network positioning (Borgatti, Everett, and Freeman 2002). In particular, the quantitative analysis concentrates on detecting factions by using blockmodeling. Blockmodels are used for grouping network actors according to their relational patters: the more similar the connections shown by two actors, the more similar they are. This enables to define groups or categories of equivalent actors. In bipartite networks, it is possible to use blockmodeling for detecting factions according to the preferential attachment of actors towards themes: factions must have a very high intra-group density (the number of active connections over the total number of potential connections), but the density of the connections between groups must be very low.

Figure 3. Text data structure: an example. Source: created by the authors.

Results

The data collection process resulted in detecting 129 speeches made by Japanese politicians and experts on social innovation between 2017 and 2021. It has been possible to identify 64 individuals belonging to the following groups: the Liberal Democratic Party/自由民主党 (28 individuals); experts (16 individuals); the Constitutional Democratic Party of Japan/立憲民主党 (9 individuals); the Japan Innovation Party/日本維新の会 (4 individuals); the Democratic Party for the People/国民民主党 (2 individuals); independent politicians (2 individuals); Komeito/公明党 (2 individuals); and the Japanese Communist Party/日本共産党 (1 individual). Since 1955, the Liberal Democratic Party has almost always been the ruling party in Japan, alone or in coalition with other parties such as Komeito. The Liberal Democratic Party is a traditional conservative centre-right party, whose socio-economic and political interests focus on bureaucratic simplification, tax reduction, cooperation with the United States, and it is characterised by moderate elements of nationalism and protectionism. Komeito is a political party linked with the Buddhist movement Soka Gakkai: it is positioned in the centre-right political spectrum, and its main aim is enabling social and economic freedom. These two parties diverge on different political topics (Liff and Maeda 2019), especially when looking at their view on the Japanese welfare state: while in the past the Liberal Democratic Party tried to reconfigure the national welfare system adopting a marketisation approach, Komeito has always emphasised its importance and opposed to the liberalisation of (most of the) social services. The Constitutional Democratic Party of Japan is currently the main opposition party in Japan, and in 2020 it merged with the Democratic Party for the People. They are considered centrist parties, adopting liberal approaches in economy, supporting pacifism and sustainable development, but with some conservative elements in internal and foreign policy. The Japan Innovation Party is predominantly libertarian, and it advocates deregulation and a reduction of the Government intervention in society. This is rather the opposite view of the Japanese Communist Party, which in contrast is against large corporations and the military alliance with the United States, and it supports environmentalism and the strengthening of the national welfare system.

Experts had different roles and belonged to different sectors and organisations: the database includes academics from the University of Nagoya, the University of Kyoto, the Yokohama National University, and Tohoku University; civil servants from the Ministry of Education, Culture, Sports, Science and Technology, the Ministry of the Environment, the Ministry of Economy, Trade and Industry, and the Cabinet Office; scientists working for the Intergovernmental Panel on Climate Change; mayors of Japanese cities; directors of research institutes such as the National Institute of Physical and Chemical Research. Experts were invited to share their opinions as representatives of the scientific community and the institutions. Overall, it was possible to map 177 claims linked to the seven themes characterising social innovation. However, some of these claims were just repeated by the same actor in different periods on time; once removed all duplicates, 108 unique claims were identified.

Table 1 shows the number of claims by group; themes are numbered according to their order in Figure 1. The role of institutions (theme 7) was the most discussed theme in the Japanese National Diet, followed by social entrepreneurship (theme 6) and sustainability (theme 4). The other themes were far less discussed by Japanese policymakers and experts (Table 2).

Table 1. Number of claims by group.

Group	Theme 1	Theme 2	Theme 3	Theme 4	Theme 5	Theme 6	Theme 7	Total
Liberal Democratic Party	5	4	6	6	3	13	14	51
Experts	–	–	2	5	4	6	6	23
Constitutional Democratic Party of Japan	1	1	1	4	–	2	5	14
Japan Innovation Party	–	–	–	1	–	3	1	5
Democratic Party for the People	–	–	–	1	–	–	2	3
Independent politicians	1	–	1	2	–	–	2	6
Komeito	–	–	–	1	–	2	1	4
Japanese Communist Party	–	1	–	1	–	–	–	2
Total	7	6	10	21	7	26	31	108

Legend.

