Evidence-Based Policy Advice and Decisions in the Nordic Region

In reviewing the research literature on evidence-based policy, very little was found about reference use in policy documentation (see Chaps. 1 and 2). Therefore, it is interesting to conduct research on how various knowledge sources are mobilized in reference networks through policy-making processes. Such sources, as mentioned in the introductory chapter, can be widely defined, including information, ideas, and arguments; well-tested beliefs; and lay, professional, and academic knowledge (Radaelli, 1995).

The POLNET study focuses on the use of knowledge sources referenced within white and green papers. In our study, we found that, in all five countries, documents made explicit references to a variety of knowledge sources. The results of our investigation show that the policy papers reference both governmental documents and draw on other types of documentation published by non-academic authors alongside research-based evidence produced by various research institutions. However, a striking similarity between the five cases is the near absence of academic references from the educational sciences.

As formulated in the Icelandic chapter, academic papers “are thought to be irrelevant or not providing ‘accessible’ knowledge in the evidence base for the policy” (Chap. 6, p. 174). Likewise, as spelled out in the Danish chapter, references to Danish academic research are almost non-existent. Due to international policies that recommend using big data and empirical evidence and the pattern of the relatively few references to academic works in the educational sciences, there seems to be a mismatch between what policy makers and experts consider relevant to cite and what, for example, students at universities and colleges read and discuss to achieve their degrees in the educational sciences. There are also relatively few references to articles printed by renowned, national, and international publishers.

Educational science in Europe and the Nordic countries in particular has, for institutional reasons, been characterized by professional-practical scholarship relevant in, for example, teacher education (Heggen et al., 2010). Educational scholarship has also evolved as an academic research field at universities since the early 1900. Yet, traditions for large-scale empirical research vary between areas of expertise in education as well as between countries. There is, for example, a long-standing tradition for comparative research studies in Sweden and Finland, while Denmark, Iceland, and Norway are latecomers in this respect (Sivesind, 2019).

A surprising observation based on our dataset is the outstanding distinction between the numbers of references cited in each of the cases. Norway is clearly on top (2312 references), followed by Sweden (1421), and Finland (677). The Danish and Icelandic cases have 231 and 203 references, respectively. One obvious explanation for this difference is the selection procedures for collecting documents. In Norway, the team decided to start with two white papers and the green papers on basic education referenced in the white papers and thereafter count all references within this corpus of documents. The same selection procedure was applied in all countries, but the reforms under study did not produce the same number of source documents (i.e., white papers).

In the particular reform process in Norway, the parliament asked first for one white paper on reforming the national curricula and second for a white paper that included recommendations for renewing the assessment system. The minister in education decided to meet this last request by including reform-related themes in a white paper in progress. This national case resulted in a larger number of references compared to the other four cases. Due to these concrete circumstances, we cannot conclude that the large variations in the number of references represent an institutionalized pattern. That is, we have studied single reforms in five Nordic countries, and there are good reasons to think of future reforms that will result in other patterns and numbers. Nonetheless, there are obviously various national traditions and institutional practices of policy making within the five countries that result in a limited number of references, such as in Iceland, and many references, documented in the reference lists and footnotes, such as in Norway. As Christensen and Holst (2017) and Christensen and Hesstvedt (2019) have concluded based on their longitudinal survey of public enquiry reports, the expertization of public enquiry bodies represents an emerging trend in Norway that results in an increasing number of references.

Moreover, as clearly expressed in the Danish chapter, the low number of references in this case does not necessarily reflect that the reform did not rely on any evidence. Based on additional qualitative data, the authors argue that stakeholder evidence and practice-based evidence have been of core significance in the preparation of the Danish reform. However, this evidence was not substantiated in formal documentation and thereby not available to map in quantitative terms with a bibliometric study. So far, our observation of how Danish governments authorize their national school reforms, seems to be a consequence of a lack of national—institutionalized procedures to write public enquiry reports, like those in Norway or Sweden.

