The more we verify it , the more truth there is ’ : Producing the truth and expertise in public policy

Between 2005 and 2006, the Czech government imposed a forced administration on the General Health Insurance Fund (VZP) because of economic problems. This unprecedented step in Czech public policy resulted in a debate. The VZP is the largest insurance fund in the Czech Republic, and this case has been one of the most severe institutional crises in Czech health policy. Employing analysis of 38 television-hosted debates, we emphasise how the actors negotiated ‘The more we verify it, the more truth there is’: Producing the truth and exp... International Review of Public Policy, 3.3 | 2021 18


Introduction 1
Our paper focuses on ways in which factuality and truth are negotiated, assessed, and criticised in public policy discussions. We intend to introduce discursive psychology as a methodological tool to understand forms of policy discourse and demonstrate its promise in the study of policy crisis management. In the literature, crises are usually considered as 'path-shaping moments' of change (Hay, 2013), signifiers of change (Roitman, 2014), moments in which actors' perception of their self-interest becomes problematised (Blyth, 2002) or 'episodic breakdowns of familiar symbolic frameworks that legitimate the pre-existing socio-political order' (Boin et al., 2008: 3). In the context of critical moments in public policy, expertise is being debated, and the boundaries of epistemic authority are drawn (Eyal, 2019). The policy crisis seems to be an excellent example of how epistemic authorities are constructed in public dialogue.
(approx. 403 million dollars), and due to poor cash flow, the payments for treatments were delayed for up to two months. The problems with its cash flow concerned most of the public. Moreover, physicians went several times on strike because of delayed payments, which caused financial troubles (in some cases, they were even facing bankruptcy and were forced to take a bridging loan). For a couple of months, the health care crisis was making daily news headlines and was an issue of television debates. 3 These TV debates represent the primary data source for our analysis. We analysed 27 debates broadcast on public and private TV channels from September 2005 to May 2006. We examine how different actors established their expertise and backed their claims in the discussion. In the theoretical section, we draw on this paper's epistemological and ontological positioning. Principles of discursive psychology and sociology of expertise are introduced. The paper deals with the role of the institutional context of debates on actors' positionings, factual accounts in public discussion, and their policy consequences. Finally, it introduces the strengths and weaknesses of discursive psychology for understanding recent public policy debates. We show how different realms of expertise produce different corroboration and other factual accounts, such as data evidence, accounting for impartiality, building reputation, and mobilising biographical experience and self-reflection. The next section describes our methodology and data corpus. The empirical section explores the story behind the case study analysed, i.e., how the financial crisis in VZP evolved, and sums up the key strategies of fact construction and expertise. In conclusion, we show how different kinds of expertise use different corroboration and factual evaluation vehicles -from data evidence to impartiality and reputation, to biographical experience and reflection. 4 As epistemic authority rests on the successful mobilisation of norms, classifications, methodologies, technologies, or rhetoric to establish or maintain the boundaries between accepted knowledge claims (Straßheim, 2017), the paper stems from the argumentative and rhetorical approaches associated with discursive psychology and employs its long-term concerns with fact construction and arguing as a situated practice (Billig, 1991;Billig et al., 1988;Potter, 1996;Edwards & Potter, 1992;Edwards, 1997;. The paper illustrates the benefits of using discourse psychology in studying policy debates. It emphasises the action-oriented nature of expertise and the importance of viewing facts as a rhetorical resource (Demasi, 2020). The application of discursive psychology makes it possible to identify rhetorical resources for constructing expertise and to study how claims are presented as factual in public dialogue. We believe that this approach can provide important lessons for studying public policy in the post-truth era.

Expertise in Discursive Practice 5
The dynamics of argumentation have already been recognised in policy analysis (Fischer & Forester, 1993;Howarth & Griggs, 2004;Fischer & Gottweis, 2013) as an effort to inform the ordinary-language processes of policy argumentation, as reflected in the talk and deliberation of politicians, policy experts and citizens (Fischer & Gottweis, 2013, p. 489). Discursive approaches to argumentation have allowed researchers to study how the articulation of meaning is associated with political dynamics including the production and normalisation of specific truths, discourses' competition for power, and discursive practices (Behagel et al., 2019). In line with these approaches, discursive 'The more we verify it, the more truth there is': Producing the truth and exp...
International Review of Public Policy, 3.3 | 2021 psychology examines how speakers construct their representation of realities (Taylor, 2001) and observes that truth is something people try to accomplish through talk rather than an objective reality.

