Leiden University
Faculty of Social Science
Institute of Public Administration
Master Thesis
Legitimacy Crisis of the Experts of Political Parties?
A Comparative Analysis of the Role of Expertise of Political Party Affiliated
Think Tanks in Germany
Submitted by Noa Sophie Korth S1833987
MSc in International and European Governance
Supervisor: Vasileios Karakasis Second Reader: Natascha van der Zwan
1 Abstract
Political party affiliated think tanks have a long history of spreading political education to support the pluralistic character of German democracy. However, since the creation of the AfD, the legitimacy of the think tanks is challenged. This thesis is going to analyse the legitimation of the think tanks by examining the production of expertise. Thereby, it will examine the factors of the expertise production which increase or decrease input and output legitimacy. Resultingly, the thesis is going to give concrete recommendations of how the think tanks can strengthen their institutional standing. From a theoretical standpoint, the research is going to adopt a new perspective on legitimation. A new model is created which addresses legitimacy as a dual-process involving the interactive actions of the authority and the audience.
Keywords: input legitimacy, output legitimacy, throughput legitimacy, political party affiliated think tanks, expertise
2 Abbreviations
AfD Alternative for Germany
CDU Christian Democratic Union for Germany
CSU Christian Social Union (Bavaria)
FDP Free Democratic Party for Germany
FES Friedrich-Ebert-Stiftung
HBS Heinrich-Böll-Stiftung
HSS Hans-Seidel-Stiftung
KAS Konrad-Adenauer-Stiftung
RSL Rosa-Luxemburg-Stiftung
3 Content 1. Introduction ... 4 2. Theory ... 8 2.1. Literature Review ... 8 2.2. Theoretical Framework ... 9
2.2.1. Audience-driven legitimacy: three types of legitimacy ... 10
2.2.2. Authority-driven legitimacy: expertise as source of legitimacy ... 15
2.3. Connecting legitimacy and expertise ... 18
3. Research Design ... 21
3.1. Case Selection ... 21
3.2. Method of Analysis ... 22
3.3. Operationalization ... 26
3.4. Reflections on Validity, Reliability and Informed Consent ... 30
4. Empirical Findings ... 32
4.1. Audience ... 32
4.2. Throughput legitimacy ... 33
4.3. Non-instrumentalized expertise production ... 38
4.4. Connecting the findings to audience-driven legitimacy ... 42
4.5. Instrumentalized expertise production ... 45
5. Analysis ... 47
5.1. Testing H1: throughput legitimacy ... 47
5.2. Testing H2: non-instrumentalized expertise production ... 51
5.3. Testing H3: instrumentalized expertise production ... 55
6. Conclusion ... 59
Bibliography ... 66
4
1. Introduction
The head of the department of politics and consulting of the Konrad-Adenauer-Stiftung stated that “almost no think tank in the whole world is neutral, in the truest sense of the word, nearly all think tanks do follow their own political agenda” (Borchard 2014, p. 2). Many different types of think tanks exist such as autonomous think tanks, government affiliated think tanks or university affiliated think tanks (TTCSP 2019, p. 13). The Konrad-Adenauer-Stiftung is a political party affiliated think tank in Germany. Deriving from the term ‘political party affiliated think tank’, one would associate a political motivation underlying the work of the think tanks. Resultingly, one would hardly conclude that this type of think tank acts neutral. However, German political party affiliated think tanks are unique. They are the only political party affiliated think tanks in the world that receive public funding by the government (Borchard 2014, p. 3). Therefore, the think tanks have the duty to function in a distance to parties and to not politically support them. In contrast, the think tanks have the function to foster the creation of the political will of German citizens through independent political education (Heisterkamp 2014). Hence, the think tanks are an institutional mechanism to safeguard the pluralistic character of German democracy. Although it sounds simple in theory, the proximity of the think tanks to the parties causes tension. This tension of the simultaneous distance and proximity to parties is denied by the think tanks (Borchard 2014, p. 3). However, recently it became an item of public criticism.
The political landscape in Germany has changed drastically since 2013. Six political party affiliated think tanks exist in Germany. They are affiliated to the liberative/conservative CDU/ CSU, the social-democratic SPD, the free-democratic FDP, the left-wing Die Linke and the green party Die Grünen. The think tanks were widely accepted in the society and especially appreciated by students who could apply for scholarships. The legitimacy of the think tanks was not frequently criticized until the creation of the Alternative for Germany (AfD). In 2013, the party gained popularity due to the dissatisfaction of the handlining of the immigration crises (Kim 2017, p. 4). The often as right-wing populist classified party is setting up their own public funded epistemological community. In 2018, the AfD officially acknowledged the Desiderius-Erasmus-Stiftung as their political party affiliated think tank. One year later, the AfD tried to enforce the public funding of the Desiderius-Erasmus-Stiftung, in the size of 1,4 million euro, by legal action. However, the Federal Constitutional Court declined the request. The court argued that the think tank does not fulfil the criteria of a political affiliated think tank as the AfD is not repeatedly represented in the Bundestag (Spiegel 2019).
5
The will of the AfD is not weakened by this setback. The new message of the AfD is that the think tanks exploit the federal government based on the waste of public resources (Spiegel 2019). Therefore, the party currently aims to delegitimize the political party affiliated think tanks system in Germany (Zeit 2019). In June 2018, the AfD submitted a proposal to the German Federal Parliament to define the legal status and funding of the think tanks. In this proposal, the AfD mainly criticised the increased public spending despite the lack of legislation. According to the AfD, the public funding of the six think tanks amounted 581.428.000 euro in 2017 (AFD 2018, p. 1). The AfD argued that the German Federal Parliament cannot guarantee the legitimate use of the public funding based on the lack of transparency and accountability (AFD 2018, p. 2). Thereby, the party emphasized that no federal output measures exist to indicate the efficiency of the think tanks. Hence, it is not transparent what functions the think tanks fulfil. Resultingly, the party emphasized the illegitimacy of the approximately 300 external departments (AFD 2019, p. 2). Therefore, the AFD proposed legal definitions and the cutback of ‘illegitimate’ organizational structures (AFD 2018, p. 3).
In March 2019, all six major political parties voted against the proposal of the AfD. The CDU argued that there is no need for a new legislation based on the existing legal provisions (AIH 2019, p. 6). The FDP argued that it is the right of the German Federal Parliament to decide which projects and institutions should be supported by public funding (AIH 2019, p. 6). The FDP and Die Linke both emphasized that the AfD uses false facts in order to delegitimize important educational institutions. In addition, they both pointed out the power-maximizing behaviour of the AfD as they state that the party tries to increase the funding of its own think tank by changing the legislation (AIH 2019, p. 6). The Green party agrees with the criticism of the other parties. However, it is the only political party which stresses that there is a need for a new legislation based on the lack of transparency (AIH 2019, p. 7).