Nonetheless, the relative number of references to domestic governmental documents compared to other types of references is significant in all five countries. However, there are also vast distinctions for this dimension (from around 25% to almost 6%). Sweden and Iceland are on top, Denmark somewhere in the middle, and Norway and Finland at the bottom. As indicated in more detail in each of the chapters, these disparities must be understood and interpreted in the context of national reforms that in all countries are launched under the auspices of national governments. Nevertheless, independent of the institutional arrangement of the bodies that provide the reports, both policy makers and experts are formally expected to reference government-published documents in all Nordic countries.

Our bibliometric analysis reveals the active utilization of international references (cf. Chap. 9). Also for this dimension, our analysis unravels interesting differences. While Denmark is the country case with the most international references (36.36%), Sweden represents the case with the least amount (18.93%). As Nordin and Wahlström conclude in their chapter, “[T]his finding shows the possibility for national politics to uphold a high level of self-referentiality even when the national political agenda to a large extent is dictated by international organizations such as the OECD” (p. 244). In the Finnish chapter, the authors write that they expected to see the significant use of international sources for policy evidence and, in particular, the use of the OECD; however, data from Finland indicates a strong state involvement and concentration of expertise in state-funded bodies rather than a trust in international expertise. Interestingly, in Finland, there is a longstanding tradition of empirical research in the field of education and therefore, perhaps, not the same need to draw on OECD studies such as in Iceland, Denmark, and Norway. Another finding that created a puzzle for the research team was the (mostly) absent use of regional Nordic references in the source documents (cf. Chap. 12). Although Norway stands out as an exception among these countries, with its regional references amounting to almost 7%, this is lower than the authors expected based on previous research on the Nordic education model, which describes the region as a common unit.

Taken together, the bibliometric analyses provide important insights into how policy makers use various types of evidence to inform, back up, and legitimate school reforms. While some findings are in line with what other researchers have pointed to, that there is a commonness in terms of how the Nordic countries organize their education system (Telhaug et al., 2006), there are also results in our study that nuance and even contradict the idea of a Nordic education model and, thereby, ideas of unity and commonness. Although Nordic policy makers and experts meet in an international context to reach a consensus as stakeholders for advocating their views, they also act highly self-referentially in regional and national settings to deliberate on various possible solutions to their own national problems (see Volmari et al., Chap. 12). Therefore, we consider the national and comparative chapters in this book to provide complementary insights. Reforms are legitimized by national knowledge, which in the Nordic countries are authorized by the state, while global and international knowledge providers place fingerprints on the reforms through their soft governance systems.

The Constellation of Knowledge Providers Within a Nordic Policymaking Context

As Steiner-Khamsi indicates in Chap. 2, references in terms of citations carry epistemological connotations that represent various forms of knowledge as well as sites for knowledge production that connect national, regional, and global policies. The constellation of knowledge providers in policymaking processes is thereby changing because of new ideas, institutions, and networks that characterize both national and transnational policy (Legrand, 2021). Due to new partnerships and networks that change customary procedures for policy development, the relevance of the traditional distinction between applied research, produced by multiple research institute types, and basic research, produced by universities, is challenged (Gibbons et al., 1994; Nowotny et al., 2003; Stokes, 1997).

In some Nordic countries, governments are funding research programs that involve both universities and research institutes in the same machinery of producing research, expected to make an impact on policymaking processes. A core aspiration of these programs, auspices by, for example, research councils, is to provide policy-relevant evidence. These programs may well transcend traditional boundaries between basic research and applied research, since evaluations assess the researchers and their research impact according to the same standards (Smith et al., 2020).