6
There is a growing interest in epistemic authority in discursive psychology (Raymond & Heritage, 2006;Heritage, 2012;Stevanovic & Perakyla, 2012). Concerning expertise, there is a relatively long history of the concepts of epistemic and epistemic rights (Heritage, 2013, Raymond & Heritage, 2006, Terasaki, 2004. Epistemic rights refer to how speakers negotiate and manage their knowledge claims in talk (Heritage, 2013). As Stevanovic and Perakylä (2012, p. 297) point out, epistemic authority is the right to know how the world is. There are direct links between participants' identities and the rights and responsibilities associated with the identities directly implicated in speaking practices (Raymond & Heritage, 2006). Members of groups might own different types of knowledge, have privileged access to their experience and use specific terms to account for their epistemic stance and the facticity of their utterances. Heritage (2012) distinguishes between epistemic status and epistemic stance. Epistemic status is the positioning of participants concerning their knowledge distribution and access to a given epistemic domain' (Mondada, 2013, pp. 599-600). Epistemic statuses can be associated with types of expertise. On the contrary, epistemic stances mark the particular epistemic quality of some knowledge held by the speaker and how the speaker knows something in terms of certainty or the source of such knowledge.

8
For classification of expertise, we follow Collins and Evans (2002), who propose distinguishing between the wider public, scientific experts and experience-based experts. They use the term experience-based experts to refer to public members with special technical expertise stemming from experience not recognised in degrees or other formal qualifications. This expertise is divided into two classes: contributory expertise and interactional expertise. Contributory expertise is about possessing sufficient expertise to contribute to the science of the field being analysed. Interactional expertise is the expertise to interact constructively with participants. Interactional expertise helps to engage experience-based knowledge with scientific research. Scientists need interactional expertise to absorb non-scientific expertise to produce an optimum outcome. Regardless of the ambiguous boundaries between the categories of expertise mentioned by Collins and Evans (see, e.g., Jasanoff, 2003), this distinction provides an important empirical tool for contextualising some aspects of knowledge work. In terms of public policy, we suggest modifying the definition of 'scientific experts' to grasp the institutional dimension of expertise. Instead of scientific expertise derived from a degree or formal qualification, we prefer to speak about institutionally recognised expertise derived from the institution's responsibility for the area of knowledge and identified knowledge resources to deal with a problem.

Based
Expertise stemming from experience not recognised in the form of degrees or other formal qualifications and which contributes to the knowledge of the field being analysed Expertise stemming from experience so as to interact constructively with other participants in the policy process Source: Collins & Evans, 2002 9 In the analytical section, we examine four types of expertise: (1) institutionally recognised contributory expertise; (2) experience-based contributory expertise; (3) institutionally recognised interactional expertise; and (4) experience-based interactional expertise. We are interested in epistemic stances in each type of expertise. Different kinds of expertise use different vehicles of corroboration and factual evaluation. We study how the parties mobilise different types of expertise and the consequences of using each type of expertise.