The problem of the lack of transparency is that it reinforces the beforementioned tension of proximity. Think tanks are obliged to be transparent as they are public funded institutions. However, as the Green party also has stated it, a lack of transparency emerges due to path dependencies rather than clear legal definitions (AIH 2019, p. 7). The think tanks have expanded rapidly from national to international actors which caused an increase of institutional functions and missions (Heisterkamp 2014). At the same time, this development caused an increased public funding which also needs to be justified. Therefore, the legitimacy of the think tanks is challenged as several crucial questions cannot be clearly answered in the current state of research. Firstly, the relationship between the mother party and the think tanks need to be
6
clarified. It needs to be analysed which influence the mother parties have in the production of expertise. Secondly, the functioning of the think tanks need to be systematically analysed. Thereby, the question needs to be considered if the new international functions of the think tanks are justified.
The main question of this debate concerns the legitimacy of the political party affiliated think tanks. As Borchard (2014) has stated the main source of legitimacy of the think tanks is its competence of expertise production. Motivated by this issue, this thesis sets forth the following research question to what extent does the production of expertise legitimize political party affiliated think tanks? Thereby, it will illustrate the impact of the main function of the think tanks, namely expertise, on the legitimacy of the think tanks. The question will help to clarify the underlying tensions of the debate by illustrating the functioning of the political party affiliated think tanks. As the empirical case has unique characteristics, the thesis has created a new model to test the legitimacy of the think tanks. The thesis will apply an interactive model of legitimation to the case which is derived from the key concepts of legitimation. In order to examine this complex model, the thesis will make use of a mixed-method approach. Therefore, data will be gathered from content analysis, statistical analysis and survey research.
This thesis is going to contribute to the academic field of public administration through the development of a new perspective on legitimacy. In contrast to scholars that mostly focus on the legitimation process deriving from the evaluation of the audience (Scharpf 1999; Jones 2009; Steffek 2015; Oates 2016), this thesis argues that legitimation is a dual-process consisting of actions of both the audience and the authority. Therefore, the thesis is going to connect the concepts of input and output legitimacy to the institutional sources of legitimacy. Hence, this thesis will connect literature on legitimacy with literature of expertise. Thereby, the models of expertise of Weiss (1979) will be elaborated. As a result, the insights of the different theories and models will be combined into a new model, the interactive model of legitimacy. Second, most theories on legitimacy are developed for international institutions such as the EU rather than for national advisory bodies (Schmidt 2013; Scharpf 1999; Oates 2016). However, parallels between legitimacy crises of the international institutions and the political party affiliated think tanks can be drawn. Therefore, the thesis is going to illustrate the applicability of concepts of the multi-level-governance literature on a national case.
Moreover, before an overview of this thesis is presented, this section is going to underscore the societal relevance of the study. Political party affiliated think tanks are prone to a legitimacy crises although society as a whole can benefit from their work. It is likely that many German
7
citizens are not aware of the uniqueness and benefits of the political party affiliated think tank system in Germany. Right extremist actors such as the AfD are basing their argumentation often on fake news or ‘half-true’ facts. Therefore, it is crucial to develop a clear terminology of political party affiliated think tanks to counter the arguments of the AfD. This is especially relevant in this point of time as the legitimacy crises can be reinforced by the coming economic downfalls. The spending of millions of public funding needs to be justified. Therefore, the functioning of political party affiliated think tanks needs to be analysed in order to clarify their societal importance.
In the next chapter, the theoretical base of the thesis will be presented. The literature will be reviewed and the theoretical framework will be explained. For the audience-driven side of legitimacy, the concepts of input, output and throughput legitimacy will be theoretically analysed. For the audience-driven side of legitimacy, the models of Weiss (1979) will be divided into the non-instrumentalized and instrumentalized expertise production. To connect the two sides of legitimacy, the interactive model of legitimation will be explained and three hypotheses will be established. In the third chapter, the research design will be explained which consist of a mixed-method approach. In the fourth chapter, the empirical findings will be presented according to the structure of the coding scheme of the content analysis. The fifth chapter is going to test the three hypotheses by analysing the data. In the last chapter, the conclusion will be drawn which is followed by a section on the limitations of the research. Finally, five recommendations to strengthening the political standing of political affiliated think tanks will be presented.
8
2. Theory
The following section is going to review and identify gaps in the literature. Afterwards, the theoretical framework is going to present the main concepts of the two sides of legitimacy. For the audience-driven side of legitimacy, the concepts of input and output legitimacy will be explained. The authority-driven side will be explained by the research utilization models of Weiss (1979). Moreover, the concept of throughput legitimacy will bridge the gap between the audience-driven and the authority-driven side of legitimacy. After the theoretical framework, the interactive model of legitimacy will be explained and three hypotheses will be derived from the model.
2.1.Literature Review
A lack of research on party affiliated think tanks can be identified in the academic literature. Literature concerning think tanks in Germany do not particularly focus on political party affiliated think tanks (Thunert 2001, Thunert 2004, Grunden 2013). For example, Thunert (2004) analyses the historical developments of different types of think tanks in Germany. But he concludes that “the question as how to proper place think tanks in political consultancy in Germany remains open” (Thunert 2004, p. 69). Four years later, he conducted a study comparing the political potential of 93 think tanks in Germany. While also analysing the six party affiliated think tanks, he argued that “it is difficult to distinguish potential think tanks functions from other activities of party foundations” (Thunder 2008, p. 38). Resultingly, the literature on German think tanks illustrate the lack of clarification on the political role of think tanks, in particular of political party affiliated think tanks.
Several comparative studies include German think tanks as case study (Kinderman 2017; Pautz 2010; Pannier and Schmitt 2014). The only comparative study that researches political party affiliated think tanks in Germany is of Pattyn et al. (2017). This study compares the autonomy and functioning of political party affiliated think tanks across the world. For the case of Germany, they argue that think tanks enjoy a medium level of autonomy based on the ideological alignment but separate legal staffing (Pattyn et al. 2017, p. 3). Moreover, they argue that political think tanks in Germany give cold advise through the long-term focus and the lack of daily interactions with political parties (Pattyn et al. 2017, p. 10). However, the authors make conclusions for the German case based on only two studies of Thunert (2001) and Grunden (2013). This limited source selection, distorts the complexity of institutional structures in Germany. As consequence, although the study provides interesting insights, it shows that further research is needed to clarify the role of party affiliated think tanks.