In tandem with new types of research and evaluation programs, the configuration of actors participating in a country’s policy advisory system is also in transition (Christensen and Holst 2017; Steiner-Khamsi et al. 2020). As demonstrated in both the national chapters and comparative chapters, the national governments and their state agencies are themselves significant knowledge providers. According to Baek et al. (see Chap. 9), government papers that summarize information as well as research comprise more than one quarter of the references in the Nordic sample. In the Swedish case, Nordin and Wahlström show that the state has produced more than half of the references, representing a system of self-referentiality (see Chap. 8). The role of references produced by governments and their policy departments and national agencies underscores the important role of the state in orchestrating and defining the policy issues at stake in the five Nordic countries under study. Moreover, as analyzed in the Norwegian chapter by Hörmann and Sivesind (Chap. 7), references published by the government are important for linking to and legitimizing policies from previous reforms and arguments, retrospectively. In Iceland and in Denmark, there are also highly politicized processes, where either the minister has a clear opinion, as in Iceland (see Magningttir & Johansson, Chap. 6), or where political parties are involved in negotiations before a paper is written (see Juul Reder & Ydesen, Chap. 4).

The number of references produced by research institutes that have had a traditional role of being knowledge suppliers to governments differs among the Nordic countries. Internationally, the label research institute encompasses many types of organizations that vary in the degree of how public they are (Late, 2019). They can according to Gulbrandsen (2011) be described as boundary organizations, as they often operate as agencies crossing the boundary of science and non-science (Late, 2019, p. 52). Not surprisingly, the use of references produced by such research institutes is central to both Norway’s and Denmark’s policymaking processes. In Norway, research institutes were initially established to perform R&D areas of interest for sector authorities. As Baek et al. (Chap. 9) and Steiner-Khamsi et al. (Chap. 10) demonstrate, the institute sector is a highly important knowledge provider that conducts educational policy research and research-based evaluations in Norway. Likewise, as Juul Reder and Ydesen’s analysis in Chap. 4 depicts, the Danish Evaluation Institute (EVA) and the Danish National Research Centre for Social Research (SFI) represent two large agencies that produce policy-relevant, specialized knowledge under the auspices of two public ministries.

There is a striking difference between Norway and Denmark on the one side and Sweden and Finland on the other side. While the research institutes dominate in the two first, public agencies are more powerful as knowledge providers in the two last. In Sweden, the public agency Skoleverket serves the role of being the most important knowledge provider. This agency generates official knowledge and research statistics about the school system and childcare in the country. The agency produces policy-relevant data as well as research reports that are frequently referenced in the white papers in the Swedish data. A similar reference pattern seems to be the case in Finland. Volmari et al. (Chap. 5) underscore the strong expert position and power of the Finnish National Agency of Education.Footnote 1 This organization collects its own data and provides analyses and evaluations. This governmental institution is in itself an expert body similar to the Swedish agency albeit shorter history as an independent agency. Nonetheless, Volmari et al. explain that two universities in Finland have two publicly funded research centers that play a vital role in producing sector-based knowledge, not the least, OECD-funded studies, with a special responsibility for PISA. One of these institutes has been part of the EIPPEE network. We find similar centers in all Nordic countries that link the universities to globalizing policy spaces with a certain impact on policymaking processes in the Nordic countries.

As mentioned, the scarcity of peer-reviewed academic references, including academic books and journals, is a striking feature of the reference pattern in our documentation. In Chap. 9, Baek et al. demonstrate that 30% of the references in the Nordic sample are journal articles and books. In Chap. 10, Steiner-Khamsi et al. present an analysis of the types of references that receive focus in the white papers and green papers in the Swedish and the Norwegian cases, using a detailed classification system that distinguishes between both national and international academic research references. They report that, while national and international academic research comprises about 40% and 25% of the references in the Norwegian green papers and white papers, respectively, in the Swedish context, such research constitutes only 20% of the references in green papers and none in white papers. To understand these variations, more in-depth analysis is needed. However, the numbers alone may point to the status of academic research in Sweden, where only 7% of the total number of references represented national academic research. Furthermore, as Wahlström and Nordin conclude in Chap. 8, we need to consider how intermediary organizations mediate academic research, summarizing and translating results and interpretations in a simplified and accessible way.