Data and Methods
10 In contrast to predominantly textual tradition in public policy (Fischer & Forester, 1993;Hajer & Wagenaar, 2003;Wagenaar, 2014;Yanow, 2007), discursive psychology brings our attention to discursive as an interactional practice. Its research methodology allows the researcher to link macro-social processes to micro discursive events. It pays attention to the terms used by participants. Finally, it looks at a policy debate as the product of conversational procedures and rules. First, it offers an analytical approach to rhetoric and argumentation that examines and evaluates arguments in real interactions (Potter & Wigins, 2007, p. 73). It might help us, as policy analysts, to identify which ideas or forms of communication work and what effects they produce.
11 In contrast to some paradigms of interpretive policy analysis, such as frame analysis (Rein & Schön, 1993;Snow, 2004), discourse analysis (Hajer, 1997) or narrative policy framework (Shanahan et al., 2011), discursive psychology is focused less on the content of policy talks in favour of studying their forms. Second, discursive psychology pays attention to the terms participants use in interactions. In contrast to critical analysts, who impose their terms on the object of analysis, discursive psychology takes 'seriously the object of inquiry on its own terms' (Schegloff, 1997, p. 171) and the importance of how people categorise their social world and what version of the social world they construct (Kitzinger & Mandelbaum, 2013;Stokoe, 2010;Stokoe et al., 2017;Whitehead & Lerner, 2009). The emphasis on categories provides a terminological apparatus that can inspire category analysis in public policy (Yanow, 2000) and broaden the scope of policy-relevant situations in which categories play an important role. Third, discursive psychology focuses on talks as the product of formal procedures and rules that can be identified and described. These procedural rules might be heterogeneous, and different  (Chafe & Nichols, 1986;Drew, 1991;Kamio, 1997;Stivers, 2005;Willett, 1988). 12 We captured the discourse through a retrospective narrative analysis of 27 debates broadcast on public and private TV channels from September 2005 through May 2006. The debates' transcriptions were collected from the Newton Media. The identified transcripts featured two or three leading political figures reflecting on the main events of the previous weeks. The transcriptions captured the period from the first mentioning and threatening of forced administration through to the actual announcement, the subsequent quarrels about the justification for the unprecedented act, and retrospective evaluation.
13 Television debates provide an ideal unit of analysis because they require politicians to explain their steps while facing their political opponents and being watched by their voters. "The televised debate can be seen as a minor device of government, a mode of ruling through showing how we should think and act under certain norms of civility" (Chouliaraki, 2005, p. 277). In 'confrontational' (Hutchby, 1996) or 'adversarial' discourse (Bull & Wells, 2012), conversation analysis helps us to grasp a dynamic of conflictual views and negotiation about what is taken as truth in public policy in the context of the spectacle of confrontation (see, e.g., Hutchby, 1996). As Demasi (2016) argues, the specific focus in work on adversarial discourse is more on substantive issues, face-saving strategies (Bull & Fetzer, 2010), institutional constraints (Robles, 2011) or oppositional moves (Hutchby, 2006) than on how politicians manage their epistemic authority, or how, through mobilising 'facts' and 'knowledge', they establish their argumentative positions.
14 The data were coded and further analysed using the Atlas-ti programme. We focused on different forms of expertise and valuations of truth claims in the analysis. We kept our quotations in their dialogical context in the coding system to look at them as part of a broader question-answer structure. In the first round of coding, we used an open coding system. We make use of what has been called 'discursive devices', such as extreme case formulation, affect display, consensus or corroboration, metaphors, reported speech, category entitlement, type of knowledge, speakers' identities, right to talk, or model verbs. In the second round of coding, we try to sort out these codes along lines defined by Collins and Evans's types of expertise -contributing, interactional, institutionally recognised and experience-based.
Crisis in the General Health Insurance Fund: The Story of the Case 15 In 2005, the health care crisis was making daily headlines in the Czech Republic for a couple of months. VZP, the largest insurance company in the country, provides health insurance for nearly two-thirds of the Czech population and therefore the problems with its cash flow concerned most of the public. Due to insufficient cash flow, payments for treatments were delayed for up to approximately two months. During the crisis, health system-related issues became a major topic of political discussions, and all relevant actors demanded changes to fix the situation. Discussions delineated a new political battlefield and all leading political figures were taking stances on the topic. Physicians went on strike several times because of delayed payments, which caused 'The more we verify it, the more truth there is': Producing the truth and exp...
International Review of Public Policy, 3.3 | 2021 financial trouble; in some cases, they were even facing bankruptcy and were forced to take bridging loans. It was unclear who was to blame for the situation within the complex healthcare system. 16 The ruling coalition of Social Democrats (left-wing), Christian Democrats (centre), and the Union of Freedom (liberal) faced rising criticism over the situation because the government regulates most of the VZP's income and expenditure. First, the state was partly responsible for the VZP's income because of the state's contributions to people (58% of the population) insured by the state (e.g., students, pensioners, unemployed, etc.). The Czech government regulates the state's contribution. The second reason was ministerial decrees, which enumerated a sum of treatments that had to be paid for by the VZP, so the VZP had limited influence on its payments (the government regulated 70% of all transactions). The crisis led to the forced administration of the public General Health Insurance Fund. According to Czech law, the ministry can place a health insurance fund under forced administration in cases of poor economic performance, if a fund is in severe debt or cannot meet its liabilities, or because of failure to comply with the public interest. This forced administration aimed to stabilise finances and carry out an in-depth audit of financial flows. It was a form of crisis management in which the state significantly limited the fund's autonomy. The management was replaced with a forced administrator who took control over the fund whose activities were governed and monitored directly by the government. Even if the Act enabled the government to announce this forced administration, it would use this tool for the first and last time in 2005. Normal policy-making 18 In normal policy-making, the VZP's situation was defined as a problem that needed solving; however, the solution was supposed to be derived from standard policy According to him, the crisis was caused by the VZP's management staff, who should be replaced as soon as possible. He even refused to go to the Senate, when senators invited him to justify the forced administration; he used the argument that there was nothing to explain, that the health crisis was not a political issue, and that there was no time to talk. He refused to be dragged into the distracting 'political games', insisting on instant action and crisis management.
21 Rath mobilised the language of crisis management to support his policy. The crisis was presented as an emergency, and such emergencies demanded a 'fast policy'. Depoliticising the issue was an essential part of this rapid policy-making. When the forced administration was put into practice, and the government gained access to key VZP documents, Rath announced discrepancies in the contracts with the VZP, accused the opposition leader of corruption and cut off the unjustified expenditure (