Theme 1: Intra-organizational well-being, business ethics, and social responsibility

Theme 2: Focus on the local aspect/community development

Theme 3: Participation and inclusion of citizens

Theme 4: Sustainability

Theme 5: Cross-sector alliances and networks

Theme 6: Social entrepreneurship

Theme 7: Role of institutions.

Table 2. Graph legend.

Group	Number of actors	Shape of the node	Colour of the node	Political coalition
Democratic Party for the People	2	Circle	Yellow	Koike’s coalition
Expert	16	Square	Grey	–
Independent politician	2	Up Triangle	White	–
Japanese Communist Party	1	Box	Red	Pacifist coalition
Komeito	2	Down Triangle	Pink	Ruling coalition
Liberal Democratic Party	28	Circle-in-box	Dark Green	Ruling coalition
Nippon Ishin/Japan Innovation Party	4	Diamond	Light Green	Koike’s coalition
The Constitutional Democratic Party of Japan	9	Thing	Blue	Pacifist coalition

The affiliation network emerging from the policy debates in the Japanese National Diet is represented in Figure 4. Almost half of the individuals who made a claim were affiliated to the Liberal Democratic Party (the main party of the ruling coalition), and they managed to cover all themes. Overall, most of the themes were discussed by a mix of individuals from different parties and experts, but one theme (theme 5: cross-sector alliances and networks) was covered by members of the Liberal Democratic Party and experts only. Experts stressed the importance of developing strong collaborations between universities, research sectors, and business companies, as reported by ID8:

Figure 4. Affiliation network (2017-2021). Source: created by the authors.

…in the Ministry of Education, Culture, Sports, Science and Technology, the Centre for Innovation (COI) is active since 2013 in order to realize reforms that seek to develop social innovation through full-scale industry-academia collaborations between universities and private companies…³.

Members of the Liberal Democratic Party confirmed the key role of such collaborations, as highlighted by ID3 in her speech on start-ups from universities, as a mean for creating new business ventures and fostering social change:

…even if a system that allows universities to invest in start-ups is established, if this system is not utilized and the environment in which star-ups are being created is not maintained, it becomes meaningless. In other words, it is necessary to push universities to start new start-ups based on the results of their research, and provide them managerial know-how and financial support so that they are able to grow. From now on, it is important to promote the creation of university-launched start-ups and create an environment in which research results can lead to social innovation…⁴.

One aspect that influenced the political debate was the Covid19 pandemic. Most of the claims on social entrepreneurship (theme 6) were expressed in 2020 and 2021, and they focussed on the social changes imposed by the pandemic. In 2020, a member of the Liberal Democratic Party (ID54) claimed that:

…there are areas in which the Japanese economy and society are lagging, such as digitization […]. I think this is an opportunity to promote social change […] While doing remote work, I think that some people have thought about the possibility of having children and living and working in regions with a lower risk of infection…⁵.

Similar opinions were shared by an expert from the Cabinet Office (ID13), which argued that the pandemic could trigger initiatives supporting social entrepreneurship:

…by using this crisis as an opportunity for social change, we would like to promote reforms such as digitization…⁶.

This is why ID13 and ID54 are both linked to the node representing theme 6 in Figure 4—which indicates that these actors have expressed a similar position in favour of social entrepreneurship.

Findings and Discussion

Social entrepreneurship and sustainability were among the main themes discussed in the Japanese National Diet, because of their increasing importance in the political and academic debate in the last years (Macke et al. 2018; Marques, Morgan, and Richardson 2018; Zhang and Swanson 2014). Other themes were covered as well, and in none of them we found elements of disagreement between actors. This indicates the presence of pro-innovation bias (Godin and Vinck 2017), an attitude that has been observed in other contexts and has contributed to the emergence of stereotypes around the idea of social innovation (Pel and Kemp 2020). When discussing about sustainable practices and social entrepreneurship, there is often an underlined positive representation of these phenomena: however, failures and unexpected consequences are always possible. In the Japanese context, institutional actors tend to assume that all strategies for supporting social innovation are acceptable: some actors concentrate on certain themes rather than others, but no one express a strong disagreement on proposals coming from other individuals—even if they belong to the opponent political party. Another element explaining this finding relates to the convergence (and alignment) of the Japanese political parties. As described by Kweon and Suzuki (2022), in recent years even the Liberal Democratic Party—notably supportive of market-oriented approaches—has expanded social benefits for specific social groups, to cope with population ageing. Hence, we observe more similarities across political parties than in the past, because: (a) they perceive social innovation as a positive phenomenon per se, and (b) they are starting to share similar policy visions to address social problems.