Our analysis thus far does not provide much empirical knowledge for the five countries on the use of references from organizations that present themselves as think tanks. Christensen and Holst (2020) refer to think tanks as organizations that aim at influencing political debates and decisions by referencing knowledge and information. The think tanks can be considered as Rich (2004, p. 150) defines the term: “independent, non-interest-based, non-profit organizations that produce and principally rely on expertise and ideas to obtain support and to influence the policy making process.” Medvetz (2012, p. 213) extends this definition by defining think tanks as a “hybrid institutional area situated at the nexus of the political academic, economic and media fields.” Notwithstanding, in the Nordic countries, both as organizations and as discursive spaces, think tanks are relatively seen, new inventions, since the organizational figuration of actors as described in the literature on think tanks differs in comparison with the more established research institutes that provide sector research.

One key disparity is that research institutes in the Nordic countries receive funding from the Research Council of Norway and conduct commissioned research that are regulated by the contracts with public and private partners. While research institutes must follow formal contracts that to some extent prohibit them from taking an active part in political negotiations, think tanks pursue ideological agendas as discursive tools. Therefore, research institutes do not act and react to political issues in the ways that think tanks can do. Therefore, think tanks claim to generate complementary knowledge in comparison with research carried out by universities and research institutes, both because of their institutional affiliation and due to their ways of deploying knowledge and expertise. Findings based on the POLNET study indicate that our documents cite think tanks only to a modest degree, if at all, in the reference lists. Yet, they deserve a closer look in future research, as they may become influential through advocacy coalitions (Sabatier & Weible, 2007; Steiner-Khamsi, 2021) and policy assemblages where policy is produced through the ancillary of political and environmental contexts (Legrand, 2021).

Taken together, the research and development systems (including innovation) and the policy advisory systems within the five Nordic countries studied in this book have certain similarities, but the sites at which policy knowledge, such as research and evaluations, are produced do reflect significant variations. A lesson learned from reading the chapters of this book is that there are no clear boundaries between the sites for knowledge production and the sites for knowledge usage. This again leads to unclear separation of roles of policy makers on the one hand and policy advisory commissions, including researchers, on the other hand.

International Organizations: The Supremacy of the OECD

As part of the rise of the evaluative state (Maroy, 2009; Neave, 1988, 2009), research-based evaluations and surveys provide evidence and knowledge that are useful for legitimizing school reforms. Moreover, as Martens et al. (2016, p. 518) declare about the effects of this evolvement: “There is little doubt that international assessments established a ‘new center of gravity’ in the field of education and that they re-shaped education policy-making and practices in many countries.” For this and other reasons; universities, university colleges, and the various forms of research institutes act as competing or collaborating tenderers/bidders in seeking to manage various types of government-funded evaluations in the Nordic countries. This is obviously the case in the Norwegian context, where the Ministry, following major educational reforms in 1990, decided to introduce research-based evaluation programs at the turn of the 2000s (see also Zapp et al., 2018).

Besides the formal research and development systems that we consider as domestic knowledge providers, this book unravels the importance of international organizations as core knowledge producers in policymaking processes. As Steiner-Khamsi (2013) suggests, international knowledge and comparative studies in particular can influence policymaking processes (a) as evidence that informs policy planning within particular contexts, (b) as normative guidelines for how to change educational processes concerning global problems, or (c) as projecting best practices that are evaluated against a set of international performance standards. A key point in Steiner-Khamsi’s work is that the production of knowledge, especially the design of comparative research projects, optimizes evidence for the research impact on features that are not necessarily structured in the same ways across the selected sample of cases. Against this background, we expected extensive references to comparative studies for legitimizing reforms of certain features, independent of regional commonalities that are traditionally associated with the Nordic model.