9
Ulrich Heisterkamp (2014) researched the functioning of political party affiliated think tanks through semi-structured interviews with the main six think tanks. His work can be seen as the primary work on German party affiliated think tanks. It is the first study which structurally aims to clarify the terminology. A clear result of his work is that the production and transfer of educational knowledge is the primary characteristic of the think tanks (Heisterkamp 2014, p. 471). Nevertheless, he also identifies in the interviews that clear confusions about the terminology of think tanks exist between the organisation themselves. Heisterkamp (2014) advises in his book that further research is needed to clarify many aspects such as the relationship between the think tanks and mother party.
Political party affiliated think tanks are getting criticized for the lack of accountability, transparency and informality in decision making processes (AfD 2018). Those concepts are commonly discussed in the multilevel governance literature in regard to legitimacy crises of international organizations (Reh et al. 2011; Papadopoulos 2010; Stephenson 2013). Moreover, the theory on legitimacy is mostly only applied to international institutions such as the EU (Scharpf 1999; Steffek 2015; Oates 2016; Jones 2009; Schmidt 2013). Literature on think tanks lack the consideration of legitimacy issues. In contrast, the literature on think tanks focuses on the production of expertise and its influence in the policy advisory system (Heisterkamp 2014; Pattyn et al. 2017; Littoz-Monnet 2017; Osborne 2004). Therefore, the link between expertise and legitimacy is lacking in the research of political party affiliated think tanks.
Taking into consideration the limitations of the literature, this thesis will argue that the theories developed for international institutions can be applied to party affiliated think tanks. The criticism of the legitimacy of the think tanks have parallels to the legitimacy crises of institutions such as the European Commission. The think tanks function in multilevel networks with overlapping accountability structures which decreases process transparency and increases informality. In contrast to other think tanks, party affiliated think tanks in Germany are publicly funded. As a result, they face more direct legitimacy questions than other advisory bodies. Therefore, the theoretical framework will use the concepts developed for the international institutions and adopt them in a latter step to the institutional context of Germany.
2.2.Theoretical Framework
Legitimacy can be seen from two theoretical perspectives. The normative perspective categorizes legitimacy as the judgment of the rightfulness of political institutions (Steffek 2016, p. 265). Contrasting, the sociological perspective sees legitimacy as measurable fact which does not involve normative but empirical testable conditions (Dellmuth and Tallberg 2015, p. 154).
10
This thesis is going to adapt a sociological approach as it argues that legitimacy is not a normative attribute of institutions but a measurable process. According to Zaum (2016) legitimacy arises through the interaction of two groups, namely, the exercising authority and the audience. The most common definition of legitimacy is developed by Lipset (1959) who argues that “legitimacy involves the capacity of a political system to engender and maintain the belief that existing political institutions are the most appropriate or proper ones for society” (p. 86). The definition illustrates the dual-process of legitimacy as both actors, the audience and the authority, have active roles in the legitimization process. The authority is targeting actions to increase legitimacy and the audience is evaluating these actions and the output. Therefore, both sides of legitimacy need to be analysed in the theoretical discussion to grasp the entire process of legitimacy.
2.2.1. Audience-driven legitimacy: three types of legitimacy
The first part of the process of legitimacy is audience-driven. This process is based on the way the audience acknowledges the legitimacy of the exercising authority (Zaum 2016, p. 7). The judgements regarding legitimacy are not universal. Zaum (2016) argues that “legitimacy is rooted in the collective belief of a particular community, which gives these beliefs a certain degree of stability, but also means that legitimacy can only be assessed with respect to this particular group” (Zaum 2016, p. 3). This shows that an institution does not need to be representative to the whole society but to a particular community. The channel of audience evaluation has been frequently researched through the development of several types of legitimacy (Scharpf 1999; Schmidt 2013; Steffek 2015; Oates 2016). This thesis will analyse three types of legitimacy: output, input and throughput legitimacy.
Output legitimacy
The concept of output legitimacy derives from the work of Fritz Scharpf (1970;1997;1999). Analysing EU institutions, he argued that output legitimacy refers to the ability of institutions to effectively produce outcomes for the audience. Output refers in this case to the decisions and actions of political institutions (Schmidt 2013, p. 4). Moreover, Scharpf (1999) claims that output legitimacy is centred around the problem-solving capacities of a political institution. This problem-solving ability refers to the way actions of political institutions respond to social problems. Scharpf also emphasises the importance of enforcement mechanism to ensure the efficiency of the political actions (Scharpf 1999, p. 20). Therefore, Scharpf mainly argues that the underlying condition of output legitimacy is institutional efficiency.
11
Highlighting the conceptual ideas of Scharpf, later scholars emphasized the audience-driven component of output legitimacy from a sociological perspective (Oates 2016; Jones 2009; Schmidt 2013; Zaum 2016). Oates (2016) states that “output legitimacy refers to the beliefs about the effectiveness of institutions in solving common problems” (p. 203). Therefore, the audience perceives output legitimacy through the evaluation of the problem-solving abilities of institutions. The evaluation of effectiveness does not derive from normative judgments of democratic behaviour but from conditions such as expertise (Oates 2016, p. 206). The underlying idea of output legitimacy is that democracy and institutional effectiveness are two unrelated mechanism in this matter. Especially in regard to the EU, output legitimacy enables political institutions to compensate democratic deficits through efficient institutional performance (Steffek 2015, p. 266). Resultingly citizen evaluate performance according through outcomes rather than through the representation of interests.
The output legitimacy of a political institution can be challenged in several ways. Zaum (2016) identifies two obstacles of maintaining output legitimacy. The first way output legitimacy can be challenged is if an institution does not achieve the outcomes which are desired by the represented community. Therefore, it is crucial for a political institution to be aware of the interest of the represented community (Zaum 2016, p. 3). Moreover, political institutions can strengthen output legitimacy by creating performance controls and enforcement mechanisms (Jones 2009, p. 1091). The second obstacle of output legitimacy is “changes in the underlying beliefs as to what outcomes are normatively desirable” (Zaum 2016, p. 3). Political institutions need to be informed about societal changes in order to be reactive in their institutional behaviour. Output legitimacy is directly connected to the values and identity of the represented community (Schmidt 2013, p. 5). As consequence, shifts in values or identity can have a significant impact on the desired outcomes of a political institution.
Input Legitimacy
In contrast to output legitimacy which is centred around institutional effectiveness, input legitimacy focuses on democracy. Output legitimacy concerns the effective performance for the audience. Contrasting, input legitimacy concerns the political involvement by the audience (Schmidt 2013, p. 4). A basic precondition of input legitimacy is that citizen can participate equally and can choose between different competing elites and political agendas (Hix 1998, p. 51). Another underlying assumption of input legitimacy is that citizen should be involved in the decision-making processes of political institutions (Oates 2016, p. 202). According to Steffek (2015) input legitimacy “refers to institutional arrangements that allow citizen to communicate
12
their interests to political decisions-makers” (p. 266). The definition emphasises the importance of the audience to express themselves in order to legitimate a political institution.