Experts shared their view about all themes apart from theme 1 (intra-organizational well-being, business ethics, and social responsibility), and six of them made claims on multiple themes. By using blockmodels (Faust and Wasserman 1992), it is possible to understand to which factions experts belong to—i.e. to which themes they are mainly interested and if they share this interest with other groups of politicians. The results of the analysis are presented in Table 3, while the goodness of fit measures are illustrated in Table 4. While there is not a clear polarisation between actors according to their political affiliation, it is possible to observe a preponderant presence of experts when the focus is on participation and inclusion of citizens (theme 3), cross-sector alliances and networks (theme 5), and social entrepreneurship (theme 6). The only political party whose majority of members made several claims about these themes was the Nippon Ishin/Japan Innovation Party. This is consistent with the political manifesto of this party, which promotes entrepreneurship and support to businesses through tax exemptions and fiscal cuts. On the other hand, the members of the ruling coalition (Liberal Democratic Party and Komeito) appear almost equally divided between the two factions, as if they wanted to cover all possible aspects of social innovation by adopting a broader approach in the policy debate. Therefore, there is not a strong polarisation in the Japanese political arena, and experts concentrate on those themes related to the idea of a bottom-up approach to social innovation and a different (social) view on entrepreneurship.

Table 3. Blockmodeling: faction analysis.

Faction 1	Faction 2
Democratic Party for the People (n = 2)	–
Expert (n = 6)	Expert (n = 10)
Independent politician (n = 2)	–
Japanese Communist Party (n = 1)	–
Komeito (n = 1)	Komeito (n = 1)
Liberal Democratic Party (n = 15)	Liberal Democratic Party (n = 12)
Nippon Ishin/Japan Innovation Party (n = 1)	Nippon Ishin/Japan Innovation Party (n = 3)
The Constitutional Democratic Party of Japan (n = 7)	The Constitutional Democratic Party of Japan (n = 2)
Focus on theme 4 (sustainability) and 7 (role of institutions)	Focus on theme 3 (participation and inclusion of citizens), 5 (cross-sector alliances and networks), and 6 (social entrepreneurship)

Table 4. Faction analysis: goodness of fit.

	Faction 1	Faction 2
Faction 1	0.339	0.107
Faction 2	0.083	0.454
Correlation to ideal: 0.355

This is an element of novelty compared to previous studies (e.g. Montgomery 2016; Moulaert, MacCallum, and Hillier 2013). While it is confirmed that experts are consulted for elaborating informed decisions about social innovation policies, from our analysis it emerges that experts are also supportive of strategies that are characterising the democratic view of social innovation: more engagement of citizens and more networking across organisations. The division neoliberal (or technocratic) versus democratic seems to have a different dimension in this context; experts are used for justifying neoliberal approaches, but at the same time they are promoting different perspectives. The below examples from two debates in the Japanese National Diet can provide a better interpretation of this phenomenon.

Actor ID48 (a bureaucrat from the Ministry of Economy, Trade and Industry) in their speech in April 2021 stressed the importance of empowering citizens and fostering social innovation via knowledge diffusion. This speech was part of a debate around the role of research institutes and universities in boosting economic growth and development; other actors from the Liberal Democratic Party participated to the debate, and they focussed on the review of the Japanese higher education system and the significance of moving towards a digitalised society. While clearly there was a point of contact between the politicians and the expert, the latter highlighted that is necessary to invest in people as main source of transformation—and empowering citizens to strengthen their attitude towards continuous learning and personal development. In other debates, politicians from the Liberal Democratic Party sustained this view asking for extra-allocation of resources for scientific research and education. This demonstrates that members of the ruling party do not concentrate on budgetary cuts only, and they recognise the economic and social returns of investing in education and citizens empowerment.

In the same period, another expert—ID52, a city mayor—made a speech about the successful experience of public-private partnerships in their area. These partnerships saw the participation of businesspeople and regular citizens, and their contribution to the co-creation and implementation phases was key for achieving a positive outcome. In the expert’s view, citizens are engaged when projects are easy to understand and have a direct implication in everyone’s life; to promote social innovation, governments should rely on bottom-up initiatives from the civil society, because citizens can immediately assess the impact of these initiatives.

