The Dialectic of Transnational Integration and National Disintegration as Challenge for Multilevel Governance

: A large number of studies have detected that within the EU multilevel governance there is a transformation toward a hybrid knowledge co-production that overcomes traditional categories such as locality or embeddedness. There, a sort of sustainable decision-making knowledge is co-developed and theoretically supposed to be applied top-down on the national level of EU member states. However, in practice such processes of uniﬁcation are always associated with a risk of limited compliance with speciﬁc national situations and with a speciﬁc national “world of relevancies”. Despite the rise in popularity of these top-down initiatives within international policy levels, there is a lack of studies that empirically analyze how national policy systems respond to these global standardization approaches. Therefore, the central aim of this study is twofold: Based on an exemplary case of an international information system co-produced by an expert group of the European Commission, it ﬁrst reconstructs whether and how transnational information is integrated on the national policy level. Second, it elucidates factors limiting an application. The results show that this international knowledge system was used for basal purposes and was mainly challenged by non-compliance with national speciﬁcities and the existence of alternative knowledge sources.


Introduction: The Internationalization of Knowledge Systems in Modern Societies
Internationalization of policy advice systems and the genesis of numerous multinational advisory organizations on the supranational level has been a focus of various lines of research (Gornitzka and Sverdrup 2011;Dür and González 2013;Décieux 2020). These processes can, on the one hand, be interpreted as a process of hybridization eroding traditional territorial boundaries of many policy issues and combining horizontal and vertical links within appropriate governance frameworks. On the other hand, internationalization has been observed on the national decision-making level through the increasing demand to consult international or transnational knowledge systems as a source of strategic advice and risk management (Adler and Haas 1992;Renn and Klinke 2013;van den Berg 2017). These developments are especially true for the European Union (EU) policy system (Boswell 2008;Münch 2010;Heinen et al. 2019).
The Lisbon reform package aimed to develop "the most competitive and dynamic knowledge-based economy in the world" (EC. European Governance 2001) until the year 2020. As mentioned in the White Paper on European Governance, the policy making process should be more open for its citizens and by this promote more accountability and responsibility for all those involved (EC. European Governance 2001). Based on this opening of policy processes, member states often act together within specific European expert groups focusing on current political problems and are able to tackle their perspectives and concerns more effectively (EC. European Governance 2001). These groups can be seen as consultative entities comprising external experts advising the commission in the preparation of legislative proposals, policy initiatives, or tasks of monitoring, coordinating, and cooperating with the member states, and "evidence-based policy making" (Gornitzka and Sverdrup 2011;Heinen et al. 2019). The idea behind evidence-based policy making is that the different member states involved should influence each other in a reflexive manner such that they benefit from sharing their perspectives and diverse "worlds of relevancies" (Décieux 2020). Their processes of knowledge exchange should enhance the sustainability of the co-produced knowledge and direct the member states toward common policy objectives (EC 2009;Jahn et al. 2012).
As a complement to the process of evidence-based policy making, the Open Method of Coordination is applied. The process of the Open Method of Coordination structures the diffusion of common European targets and guidelines through the policies of the member states (Radaelli 2003;Büchs 2008). Here, knowledge co-produced within evidence-based policy making should be applied within a top-down process on the national level. The central aim is that these common processes should allow for comparing efforts and performance of the member states, for detecting and exchanging best practices and by this mutually learning from the experience of others (EC. European Governance 2001). However, the application of internationally co-produced knowledge at the level of member states is by no means mandatory or legally binding within the Open Method of Coordination. The European Commission assumes that the Open Method of Coordination will effectively influence policy making at the national level and that national policy makers will adopt the transnational strategies and objectives (Büchs 2008).
Thus, the question arises, why member states should implement transnationally coproduced objectives and tools if they are not obliged to do so (Büchs 2008). Even if such approaches are worthwhile from a theoretical perspective, questions of adequate fit and appropriateness arise at the national policy level (Zeitlin 2005). Given the popularity of this top-down application of initiatives on international policy levels, it is quite remarkable that there is a lack of studies that analyze how national policy systems respond to such international policy standardization approaches. Therefore, the present study empirically reconstructs whether and how such a transnational information tool co-produced on the EU level is integrated in national policy systems and elucidates potential reasons and factors limiting its application on the national level. This paper contributes to the existing interdisciplinary debates on knowledge co-production, processes of evidence-based policy making and multilevel governance.