The input can be defined as the demands and support of the audience (Schmidt 2013, p. 4). According to David Easton (1965) input legitimacy can be derived from two main channels. Firstly, input legitimacy can be derived from democratic arrangements such as elections in the recruitment policies. However, most political institutions face a democratic deficit because of the lack of participation channels based on an elite-driven character (Oates 2016, p. 203). Therefore, the second channel which is concerned with identity formation can be the dominant factor in ensuring input legitimacy (Schmidt 2013, p. 4). For example, the European Commission is weakened through the limited democratic participation channels. However, the supranational, European identity of the audience can stabilize the input legitimacy of the European Commission because they belief that the experts in the Commission are the most adequate representors of their interests (Oates 2016, p. 203).
Similar as with output legitimacy, input legitimacy can also be challenged in several ways. As described above, input legitimacy can be challenged through the lack of democratic channels in the institution (Oates 2016, p. 203). The second obstacle concerns the awareness of the demands and interests of the audience (Schmidt 2013, p. 4). If the demands in terms of ‘articulated interest’ are not taken into consideration in the decision-making process, the audience perceives the input legitimacy of a political institution as low (Steffek 2015, p. 267). Therefore, the audience will support the actions of the political institution less. If the public support of the institution decreases, the demand for the actions of the institution diminish in a political competitive environment with several alternatives (Hix 1998). Therefore, the dynamic illustrates that legitimation is an active process which requires constant actions by the authority to maintain the support of the audience.
Throughput Legitimacy
Concepts such as transparency, accountability or inclusiveness are common research topics in public administration (Reh et al. 2011; Papadopoulos 2010; Stephenson 2013). However, seldom are those concepts directly connected to audience-driven legitimacy. The first time that the concept of throughput legitimacy was clarified was by the work of Vivian Schmidt (2013). Schmidt argues that legitimacy theorists are blind to the processes which occur in the ‘black box’ between the input and the output. Instead of a performance- or participation-orientation, throughput legitimacy can be identified as process-oriented (Schmidt 2013, p. 5). Therefore,
13
this type of legitimacy is concerned with the institutional interactions of administrators which use inputs to produce effective outputs.
Schmidt (2013) argues that throughput legitimacy has three underlying conditions: accountability, transparency and inclusiveness. The lack of accountability of political institutions is often discussed in the multilevel governance literature (Papadopoulos 2010). Based on the increasing complexity of political networks “it is often not clear whom the collaborative is accountable to and for what” (Bryson et al. 2006, p. 51). Accountability is often not clear cut because it involves multiple stakeholders with diverging interests. Schmidt (2013) states that political institutions need to be accountable to the actors that participate in the decision-making process. Thereby, the process of decision-making needs to take into account the input dimension as well as the output dimension (Schmidt 2013, p. 6). In addition, the process needs to involve accountability according to ethical standards which refers to equality in the decision-making processes. For example, the process needs to prevent the abuse of power or the unfair influence of lobbyist interest (Cameron 2004, p. 59).
Next to accountability, another underlying condition of throughput legitimacy is transparency. Transparency can be categorized as precondition of accountability but cannot be equated with accountability as accountability takes into consideration different institutional structures and mechanisms of political networks (Schmidt 2013, p. 6). In addition, transparency is a prerequisite of public involvement as it allows the audience to follow political processes (Berthier 2016, p. 424). In terms of throughput legitimacy, Schmidt (2013) regards transparency as the openness of information in two ways. Firstly, the information about processes underlying the decision-making need to be public available for the audience. And secondly, the decisions of political authorities need to be publicly accessible (Schmidt 2013, p. 6). As a result, it is not sufficient if institutions only publish the final decision-making without ensuring the transparency of the underlying processes.
The final condition of throughput legitimacy is inclusiveness. This condition refers to the question of “the balance in access and influence among organized interests” (Schmidt 2013, p. 6). Inclusiveness regards the intermediation of the interests through the creation of institutional mechanisms. Schmidt (2013) argues that the intermediation of interest is different from the intermediation of input legitimacy. The input demand can be indicated as the interest articulation of the audience in form of social activism. In contrast, throughput intermediation refers to the institutionalist channels through which demand is implemented (Schmidt 2013, p.
14
7). Therefore, political institutions can strengthen throughput legitimacy through the creation of inclusive channels of interest intermediation.
Steffek (2015) criticizes the concept of throughput legitimacy. He claims that the process dimension cannot be differentiated from the input and output dimension. He argues that “neither input nor output legitimacy are intelligible without reference to procedure” (Steffek 2015, p. 270). The legitimacy of effective outcomes of political institution cannot be established without accountability or transparency in the process. Moreover, the input to a political institution cannot lead to legitimacy without inclusive, transparent and accountable process which implement the demands. Therefore, throughput legitimacy can be rather seen as the means to input and output legitimacy rather than an end in itself.
Throughput legitimacy will function in this thesis as a bridge between the audience-driven and authority-driven form of legitimacy. Being aware of the conceptual criticism, the thesis visions throughput legitimacy as its own process which includes both the actions of the authority and the audience. The authority can have active influence over the process of expertise production which can impact the way the audience receives the legitimacy of the institution. The research is going to disentangle throughput legitimacy from the one-sided audience-driven side. It will consider accountability, transparency and inclusiveness as the indicators can have significant impacts on the input and output legitimacy. It could be the case that the audience believes in the effectiveness of the input and output but the lack of transparency in the process leads to a legitimacy crisis.
Summary Box: Audience-driven Legitimacy
Concept Orientation Key Words
Output legitimacy Performance-oriented
Institutional effectiveness Problem-solving capacities
Effective performance for the audience Input legitimacy
Participation-oriented
Communication via institutional arrangements Democratic channels and identity
Political involvement by the audience Throughput
legitimacy
Process-oriented Accountability in decision-making Transparency of process
15
2.2.2. Authority-driven legitimacy: expertise as source of legitimacy
The second part of the process is authority-driven. The process of legitimization requires the exercising authority to claim legitimacy from the audience (Zaum 2016, p. 7). Therefore, institutions need to build up the capacity to maintain the trust of a particular community. This can happen through multiple factors. For example, Ellinas and Suleiman (2012) argue that legitimacy can be derived from three sources which are legal authority, technical expertise and functional uniqueness. According to Heisterkamp (2014) the main function of political party affiliated think tanks is to produce expertise. Hence, this thesis will focus on the source of technical expertise in the legitimation process. By taking into account the authority driven part of legitimization, this thesis will analyse what active activities of expertise production think tanks undertake in order to create and stabilize legitimacy.