Based on our sample of documents, we find that OECD is the most significant knowledge producer of the international publications referenced in all the Nordic cases. In the comparative Chap. 11, Ydesen et al. conclude that the five country-specific cases reveal multiple layers in OECD-related references. Based on their analysis, the authors emphasize that the OECD policy instrument carries more weight than that found in a quantitative analysis of references only. By forming a powerful epistemic space, the OECD is more powerful than all other international organizations. Nordic countries have a long-lasting tradition of being advised and supported by the OECD on how to govern their educational system (Ydesen, 2019). By the turn of the century, the attention of the OECD became extended by the PISA study, in which all five countries participated from the year 2000 (Sivesind, 2019).

Verger et al. (2019) recently compiled a literature review on how the OECD influences education policy in diverse countries by governance mechanisms and demonstrated how the organization has affected member countries differently by their way of collecting data, evaluating quality, and generating ideas for how to improve education systems. The POLNET study shows variations in the usage of OECD references. As illustrated by the Swedish case of Nordin et al. (Chap. 8), although the number of references to the OECD is modest compared with the number of references to domestic and government sources, the Ministry turned explicitly to the OECD for analytical help to tackle its national school crisis in 2014. Likewise, in the Finnish case, as shown in Chap. 5, while the bibliometric analysis reveals evidence used in the 2014 curriculum reform was predominantly domestic and self-referential, a content analysis discloses the OECD and, especially, the PISA results were clearly visible in numbers in the policy documents. The same was demonstrated for the Icelandic case in Chap. 6; OECD evidence was used particularly to legitimize policy recommendations and design together with knowledge sources from Canada and a consultancy company, such as McKinsey.

As Steiner-Khamsi (2013, p. 27) argued, not all forms of comparison will necessarily lead to policy borrowing and lending across countries. Policy transfer depends on various conditions, not least the methodological design used to construct the tests and surveys within OECD studies. Within the Nordic region of Europe, PISA studies, together with similar large-scale assessments, create what Waldow (2019) interestingly conceptualizes as projections of best practices. This concept implies that references to best practices are not outcomes of particular conditions that regulate education policy, but rather socially constructed narratives that policy actors make to reduce the complexity they experience (pp. 4–5). Nonetheless, as noted by Steiner-Khamsi in Chap. 2, due to differences between the successful performance scores of Finnish students in PISA, the Finnish education has been glorified as one of the most successful systems in the world, and their results are outstanding also in comparison with their Nordic neighbors. Thus we ask: Given the noticeable variations in performance scores between Finland and other Nordic countries, how are policy makers and experts referencing Finland within the policy papers?

Our material does not indicate that national publications about the Finnish education system and Finnish educational practices as such attract extraordinary attention by being highly referenced in our dataset. Interestingly, in the comparative chapter by Volmari et al. (Chap. 12), documents from three Nordic countries refer to two national curricula from Finland, whereas other Nordic curricula are not co-cited. In general, white and green papers from Iceland, Finland, and Norway reference national publications from Sweden and Denmark more often than Sweden and Denmark refer to publications from other Nordic countries. Overall, regional collaboration, as presented by the interviewees, does not seem to lead to policy convergence or diffusion albeit Finland is highlighted as an outstanding system. Rather, national authorities decide in their own contexts what to learn from Finland and others, following their own institutional paths for deciding what to do within their own contexts. This observation leads us to conclude that option (a) within the typology of Steiner-Khamsi seems to be the most typical feature of policy borrowing and lending among the Nordic countries, while we are also open to the possibility that OECD studies and transnational knowledge about performance scores on PISA influence national education reforms more silently (Waldow, 2009).