Co-Production and Integration of European Knowledge on the National Decision-Making Level, a Theoretical Reflection
From a phenomenological perspective on modern systems of knowledge production within evidence-based policy making, frameworks can be interpreted as hybrid knowledge production. In that interpretation "supranational layers of governance"-in our case within the framework of the European Union-allow members of the domestic policy advisory system to participate in the intergovernmental policy advisory systems (Geuijen et al. 2008;Braun and Van Den Berg 2013;van den Berg 2017). From a sociological perspective on knowledge, this might be characterized as a "Hybrid Forum" (Gibbons et al. 1994;Décieux 2020). The major goal of such a hybrid forum is to co-produce expert knowledge by reflecting and bargaining about a heterogeneous world of relevancies and to come to a common solution and by this to direct the different member states toward common sustainable development goals (Décieux 2020). Theoretically, actions of a hybrid forum aim to find a consensus of all its members, and traditional categories such as discipline, locality, or nationality are overcome, resulting in a one-size-fits-all solution that is "socially robust," which means that it is precise and applicable to the specific case or decision-making context (Lang et al. 2012;Steiner 2015a, 2015b).
However, this is exactly where the problem lies: If this standardized knowledge co-produced on the European policy level is applied within such a universal strategy as the Open Method of Coordination, it is only productive if national and regional specificities are taken into account (Kaiser and Prange 2004;Lafortune et al. 2020), since the EU consists of countries with very different levels of development and a high diversity of national situations. Hence, the specific national worlds of relevancies and national policy strategies are heterogeneous, as well. In consequence, the real-life scenarios of decision-making are usually very context-specific, and the situations are complex. Therefore, such processes of standardization on the supranational level are often associated with the risk of limited or non-compliance with specific local situations within the member states and an inadequate knowledge translation to local application needs (Heinisch 2021). Hence, a major challenge for the sustainability of the Open Method of Coordination is the extent to which the universal supranational knowledge is de facto socially robust enough for the specific national decision-making levels. Within the processes of oscillation between the particularization and regionalization of knowledge, universal knowledge is systematically contested by a variety of national relevancies such as national specificities, traditional local practices, and path dependencies within national policies. Here, an institutional competition of national and supranational relevancies is taking place caused by a "dialectic of transnational integration and national disintegration" (Münch 2010). Hence, transnationally co-produced knowledge-produced within evidence-based policy making processes-often lacks in "social robustness" (Gibbons et al. 1994) for application on national policy levels (Zorondo-Rodríguez et al. 2014;Décieux 2020). One consequence is that this decreases the likelihood of a shift toward the use of supranational knowledge within the framework of national policy levels (Zeitlin 2005). Within such cases of "multi-level governance" with competition or even rivalry between local peculiarities and specialties, it is especially important to translate supranational content into the national system as precisely as possible. Overall Klimkeit and Reihlen (2016) distinguished three possible responses to such an expectation of compliance: "full compliance," "limited compliance," and "non-compliance." If the global, supranational bundles are convergent and complementary with the local and national ones, then negotiations on cooperation are lucrative. However, if they diverge, the space for cooperation is severely limited (Steinhilber 2005).

Materials and Methods
As mentioned above, transnational knowledge systems are very popular in modern knowledge societies, and there are some studies dealing with knowledge co-production at a transnational level. However, up to now there exists little empirical information on whether and how such knowledge is or could be applied on the national level, e.g., (e.g., Klimkeit and Reihlen 2016). Here, the policy conflicts between transnational integration pressures toward such a one-size-fits-all approach stipulated within the Open Method of Coordination framework and the national responses to those pressures are especially interesting, as local actors often mainly care for their specific case of decision-making, and not for any general knowledge or some overall development goals (Krohn 2008;Klimkeit and Reihlen 2016).