According to Littoz-Monnet (2017) expertise can be produced in three ways in which the expertise production is party instrumentalised. The differentiation between instrumentalized and instrumentalized expertise production can be understood as followed. With non-instrumentalized expertise production underlying mechanisms are meant which follow a collective goal. For example, research utilization is used to solve social problems of the society. In this scenario, the policymaker can play an active role in determining the expertise production. However, the policymaker does not encourage the expertise production with the primary purpose to benefit from the expertise production tactically or politically. In contrast, through instrumentalized expertise production the policymaker primarily wants to gain something such as political justifications for actions or prestige.
Combining the conceptual ideas of Littoz-Monnet (2017), this thesis will elaborate on the models of Weiss (1979). The models help to understand how expertise is produced by epistemological groups and used by policymakers. Therefore, the models can also be distinguished into instrumentalized and non-instrumentalized expertise production. Every model represents a different channel of expertise production. Some models, especially those in which expertise is instrumentalized have a clear mentioned link to legitimacy. For others, the link to legitimacy still needs to be constructed.
Non-instrumentalized expertise production
According to Littoz-Monnet (2017) ‘expert-shaped policy’ is produced when expertise is created before it gets in contact with political decision makers. In most of those cases, expertise is produced in independent NGOs or in academia. The knowledge-driven model of Weiss (1979) can be compared to the process of expert-shaped policy. In this model, research is
16
produced without the clear aim to enter the policy cycle. Rather social research is done to generate theories. However, demand for the expertise is created in politics because of the supply of academic research. Policymakers are incentivized to use the results of academia based on the revolutionizing new ideas (Weiss 1979, p. 427). For example, developments in technology led to the incentives for policymakers to create new channels of e-government (Moon 2002, p. 425). However, Weiss (1979) argues that this model occurs less in policymaking than others. More often, expertise is produced to respond to a particular social problem. In the problem-solving model researchers and policymakers have the same goals to create solutions of social problems. According to this model “the main contribution of social science research is to help identify and select appropriate means to reach the goal” (Weiss 1979, p. 427). Therefore, the main aim of this type of expertise is to increase responsiveness to the audience. In this process expertise can enter politics in two ways. Either the expertise is produced and selected through a search in the need of solutions. In this way expertise enters the political arena in a more indirect way through the advice of analysts or consultants. Or policymakers actively recruit researchers to conduct research through a commission or specialized task force. This way enables expertise to enter the political area in a more purposeful way (Weiss 1979, p. 428).
The enlightenment model of Weiss (1979) follows a different dynamic. In this model goals between researcher and policymakers do not have to be aligned. Rather, researchers conduct research to enlighten society in the long-term. Research can even be opposed to the opinion of the current government and therefore aim to create a shift in public values (Weiss 1979, p. 430). The research is spread through academic journals and media channels and thereby influence the way society thinks. Policymakers will also get influenced through this public influence. Therefore, not single studies influence specific policies but the entirety of research of an epistemic group influences the missions of policy makers (Weiss 90179, p.429),
Instrumentalized expertise production
In the second process described by Littoz-Monnet (2017) policymakers have a more active role in the production of expertise. Policymakers strategically instrumentalize the production of expertise in order to meet political agendas and objectives. The resulting policy product of this process is called ‘policy-shaped expertise’. Littoz-Monnet (2017) states that policymakers aim to directly legitimize actions through the expertise production. Therefore, the link between legitimacy and expertise is clearly constructed. Two models of Weiss (1979) demonstrate parallels to the dynamic of policy-shaped expertise.
17
The first model of instrumentalized expertise production is the political model. According to Weiss expertise is used in this case “for self-serving purposes of agency justification and personal aggrandizement” (Weiss 1979, p. 429). Social science is used to support the fixed preferences of policymakers. Therefore, policymakers use expertise in order justify actions or convince the opposition of their policy programmes. The use of expertise in the political model can be based on several underlying motivations. Littoz-Monnet (2017b) argues that the instrumentalization of expertise of policymakers can have four reasons. Firstly, the use of expertise can increase the input and output legitimacy of the actions of policymakers. Based on the increased responsive through conducting research, the audience believe that their demands are implemented. Moreover, through the use of research in the decision-making process, the audience believes that the policy outcomes will be more effective (Littoz-Monnet 2017b, p. 588). Second, policymakers can use external expertise to maximise their capacities to act in other policy areas. For example, national institutions can increase their capacities through expertise to act in the international arena. Third, through the framing of policy issues in technocratic language, policymakers can purposefully depoliticize issues. Thereby, the authority can shift the attention of the audience to other policy issues. And finally, institutions can “build an instrumental constituency of support on the ground” (Littoz-Monnet 2017b, p. 587). In the competitive environment of politics, it is crucial for institutions to maintain the support by the audience or higher authorities. Therefore, the use of expertise can increase this support by legitimizing the existence of institution itself.
The second model of instrumentalized expertise production is the tactical model. In this model social science is also instrumentalized by policymakers for political benefits. However, in contrast to the political model, the expertise is used differently. Weiss states that “it is not the content of the findings that is invoked but the sheer fact that research is being done” (Weiss 1979, p. 429). Therefore, policymakers can increase the pretence of responsiveness to the audience demands. For example, they can show the prioritization of demands by demonstrating the research is done. However, at the same time policymakers can also postpone decisions with the excuse that not sufficient research is done. Moreover, they can justify policy outcomes with limited support or strong criticism by shifting the blame on social science. As a result, epistemic communities within institutions can tactically increase the reputation of an institution solely through presence and not through effectiveness (Weiss 1979, p. 429).