Possible reasons for the self-referentiality of school reform policy are the historical and organizational conditions that policy makers refer to and which differ between countries. Conditions are important levers in reforming schools, which implies the steady need to search for context-dependent knowledge that in the next step shapes the narratives of best practices. Moreover, such disparities may not constitute research objects or topics favored within international assessments because context-sensitive knowledge is neither necessarily capable of being measured by research items used for comparative studies, nor easily transferred between the contexts that we have studied. Therefore, in general, only those dimensions that are constructed as generic, that be, in terms of values and standards, are possible targets for international knowledge transfer.

The Knowledge Work of the Government and Its Agencies

As the chapters in this book uncover, the state plays a vital role in policymaking processes in all the Nordic countries, and besides the ministries, semi-independent central agencies are core bodies that have a say in these processes (Greve et al., 2020). Although the level of independence of these agencies varies, their tasks are to obtain, translate, and implement reform ideas (Røvik et al., 2014). As already mentioned, public agencies are not per se an think tanks organization with political ambitions. Yet, they can organize international cooperation and activities, orchestrated by powerful policy actors, such as the OECD and the European Union (EU). Thereby, they function as agenda setters that mediate knowledge “from the outside” in policymaking processes that are regulated by the state. By coordinating and inviting others to join their research and development fora and to participate in various forms of knowledge work together with for example professional associations (Nerland and Karseth, 2015), the agencies advise ministries on how to deploy both a repertoire of reform ideas and standards of which national policy makers and politicians should be aware.

However, central agencies at the national level can also be viewed as gatekeepers in orchestrating ideas of how to translate, whom to involve, and how to connect topics and realms in policymaking processes. Although ministries mandate and oversee activities within boards and agencies, they can also be considered a hub that regulates its own knowledge flow between policy makers, international organizations, various stakeholders as well as researchers. This evolving layer of knowledge can be observed in how knowledge sources produced by the agencies are referenced in governmental papers, and is therefore an interesting question what these agencies prioritize to produce as well as reference in their reports.

The POLNET study has not identified reference patterns in reports produced by the state agencies. Yet, as some of the country chapters have demonstrated, national agencies seem to play an important role for providing evaluations and assessments in their education systems, that resonate policy makers and experts’ need for evidence. Simultaneously, it is important to recognize that the ministries themselves are both receivers and translators of this evidence (see Chap. 7). Anyhow, the governance arrangements of national agencies seem to be in a mode of constant change through processes of mergers between entities, organizational rearrangements, and new establishments (see Ärlestig & Johansson, 2020). Therefore, it is of interest to study how agencies and ministries interact in their search for policy solutions that potentially results in education reform. We may argue that ongoing transitions reflect the search for a workable and legitimate balance of responsibilities and tasks between politicians, bureaucrats, and experts and not at least, between ministries, boards/agencies and coalitions with stakeholders. There is a need to study the intersection between these bodies in future research.

Nonetheless, by investigating reference patterns in our database, we observed that the importance of the state and the ethos of a robust public administration system still seem a characteristic of the Nordic mindset (see Volmari et al., Chap. 12). While processes for the deregulation, privatization, and marketization of Nordic education are ongoing, researchers like Dovemark et al. (2018) concluded that the changes are not as dramatic as those occurring in other countries. Nordic countries can be described, according to Maroy et al. (2017), as closer to a Neo-statist variant of managerialism than a neoliberal variant. Alternatively, as observed by Greve et al. (2020), a mixture of multiple reform packages is in use in Nordic countries that reflects institutionalized forms of coordinating the public sector, although managerial tools are at the forefront (p. 706). This leads in the next step to the conception of a “welfare mix” (Sivesind & Saglie, 2017) that characterizes the Nordic education systems. Therefore, unraveling the complexity of governance structures and policy borrowing and lending across nations is imperative. Taken together, to understand the practice of evidence-based policy making, we need to capture how political institutions work: their procedures, routines, regulations, and relations. Moreover, further research on the central educational agencies is needed to capture how science and politics are structurally coupled (see Steiner-Khamsi et al., Chap. 10; Steiner-Khamsi, 2021).