Study: International Indicator Set as an Example of an International Knowledge System
To elucidate those conflicts is the central aim of this study, which uses a case study approach that focuses on a statistical indicator system that was part of a long-term policy strategy in a specific key policy area and was developed and reviewed by an European Expert group chaired by the European Commission. This statistical indicator system should promote evidence-based policy making on national level and direct the member states toward common European development goals.
Therefore, this statistical indicator system is an exemplary case for an international knowledge system implemented to promote and provide evidence-based policy making on the EU policy level that should be applied on the national level within an Open Method of Coordination framework. Structurally, this expert group consisted of national representatives who worked in national ministries and national statistical institutions or research centers, as well as international representatives coming from the relevant department of the European Commission and from Eurostat. By examining this exemplary case, this paper reconstructs possibilities and limits of the application of this transitionally co-produced knowledge system on the national level. The reconstruction is guided by the two following research questions:

•
How is transnational knowledge applied on the national level? Referring to the integration of transnational knowledge within the decision-making processes of national policies.

•
What are the possibilities and limits of an application on the national level? Covering reasons for processes of disintegration within the member states.

Data Collection Methods
Following the recommendations of Martin (2012) as well as Koier and Horlings (2015), the present study used a multi-method case study approach to reflect whether and how the transnationally co-produced knowledge is applied on the national level. Data for the case study were taken from two different data sources (method triangulation): A documentary analysis of relevant publications written by the expert group and by the European Commission (e.g., meeting minutes or commissions papers). These documents provided background information about the expert group, helped me to understand the structure of this group, and gave a hint about the political aim of this group. In addition to the documentary analysis, six expert interviews were conducted with interviewees selected through theoretical sampling (Cohen et al. 2007). Here, the major aim was to get insights from heterogeneous actors that are potential applicants of the knowledge system within the framework of the Open Method of Coordination. These are especially the members of the expert group that co-produced the knowledge system. As this expert group can be interpreted as a hybrid forum the theoretical sampling aspired to select applicants of the knowledge system who differed in relevant properties. These were group members from different member states of the EU with different national worlds of relevancies and application as well as representing a wide range of disciplinary backgrounds (statistics, policy, science, practice) and, by this, heterogeneous relevancies and areas of application. This design allowed for empirical reflections of processes of knowledge application from all relevant stakeholder perspectives ("perspective triangulation") (McKim 2017). The main sources of information were the interviews, which were organized as semi-structured, most lasting between 60 and 90 min. With each interviewee's permission, each interview was conducted in person by the author and recorded, then transcribed. Depending on the preference of the interviewee, the interviews were conducted in English or German. German interviews were translated into English.
The aim of the semi-structured interview strategy was to generate a consistent baseline of topical coverage and a flexible strategy of discovery (Berg 2004;Van Maanen et al. 2007). Thus, in most cases, I allowed participants to speak at length and uninterrupted. Within the analytical strategy of this case study, data collection and analysis were developed together in an iterative process (Hartley 1994;Kuckartz 2019). This allows a theory development that is grounded in empirical evidence. All interviews were analyzed using qualitative content analysis with a mixed strategy of deductive and inductive category development (Kohlbacher 2006;Koier and Horlings 2015;Kuckartz 2019). A central goal of this case study was to search for patterns in data and to interpret them in terms of the theoretical background of the case study, the social setting in which it occurred, and the results of the document analysis (Patton and Appelbaum 2003). Thus, the overall goal was to match these patterns and to develop a detailed case description. Finally, the findings of the study were cross-checked and discussed by three external experts (who were academics in the related research areas of transdisciplinary and cooperation research, policy advice, and national social policy) as well as by three participants to ensure validity (Hartley 1994).