18
Summary Box: Authority-driven legitimacy
Concept Model Key word
Non-instrumentalized expertise production
Problem-solving model
Responsiveness to social problems Expertise aligned to policymakers
Knowledge-driven model
Theory generation outside the policy cycle
Impartiality of academia
Expertise aligned to policymakers Enlightenment model Production of knowledge with
long-term objectives
Expertise not necessarily aligned to policymakers
Instrumentalized expertise production
Political Model Self-serving purposes Support of fixed interests of policymakers
Tactical Model Sheer fact of expertise production Pretence of responsiveness 2.3.Connecting legitimacy and expertise
The theoretical framework has shown that legitimacy is not only based on the evaluation of the audience. In contrast, the authority has also a crucial role in the legitimation process. The authority has to create and maintain sources of legitimacy in order to increase the audience-driven side of legitimacy. The different concepts and models of both sides of legitimacy are present in the literature. However, the connection between the concepts is missing. Therefore, this thesis has created a new model which illustrates the interactive process of legitimation. The interactive model is based on two processes, one is driven from the audience and one is driven from the authority. Firstly, the audience-driven process includes the concepts of input and output legitimacy. Throughput legitimacy is classified in the audience-driven side of legitimacy but has a bridging function. The arrow illustrates that throughput legitimacy is directly linked to the authority. Although the audience evaluates the throughput legitimacy, the process of the expertise production is also part of the sources of the authority. Secondly, the authority-driven side of legitimacy is based on the two sources of legitimacy, namely the instrumentalized expertise production and the non-instrumentalized expertise production. Hence, the model illustrates that legitimation is a complex process including the actions of multiple actors.
19
THE INTERACTIVE MODEL OF LEGITIMATION
To test the interactive model of legitimacy and to understand how expertise production legitimizes political affiliated think tanks, three hypotheses will be tested. The first hypothesis concerns the impact of throughput legitimacy:
H1: The lack of throughput-oriented expertise production decreases the audience-driven
legitimacy of party affiliated think tanks.
The underlying expectations are that the audience will less strongly believe that an institution is effective if it does not provide clear accountable structures. In addition, missing transparency in the process of expertise production can also reduce input and output legitimacy as the effectiveness and democratic involvement possibilities are not visible. And finally, if the channels of intermediation are not actively constructed, the authority is not able to transform the demands of the audience into expertise. Therefore, the lack of inclusiveness in the channels of intermediation can also diminish input and output legitimacy.
20
In addition, this thesis will focus on the non-instrumentalized expertise production as source of legitimacy for party affiliated think tanks. In this part of the theory the linkage between expertise and legitimacy is the least researched. Consequently, the thesis will investigate how the non-instrumentalized from of expertise influences input and output legitimacy. Therefore, the following hypotheses will be tested.
H2: The production of non-instrumentalized expertise by political party affiliated think
tanks leads to an increase of audience-driven legitimacy.
The underlying expectations are that the use of knowledge-driven expertise can increase the perception of equality standards in the process of expertise production. Moreover, the use of problem-solving expertise can increase the responsiveness to social problems of the audience. Think tanks can use expertise to enlighten the audience in the long-term which aligns the goals of the authority with the audience. Therefore, it is to be expected that non-instrumentalized expertise production has an positive impact on input and output legitimacy.
Finally, the thesis will focus on instrumentalized expertise production. As the link between legitimacy and expertise is already mentioned in the theory, this part will primarily test the theory. Therefore, following hypothesis will be tested.
H3: The expertise production of party affiliated think tanks is instrumentalized by the mother parties to increase audience-driven legitimacy.
The underlying expectation is that mother parties use the expertise of the think tanks to increase the perception of responsiveness and effectiveness. Not only the actual expertise production, but also the sheer fact that expertise is produced can increase the responsiveness of the mother parties. Therefore, the parties are able to expand the legitimacy to other areas through the production of expertise and use expertise to depoliticize policies. Resultingly, think tanks create institutional support through the production of expertise.
21
3. Research Design
This section is going to present the research design through which the hypotheses will be tested. Firstly, the case selection will be presented and justified. Afterwards, the method of analysis will be explained. Therefore, the content analysis, the statistical analysis and the survey research will be explained. Afterwards, the operationalization will follow which is guided by the coding scheme of the content analysis. The last selection will reflect on the validity and reliability of the study.
3.1.Case Selection
Currently, six think tanks exist in Germany which are affiliated with the main political parties. Based on the functional expansion of the institutions in the last 20 years, the functioning of the think tanks became increasingly complex. Therefore, it is necessary to concentrate the research on a particular part of the institutions instead of focusing on the think tank as a whole. Every political party affiliated think tank in Germany has a separate department for political education. Those departments qualify as unit of analysis for this research as they are the centre of the expertise production and transfer (Ulrich Heisterkamp 2014).
In order to qualify as small-n research four think tanks are selected out of the six political party affiliated think tanks. The case selection follows the Most Different System Design (MDSD). The selected cases differ in institutional political affiliation, size, budget and the size of the mother party but share a similar outcome. The outcome which all think tanks share is institutional stability based on the until 2014 seldom criticized legitimacy. The research aims to investigate the main explanatory variable that explains how the diverse think tanks maintain legitimacy through the production of expertise.
Table 1 illustrates the case selection. The four case studies have been chosen based on data availability. From beginning on, the think tanks of the SPD and the FDP were not able to participate in the research. They justified their withdrawal with the limited institutional
Table 1: Case Selection
Think Tank Party Affiliation Unit of Analysis
Konrad- Adenauer-Stiftung CDU Department ‘Bildung’
Hans-Seidel-Stiftung CSU Institut für Politische Bildung
Rosa-Luxemburg-Stiftung Die Linke Akademie für Politische Bildung Heinrich-Böll-Stiftung Die Grünen Department ‘Soziale Teilhabe’
22
capacities caused by spread of COVID-19. The second reason for the case selection is the fit to the MDSD. For example, based on the political position, voters and size, the think tanks of the SPD, FDP and CDU share similar characteristics. Therefore, the selection of all three cases would contrast the MDSD research design. Resultingly the Konrad-Adenauer-Stiftung is selected to represent the large size think tanks which are supported by the majority of the German population. Although the Hans-Seidel-Stiftung shares the political affiliation with the CDU, it differentiates from the other think tanks in its geographical dimension. In contrary to the other parties, the CSU is an ethno-regional party which means that it only represents the state of Bavaria (Falkenhagen 2013). The third case, Rosa-Luxemburg-Stiftung, significantly differs in its political position from the other think tanks based on the its left-wing political affiliation. Moreover, it has a significantly lower size and budget than the Konrad-Adenauer-Stiftung (DBWD 2006, p. 7). The Heinrich-Böll-Konrad-Adenauer-Stiftung, the fourth case study, is with 195 employees and a budget of 36,5 million-euro in the middle range of the think tanks. What distinguishes the Heinrich-Böll Stiftung from the other think tanks, is its mission regarding sustainability and combatting climate change. As a result, the four case studies are selected because they depict various political affiliation, sizes, budgets and functional orientations. 3.2.Method of Analysis
To analyse the complex process of legitimation, the research design will make use of a mixed-method approach. The mixed mixed-method approach consists of three elements: content analysis, statistical analysis and survey research. The threefold approach makes it possible to compare the mission of expertise production with the actual implementation. Through the content analysis the mission statements can be analysed. The survey research will deepen the understanding of the underlying intentions of the expertise production. And finally, the statistical analysis enables the comparison of the intentions of expertise production with the actual execution. In order to understand the research design, the following three sections will shed light on each method of analysis.