Therefore, while it is important to recognise that expert knowledge is used in the Japanese political arena for promoting a technocratic view, it is also true that this view is not completely in conflict with the democratic one. A hybridisation is possible (Krauz 2016), and experts may become the vehicles for encouraging (scientifically robust) approaches and policies that have been tested and implemented at local level. However, we also need to recognise that experts’ suggestions could have been influenced by the central government past strategies. Japan has a strong central government, and the budgetary cuts imposed in the last years have incentivized co-creation initiatives led by citizens and non-governmental organisations—sometimes intentionally, sometimes not (Cinar, Simms, and Trott 2022; Normile 2007). Naturally, these initiatives emerged in response to the government’s cuts, and experts could only ascertain their success and introduce them in the political arena as examples of good practices.

Conclusions

While the importance of evidence-based policymaking is highlighted by scholars and policymakers, there is a lack of studies on social innovation examining how expert knowledge is used in policy debates. By focussing on Japan as a case study, this research applies Discourse Network Analysis to detect the discourse factions emerging in the Japanese National Diet between 2017 and 2021, identifying the positioning of experts and political actors and assessing how expert knowledge is used in the policymaking process.

Differently from previous studies (Fisher, Waggle, and Leifeld 2013; Fisher, Leifeld, and Iwaki 2013; Ghinoi and Steiner 2020), it has not been found a strong polarisation between politicians. At the same time, experts are covering almost all the themes, but they mainly concentrate on the following: participation and inclusion of citizens, cross-sector alliances and networks, and social entrepreneurship. Hence, experts do not just concentrate on those themes that are considered relevant for (neoliberal) policymakers: they are also emphasising the opportunities offered by other approaches. This situation can be seen as a first step towards the hybridisation of two different paradigm, in which expert knowledge is not only used as a technocratic leverage for justifying neoliberal policies, but a tool for legitimising the inclusion of the civil society in the policymaking process.

This work makes a twofold contribution to the literature on social innovation. First, it uses Discourse Network Analysis for investigating the role of experts in supporting social innovation policies. This is the first study adopting this method for analysing the social innovation policymaking process, and further applications can concentrate on other cases studies to understand the dynamics of this process in other contexts. More empirical studies are necessary to understand how different societies react to social innovation, what strategies are implemented, and if expert knowledge is used for justifying neoliberal approaches (Montgomery 2016) or a stronger involvement of the civil society. Second, this study contributes to the empirical literature on expert knowledge by shedding light on the (discourse) interactions between experts and politicians. We empirically tested the presence of pro-innovation bias (Godin and Vinck 2017) and on overall support of certain elements characterising social innovation such as networking, social entrepreneurship, and sustainability. Meanwhile, we observed that experts are not supportive of top-down approaches only: they also promote bottom-up initiatives, and this phenomenon can be the natural consequence of past neoliberal policies as well as an hybridisation (Krauz 2016) where expert knowledge is not just the foundation of a technocratic paradigm but a bridge between institutions and the civil society.

This study is not exempt from limitations. First, it has not been possible to identify if experts’ opinions have been able to influence politicians’ opinions, and if this had an impact on the implementation of certain policies. Further research can be dedicated to assessing the presence of social contagion between stakeholders. Second, the seven themes identified via the literature review can be considered as rather general; for example, sustainability can be decomposed into social, environmental, and economic sustainability. Future studies should concentrate on single initiatives that have emerged as particularly important for social innovation and assess policy debates focussing on these specific initiatives—in Japan or even other countries. Another limitation is that we have included only politicians and experts in our analysis. We are aware that other stakeholders from the civil society are contributing to social innovation, and further research may contribute to integrate our findings with those deriving from a review of bottom-up initiatives. Finally, there are some limitations in the method of analysis as well. Coding can be an issue when using Discourse Network Analysis, even when researchers are double checking the results: further research should be dedicated to exploring new tools for adopting advanced text analysis techniques in combination with Discourse Network Analysis.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 All reports can be found here: https://kokkai.ndl.go.jp/#/

2 Actors can also repeat the same claim over the same theme in multiple speeches; however, it was considered only the presence/absence of ties between actors and themes, not the frequency of claims over the same topic.

3 Authors’ translation from the original Japanese text.

4 Authors’ translation from the original Japanese text.

5 Authors’ translation from the original Japanese text.

6 Authors’ translation from the original Japanese text.