How Is Transnational Knowledge Applied on the National Level?
Officially, EU expert groups give guidance regarding policy making and support the European Commission in decision-making (EC 2015). As the first quote (below) shows, the communicated motive to create a statistical indicator set (the exemplary case of a knowledge system) was to generate a database for evidence-based policy making for the EU and also on the national level for the member states. The first quote indicates the perspective of the chair of this expert group, which covers the perspective of the European Commission on this group and the group targets.
"The major goal of this group is to reinforce the evidence-based policy on both sides, EU and national level. So it was decided to set up this expert group in order to identify and decide about common indicators that can be a kind of support of the implementation of a cooperation framework." Expert 1: European Commission However, as the following quotes indicate, the indicator set was not primarily used as a base for finding a rational decision within processes of evidence-based policy making. It was, for example, used as an instrument of orientation for national policy makers. On the national level, such a European evidence-based policy making instrument is largely seen as a knowledge tool to reflect the key areas of the political strategy on the EU level. For example, it may be consulted to get an overview on the main interests of European policy and its key areas.
"The dashboard [indicator system] is a kind of a statement, a summary of the EUROSTAT databases, so it reflects and represents the main interests of the EU. And therefore, I think it is important for the EU to communicate the important areas of their policy. And for us to inform us about their key areas." Expert 2: National Actor Taken together, this quote and the following (Expert 3 and Expert 4) quotes support the same view of the indicator system. The different experts agreed that the knowledge represented by the international knowledge system is primarily used to get a rough and basal orientation.
"It is good to get a first impression as you can easily compare the harmonized data with on view." Expert 3: National Actor This rough orientation is, for example, used for a raw international comparative approach. In such application cases, national policy makers use the knowledge tools to get an indication of how the national policies have performed compared to other countries. Such benchmark analysis is usually done at specific steps of a policy process, and also to, e.g., create agency.
"I think we use it most when we are doing some broad exercises. When we are at the beginning or the end of a government period, then we make an overview, and consider questions like 'ok where are we with our policy?' [ . . . ] we use the indicator dashboard to get a grip on how we are within the European fields." Expert 4: National Actor When the interviewees talked about the direct political use of the indicator system on the national level, it became obvious that the indicator system is used in different ways. However, the initial purpose to guide national policies toward a standardized approach was not frequently mentioned. Even if this is an important function within national policies, those forms of use are far from the intended use case of the indicators system and the actual motive for the use of evidence-based policy making: to be the base for national policies. For the purpose of evidence-based policy making, the internationally produced knowledge does not seem to be precise or specific enough, as shown in the next two quotes.
"It could perhaps be a first indication. But it isn't even enough to describe what really makes the difference between the countries or even between two countries, if you would like to go into detail." Expert 3: National Actor Here, Expert 3 expressed that the international knowledge tool can give a brief descriptive indication of differences between countries, but the information available does not allow for finding reasons for the differences as the knowledge is not detailed enough. In order to gain more precise insights for the decision-making context on the level of the member states, which may also be able to guide national policies, other evidence sources are consulted, as described in the following quotation from Expert 2.
"It is a reference or starting point for an analysis. After having looked at the dashboard you can check other sources to go more in detail." Expert 2: National Actor Besides this function of a rough orientation and activation of engagement, it became apparent that such an international knowledge system seems to have something like a hidden function within national policies. The following two quotes reveal that such international knowledge can also be useful in negotiations with national colleagues, such as opposition parties. Here it seems to be a strong argument if policy makers can substantiate their purposes or issues by arguing that they are in line with a priority of the European policy strategies. So, the possibility of basing relevancies on independent supranational knowledge tools that are easily available seems to facilitate having an influence within national political discussions. Hence, they base their argumentation on it not because they are convinced of the substance of this knowledge, but because it tends to make a good impression within a national discussion.
"And for the national context itself it is also very important for policy makers if they could tell that a problem is also reflected in the European policy. It is always a nice argument to be able to present your arguments and your opinions based on evidence taken from an independent source and especially from the European context. . . .
This really helps to make strong statements of certain issues. To show that it is not only their position that you have to change something. If you can say-hey it is reflected by European statistics and it is also there a European priority." Expert 4: National Actor Concerning the usage of this exemplary case, it may be concluded overall that this knowledge system has specific use cases, and that national actors in general find the idea of such a common tool interesting and sympathetic. For example, they use it to gain an initial, rough orientation, as the first step of a deeper analysis, and in situations when it suits their argumentation. However, what also became apparent was that national policy makers do not use this European tool in the originally intended sense of evidence-based policy making, to guide their political decisions with that knowledge. For this purpose, the available information seems insufficient or unsuitable.