A) Content analysis
The main part of the data will be collected through content analysis of the websites and official publications of the think tanks. Content analysis is a powerful tool in analysing themes, frequencies, values, and symbols in media sources (Neuman 2014, p. 372). It is selected as method of analysis because it enables the tracing of the use of expertise. By coding the produced information of the think tanks, the method makes it visible which concepts of legitimacy are prioritized. The content analysis of the official websites of the think tanks will be mainly used
23
to analyse H1 and H2. Nevertheless, the information gathered from the non-instrumentalized expertise production could be applied to H3.
In the operationalization section of the research design, the coding system of the research will be presented. For each concept, indicators will be selected which help to test the hypotheses. During the research, phrases will be selected that contain the indicators. Therefore, it will be made visible how often think tanks mention the indicator and in which contexts the indicators are mentioned. This thesis will make use of both manifest and latent coding to strengthen the arguments made in the research. The manifested codes are derived from the theoretical framework. The latent coding will be used to identify important reoccurring themes which not have been investigated by the manifested codes.
To analysis the data, the qualitative software atlas-ti will be used. Atlas-ti enables the simple selection of indicators, the tracing of codes as well as the creation of linkages. The data sampling plan of the content analysis is based on random sampling. The websites are the forums of communication between the authority and the audience. The authority publishes its expertise on the websites as well as the missions of the institutional actions. Resultingly, all information from the website regarding the unit of analysis can be seen as useful. And finally, in order to increase replicability, the data collection will follow a strict data management plan. As the websites transform during time, the data of the content analysis will be restored (see appendix).
B) Statistical analysis
The content analysis is only able to make the mission of the think tanks visible via the analysis of semantic fields. In order to understand how the expertise production is executed, a statistical analysis of the institutional actions is needed. Therefore, three statistical analyses are conducted which measure the expertise production. In the first statistical analysis, the usage of the manifested codes is measured. For instance, the three indicators of throughput legitimacy will be compared by illustrating how often they were coded in the data. The comparison of used codes enables an understanding of the prioritization of aspects of the expertise production. The second statistical analysis will compare the topics of intermediation events in the time period of one year. For instance, the intermediation events of the Hans-Seidel Stiftung will be categorized into the topics and target-groups. Thereby, it will help to understand how responsive the events are in reality. Similarly, the third statistical analysis will compare the publications in a time frame of one month. This analysis should compare the responsiveness to different types of audience, especially to the national and international audience types. Moreover, the responsiveness to recent social problems will be analysed through a focus on
24
publications that address the COVID-19 virus. To validly compare the results, all statistical analysis will be calculated as a percentage of the total N of each case.
The final, and most important, statistical analysis is linking the used codes to the audience-driven legitimacy. It will do so by calculating a percentage of the of the frequency of the linkages between each concept. Only codes are used in this calculation which have a direct link to effectiveness and participation. Meaning, that the codes need to mention participation indicators such as audience demands and effectiveness indicators such as problem-solving capacities in order to qualify for the consideration. Hence, not the total N of codes which were depicted in the other three statistical calculations will be used as not all have a direct linkage to the audience-driven legitimacy indicators.
The data selection plan for the statistical analyses are limited on specific data of the websites. As the one of the main functions of think tanks is to publish information, an amplitude of data can be found on the websites. However, as the research has the aim to analyse the political education of the think tanks, only publications of the specific departments will be used. In addition, specific time frames will be selected as mentioned above. Moreover, to avoid the selectivity of the researcher, only publications and intermediation events which were automatically linked to political education by the filter option ‘politische Bildung’ of the websites will be selected.
C) Survey research
The last part of the research design will gather data from survey research. The content analysis and statistical analysis of the website will not be able to answer all the questions. Especially to test H3, another method is needed as content analysis is not able to “reveal the intentions of those who created the text or the effects that messages in the text have on those who receive them” (Neuman 2014, p. 378). Content analysis is not able to make inferences about the instrumentalization of the expertise production by the authority and the audience-driven side of legitimacy. Therefore, to add more depth to the data collection, qualitative surveys will complement the results. The qualitative survey which will be answered by the directors of the department of political education. In the following, the questions of the survey research will be described following the structure of the theoretical framework.
The audience-driven side of legitimacy will not be asked in distinct questions. Rather it will be analysed by follow-up question which follow each main question of expertise production. In these follow-up questions the directors will be asked to reflect on the effect of the expertise production on the audience. The question will only follow up the questions regarding
25
throughput legitimacy and the non-instrumentalized expertise production. This method is based on the already theoretical developed linkages between legitimacy and the instrumentalized expertise production. Therefore, following question will be asked after Q3 -Q8:
FQ: How does the dynamic of expertise production influence in your opinion the target’s group perception of the effectiveness and the democratic character of the think tank?
Moreover, two questions will be asked which clarify the audience of the think tank. The audience could consist of citizen but also other actors with institutional authority. Especially, in regard to the mother parties it is not fully clear which relationship exists between the two actors (Heisterkamp 2014, p. 467). It is not clear if the think tanks identify as authority or if they primarily serve the mother party as they are the proximate decision makers. Therefore, following two questions will be asked:
Q1: What community or actors does the think tank serve? Who is the think tank accountable to? Is the think tank a part of a collective beliefs system?
Q2: Does the think tank primarily serve the political party or is it an authority in itself? After the clarification of the audience, the following question will aim to inquire the dynamics of expertise production. Firstly, the process of expertise production will be questioned. By asking the interviewees about the process underlying the expertise production, the research can analyse to which degree the audience-driven side of legitimacy is taken into consideration in the process. Hence, the three indicators of throughput legitimacy will be addressed. Therefore, following question will be asked:
Q3: Does the think consider accountability frameworks in the production of expertise or has created enforced output controls?
Q4: Has the think tank developed institutional channels which intermediate the interest and demands of the audience in a balanced way?
Q5: Does the think tank make undertake measures to make the process of expertise production transparent for the audience?
To test H2, the operationalized concepts of the non-instrumentalized expertise production need to be examined in the survey research. Therefore, the concepts of the problem-solving model, the knowledge-driven model and the enlightenment model will be tested. Therefore, following three questions will be asked:
26
Q6: Does the think produce knowledge without the clear aim of entering the policy cycle?
Q7: Does the expertise production respond to social problems and does the think tank keep track of social value shifts?
Q8: Does the think tank have the goal to enlighten the audience through expertise production in the long-term?