Evidence and Expertise in Transitions Between State Government and Network Governance

The core aim of this book has been to explore distinctions in reference patterns between reform policies in the five Nordic countries. We have tried to unravel characteristics of what Eyal (2019, p. 33) labels “distributed cognitions of expertise,” that is, how expertise outside individuals is visualized through bibliometric patterns and reference networks. Certainly, it is not easy to uncover why some knowledge sources are selected by actors at the cost of others. Often, policy making is based on tacit, practical knowledge and an outcome of assemblages beyond conscious decision making (Savage, 2020). Evidence-based policies can draw on knowledge from both the outside and inside of public policies. Therefore, the location of knowledge use is equally significant as the kind of knowledge produced and used. For that reason, excellent reasons exist for clarifying various reference patterns and for developing what Eyal (2019, p. 33) labeled explicit, abstract knowledge that expands and advances the sociology of expertise. This knowledge may well be mediated through books and articles that individual experts and others can read to enlighten conversations on public policies. Moreover, researching the sociology of expertise can provide knowledge that makes sense in contexts where people and bodies develop reforms on behalf of the state. In that case, various types of conditions can stimulate collaborative processes that help policy makers and experts to make recommendations on valid knowledge.

Although not at the forefront of our analyses, the national and the comparative chapters have pointed to strategies that stakeholders deploy in policymaking processes (Chaps. 4 and 10). While public hearings have been important devices for recognizing various opinions and voices in public administrations that have been highly departmentalized (Sivesind & Skedsmo, 2020), today, many more channels are in use for collecting information and influencing policy. Organizations of stakeholder conferences, blogs, social media, the establishment of different types of reference groups, think tanks, and public-private partnerships can evolve in various directions, mobilizing decision processes that call for new constellations of innovative reference patterns that we could not unravel in our analysis of white and green papers and which need further examination. The rearrangement of public policy implies that the modern state is “beset by a burgeoning array of domestic-global political, social and economic influence” (Legrand, 2021, p. 37). Ministries and state agencies are core bodies in this landscape of expertise, orchestrating meetings with national and international experts to communicate about kinds of evidence and expertise that are relevant for revising or renewing education policy.

Simultaneously, new technology and co-governing strategies can in many ways give policy makers new opportunities to seek policy-relevant information. Crowd-sourcing procedures are nowadays used to collect information in reform processes from a range of actors; however, the respondents are not necessarily dedicated experts, specialized scientists, representatives of a particular knowledge field, or powerful stakeholders. Who is invited to participate is technically seen, contingent. Due to the ubiquitous access to digitalized information in society and new constellations of collaborators, there are good reasons to think of policy making as becoming pluralized. This pluralization of policy may transcend institutional boundaries that have guaranteed corporate decision making into a discursive policy space that allocates the attention of policy makers to new agenda-setters in education policy.

Nordic countries are known for their corporative traditions where the state and the government have granted access to certain types of interest groups and organizations in arrangements such as public advisory bodies (Åberg et al., 2019). Such arrangements have induced both the stability and legitimacy of policies within civil society that in our case include powerful organizations, such as teacher unions. However, there are good reasons for seeing the policy system as well as traditions for developing professional expertise potentially transformed by evidence-based policy and thereby challenged. Therefore, to create legitimacy for changing policies in the field of education, the complexity of both conditions and expectations urges academic enquiry.

As Eyal (2019, p. 36) acknowledges, the application of expertise depends on “being connected with a network of expertise composed of other actors, devices and instruments, concepts, and institutional and practical arrangements, distributed in multiple loci, yet assembled into a coherent, collective agency.” This agency must deal with self-referential problems about the reflective use of policy knowledge that, in our case, refers to state-authorized school reforms. However, in our time, it must also look outside its own boundary to seek solutions to global problems in collaboration with others. Therefore, the international orchestration of policy spaces and the interdependent matrix of processes between various knowledge providers, deserve researchers’ attention in future research.