What Are the Possibilities and Limits of an Application on the National Level?
When examining reasons for limits of usage on the national level, national specificities seem to play a very important role, as, for example, pointed out by the following two quotes. Here experts use expressions such as "to find a compromise" (Expert 3) or "a tradeoff between standardization and information" (Expert 6), both of which are associated with "exclude[ing] a lot of [national] specificity." "You always have to find the common denominator and to exclude a lot of specificity. I understand that you have to come together and find a compromise at a certain point." Expert 6: Eurostat "It is a trade-off between standardization and information." Expert 3: National Actor This inclusion of national specificities within an international compromise then results in a solution that is not able to adequately cover national priorities and relevancies, as exemplarily proposed by Expert 5 within the following quote: "It often does not cover the national priorities [ . . . ]. The dashboard in some cases describes things that are not relevant for my country or even not problematic." Expert 5: National Actor Expert 2 described how they even tried to work with the EU knowledge system and had to abandon this project very quickly, as the information was simply too limited and uncertain for their national policy level issues.
"We tried to use the dashboard and came to the point that such an international system to rigid for our purpose." Expert 2: National Actor Another point was that the perspective of this international information source provided information for cross-country comparison but was not able to offer information for subnational comparisons, e.g., regional comparisons within a country, which are at least as important as international comparisons in many national policy contexts.
"And we have the problem that the dashboard is made for our whole country and here we have very different regional parts." Expert 4: National Actor Thus, knowledge represented by the international knowledge tool was too rigid for most national research goals. As this quote from Expert 4 mentions, the knowledge tool is limited to allowing comparisons between countries and not within a country.
Strongly connected to this rigidity of the information, some experts communicated doubts concerning the quality of the data the statistical indicators had been based on. Exemplifying this perspective within the following quote, Expert 1 explained that most of the information was in an extremely reduced and only very simple form, which is not appropriate to use on the national level.
"I checked all indicators of the dashboard and found them interesting. However, from a scientific perspective the extreme reduction of information was not appropriate for my use. Indicators are not contextualized and most of the indicators are too simple." Expert 5: National Actor A final and major point concerning the non-usage of this international knowledge tool within the national knowledge system was that there are alternative sources of knowledge available on the national level. Here the interviewees mentioned that they can use better fitting national data sets and information systems like the national censuses (quote from Expert 2) or seek out an exchange partner in a suitable country and have a direct dialogue with colleagues focusing on specific questions (quote from Expert 5).
"On the European level the dashboard seems to be a part of the social reporting system. However, on the national level, from my point of view, there is no real value. This is especially because the different countries have their own, better fitting and more detailed information systems." Expert 2: National Actor "What we do at my institution is, we talk to people from other countries. And this in a growing tendency. Going abroad to study other ministries. [ . . . ] I was traveling quite a lot to different countries and meetings with people from different institutions at the same level. That makes the comparison easier. I do not only have the figures; I really was there and was observing how they are thinking." Expert 5: National Actor Thus, availability as well as the ease of access to alternative knowledge sources seem to be key inhibiting factors that hinder knowledge diffusion from the transnational to the national level.