The last part of the survey is going to test the use of the tactical and political model in the expertise production. The instrumentalization of expertise is the most sensitive aspect of the research because it directly questions the active behaviour of legitimization of the institution. Especially, the tactical model neglects the core functioning of the think tank, because it questions if the existence of those kind of institutions has a bigger effect than the actual produced content. Therefore, the following two questions are purposefully asked in the end:
Q9: Do you think that the think tank or the mother party uses expertise production of the think tank politically? (For example, to expand the institutional capacities in other areas, depoliticize issues of the mother party, or create constant institutional support?)
Q10: Do you think that the think tank or mother party uses the sheer fact that expertise is produced tactically? (For example, to prioritize the demands of the audience, to justify policy outcomes, or postpone decisions?)
3.3.Operationalization
After the complex methodology has been explained, this section is going to present the operationalization of the theoretical framework. The operationalization will follow the coding system of the content analysis. Thereby, it will make the method of content analysis clearer. To structure the operationalization, first concepts regarding the audience, input and output legitimacy will be operationalized. Afterwards, the concepts of throughput legitimacy and the non-instrumentalized expertise production will be operationalized. As described above, the concepts of H3 will not follow the same operationalization based on a different research approach.
Audience
Before the production of expertise can be questioned as source of legitimacy, the research needs to clarify for who outcomes are produced for. Through the lack of research on political party
27
Table 2: Coding System
Aim Underlying
Theory
Manifested Codes Possible indicators
Clarification of Audience Collective belief system (Zaum
2016)
Audience Target groups
Participants Actors in Network
Linking audience-driven legitimacy
Input legitimacy Participation Audience’s demands Audience’s interests Output legitimacy Effectiveness Problem-solving capacities
Enforcement mechanisms Test H1: The lack of
throughput-oriented expertise production decreases input and output legitimacy of party affiliated think
tanks.
Throughput-Legitimacy (Schmidt 2013)
Accountability Accountability frameworks Output controls Feedback mechanisms Mission and espoused values
Intermediation Events
Seminars Workgroups Social Media Mission and espoused values Transparency Open evaluations
Performance reports Open decision-making Mission, values and goals Test H2: Political
Party affiliated think tank adjust non-instrumentalised forms of expertise production
to the audience to increase input and output legitimacy.
Problem-Solving Model (Weiss
1979)
Responsiveness Focus on social problems Acute focal points Response to COVID-19 Mission and espoused values
Knowledge-Driven Model (Weiss
1979)
Impartiality Party influence in expertise production
Recruitment Finances
Mission and espoused values Enlightenment
Model (Weiss 1979)
Enlightenment Long term horizon Short term horizon Mission and espoused values
28
affiliated think tanks it is not fully clear which actors can be considered as the audience. In a theoretical perspective, the question of the audience is crucial as legitimacy can only be assessed by a particular represented group Therefore, think tanks need to be aware of their audience to claim legitimacy. The thesis will look for actor groups according to age, nationality, employment and political affiliation. The research aims to examine if the authority sees the entire population as the audience or only the political party affiliated voters. In addition, it also investigates which relationship the think tanks have to other authorities. It will analyse if the think tanks categorize other authorities such as the mother party as their audience. Therefore, the manifested code audience will make the comparison of the audience possible. It will make clear if all think tanks consider the same actors as their targeted audience.
Audience-driven legitimacy
In order to analyse the audience-driven legitimacy, indicators for input and output legitimacy need to be developed. Input legitimacy was defined in the theoretical framework as the degree to which the audience perceives the prioritization of their demands and interests through institutional arrangements (Oates 2016; Steffek 2015, Schmidt 2013). Concepts which are linked to input legitimacy are democracy, participation and the intermediation of articulated interests and demands. Input legitimacy can be indicated through the demands and interests of the audience. In the content analysis, the coding of these indicators will make it visible how they are linked to other codes such as intermediation or responsiveness.
Output legitimacy was defined in the theoretical framework as the ability of a think tank to effectively produce outcomes for the audience. Concepts linked to output legitimacy are institutional effectiveness, problem-solving capacities and enforcement mechanism. Therefore, input legitimacy can be indicated through the creation of output controls which include the interests and demands of the audience. The coding of output legitimacy makes it also visible how often they are linked to other codes such as transparency or accountability. Creating those linkages of codes will enable the indirect measurement of the audience-driven legitimacy. Throughput legitimacy
Throughput legitimacy emphasizes the process which transforms the demands of the audience into effective outputs (Schmidt 2013, p. 5). Throughput legitimacy will be operationalized according to its three underlying concepts of which are accountability, transparency and inclusiveness. Accountability refers to the need of institutions to be responsive and responsible to the audience that participates in decision-making processes. In the website analysis accountability can be indicated by the enforcement of accountability frameworks and output
29
controls. In addition, accountability can be indicated by the espoused values which prioritize accountable structures in the process of expertise production.
Inclusiveness refers to the balanced intermediation of interests through the creation of internal channels of communication. Those channels transform the articulated interest and demands of the audience into effective outputs which respond to the input. It is not only the question if the interested are intermediated but also if the intermediation follows equity principles. Therefore, inclusiveness can be indicated through the creation of channels which incorporate inputs in a balanced matter. In the website analysis possible indicators of intermediation could be events, seminars, workgroups or social media usage. Moreover, the mission statement and espoused values of intermediation will be also analysed.
Transparency is a crucial condition in the debate of political party affiliated think tanks in Germany. According to Schmidt (2013) transparency concerns the openness of information of the process as well as the decision-making. In the research, transparency emphasises the availability of information of the processes more than the openness of the final outcomes. The transparency of the final outcomes, the publications, are an obvious function of the think tanks. Therefore, the question of transparency in the research regards if its accessible why and how certain information is produced. Therefore, transparency can be indicated by measures which are created to make information accessible for the audience. In the website analysis, indicators of transparency could be derived from open evaluations, performance reports, and channels that inform about the openness of decision-making. As in the other concepts of throughput legitimacy, the mission statement as well as the espoused values are helpful sources in analysis of accountability.
Authority-driven legitimacy
To test H2, concepts of the non-instrumentalized expertise production need to be operationalized. The first concept of non-instrumentalised expertise production derives from the problem-solving model. It refers to the production of information which responds to particular social problems of the audience. Therefore, problem-solving expertise can be measured through the responsiveness of the think tank which will also function as code in the content analysis. The clarification of the audience is especially needed in this concept. To measure responsiveness, it needs to be clear who the receiving end of the expertise production is. In the website analysis, responsiveness can be indicated to aspects that relate to the focus on social problems of the audience. The example of COVID-19 will be taken to analyse how the think tanks has changed their expertise production to respond to the needs of the audience.