Discussion
Multinational advisory organizations on the supranational level, especially at the EU policy level, are a central characteristic of modern policies (Adler and Haas 1992;Legrand 2012;van den Berg 2017). These aim to foster processes of evidence-based policy making within hybrid processes of knowledge co-production and by this to overcome traditional categories such as locality or embeddedness (Décieux 2020). They also aim to form a sort of sustainable decision-making knowledge, not only for the international level, but also for the national policy levels. Within the EU policy framework, the Open Method of Coordination is one approach that aims to influence national policies by transnationally co-produced objectives within a non-binding legal framework. Although such an initiative sounds promising in theory, processes of unification, aggregation, and standardization on the supranational level are always associated with a risk of limited compliance with the situations within the member states and their specific local needs (Zorondo-Rodríguez et al. 2014). Using the data of six expert interviews and a documentary analysis, this study reconstructed the exemplary case of an international information system co-produced by an expert group of the European Commission, reflected on whether this international knowledge system was integrated into the evidence-based policy making processes of national policy systems, and investigated factors limiting its application. By this it was possible to empirically reflect processes of a dialectic of transnational integration and national disintegration.
The results showed that the knowledge tool was only used for very basal purposes on the national level and that the key objectives-to promote evidence-based policymaking on the national and the regional level and to promote member states toward common EU sustainable development goals-were not diffused. The knowledge provided by this international tool was not suitable as evidence in national decision-making processes, as it was not consulted to form a decision in the sense of rational decision-making. This international knowledge is reportedly strategically used to substantiate the arguments for already reached decisions, e.g., within national discussions, in cases where the arguments are in line with the international priority of the Open Method of Coordination (Boswell 2008;Boswell 2009;Décieux 2020). Büchs (2008) defined such a scenario as an "invited dutifulness", where the Open Method of Coordination agenda provides justification for previously planned but unpopular policy reforms at the national policy level. However, it has to be noted that in the case of an invited dutifulness the knowledge system was not consulted as an evidence system in the original sense of evidence-based policy making on a national or regional level (Büchs 2008;Décieux 2020). Thus, the motive for use within processes of invited dutifulness at the national level can rather be characterized as "cherry picking for evidence" (Patton and Appelbaum 2003) or "policy-based evidence making" (Strassheim and Kettunen 2014) than as evidence-based policy making.
A first central reason that it was not preferred as an evidence source was that the transnationally co-produced decision-making knowledge lacked in social robustness (Gibbons et al. 1994) for the national application context, as it was not specific enough, not precise enough, or even did not cover national political priority areas. Thus, following the compliance typology of Klimkeit and Reihlen (2016), this transnational knowledge source only achieved limited compliance to national needs and relevancies.
Moreover, it became obvious that alternative knowledge systems mostly coming from national knowledge sources are a central reason for the non-usage. The existence and easy availability of such a data source in a member state that is usually able to cover more national relevancies seemed to be the central basis for the non-usage. Thus, the transnational integration was challenged by concurring and better fitting alternatives that caused the national disintegration of transnational knowledge systems at the national level. From that perspective it could be argued that this transnational knowledge system lacked an appropriate translation to national or local needs (Heinisch 2021).

Conclusions
Theoretically the idea of transnational knowledge standardization is beneficial, and I strongly support such approaches as they offer synergies. It is an attempt to change some of the traditional policy practice and styles. Programs such as the Open Method of Coordination in combination with programs of evidence-based policy making have considerable potential to improve EU governance, but there is a lot of space for improvement in the implementation (Radaelli 2003;Büchs 2008). One central challenge such a shared knowledge system has to overcome remains the dialectic of transnational integration and national disintegration. Thus, international actors providing such knowledge tools should take national concerns very seriously and put their efforts into the integration of the specific needs of its applicants, for example, in this case by considering information at both the national and regional levels. Such a knowledge system that indeed requires a lot of political resources is only useful if it is de facto used appropriately, and this requires what Haas (2006) described as "replacing the national".

Limitation
It could be argued that this expert group might represent a highly specific case that is not representative of international expert groups per se. Alternatively, this knowledge system can be seen as a typical outcome of an exemplary expert group of the European Commission. The group can be seen as typical because the legal basis and structure of decision-making processes is similar for all expert groups of the European Commission. Their organizational structures are spelled out in the following passage of the "Register of Commission Expert Groups and other similar Entities": (EC 2015) "None of this input is binding . . .
[it] remain[s] fully independent regarding the way they take into account the expertise and views gathered." Thus, the transnationally co-produced decision-making knowledge is not binding within the overall policy processes on the EU level and as well for the national policy makers. Moreover, as Büchs (2008) provided, the Open Method of Coordination is per se "softer than the 'classical', 'regulatory' or 'redistributive' mode of EU governance" and not binding for the policies within the member states. Her article presents a larger number of political examples for the dialectic of transnational integration and national reactions of invited dutifulness within several EU member countries (e.g., Germany, France, and Great Britain).
This enables the perspective to generalize results of this specific case by extrapolating and testing them across various cases and contexts, and to find similarities and dissimilarities (Mayring 2007;Krohn 2008). Such a moderate generalization allows researchers to develop models that are consistent over more than one case in order to reflect reasons for similarities of cases and also reasons for dissimilarities across cases (Krohn 2008). It is highly likely that there are many similarities concerning structures and processes of knowledge co-production in other expert groups of the European Commission and transnational "co-products" that are produced as rational decision-making evidence, but ignored or in the best case used strategically at the national level.
Funding: Open access funding enabled and organized by the Open Access Publication funding of University of Duisburg-Essen.