Normalisation 23
Almost a year after the announcement of forced administration, VZP's total debt was reduced from 12 to five billion, partly because of injections from the state budget. The government poured almost 6.5 billion crowns into the VZP. The Parliamentary 'The more we verify it, the more truth there is': Producing the truth and exp...

International Review of Public Policy, 3.3 | 2021
Commission stated that the health insurance system could be blamed on VZP's debt. Jiřina Musílkova, the former director of the VZP, acted as an independent counsellor in the health service. She was convinced there had been no reason to appeal to her. On the other hand, Rath insisted that there was a reason since Musílkova had not cooperated in stabilising the VZP (the Ministry of Health, David Rath, 9th April 2006). Pavel Horak, the new head of the insurance company, pointed out in a television debate that 'it is possible to maintain a balanced economy. However, you must know that you will receive adequate consideration for the money invested, but past management did not check it properly.' (director of the VZP, Pavel Horak, 30th April 2006).

Four types of expertise and their arguments 24
In the next section, we examine four types of expertise described in the theoretical background: (1) institutionally recognised contributory expertise; (2) experience-based contributory expertise; (3) institutionally recognised interactional expertise; and (4) experience-based interactional expertise.

Institutionally Recognised Contributory Expertise
25 Expertise is often associated with contributory expertise based on institutional knowledge resources that produce legitimate knowledge of the field being analysed. In our context, contributory expertise was heavily mobilised by the representatives of health insurance funds, who posited themselves as having more expertise than interest groups and politicians to justify their position as being based on facts rather than on ideological standpoints. Contributory expertise is based on demonstrating ownership of particular knowledge and knowledge superiority is a means of establishing an appropriate measure of relevance. It does not refer to someone's epistemic status only (what they are entitled to know) but also to their epistemic stance (the extent to which they position themselves as knowing more than others). In this example, health insurance funds justify their truth through the superiority of their data over others. They based this on the scope and completeness of their data being essential characteristics. However, at the same time, they also externalised the truth and referred to the epistemic status of others, mobilising a broad network of epistemic resources to justify their stance.
Jiřina MUSÍLKOVÁ, director of the General Health Insurance Fund: These predictions are derived from the macroeconomic number the Ministry of Finance gives us. We rely on real numbers. (Otázky Václava Moravce, the Czech TV, 5 September 2005) 27 The speaker whom the government had criticised defended her position on governmental knowledge. Diversifying a pool of knowledge resources and transgressing boundaries between conflicting fields helped to establish a durable network of arguments; this contributed to 'a stake inoculation' (Potter, 1996, p. 125) as a precondition of facticity 'where descriptions are constructed to head off the 'The more we verify it, the more truth there is': Producing the truth and exp...
International Review of Public Policy, 3.3 | 2021 imputation of stake or interest'. In our example, stake inoculation is used to describe representatives of health insurance funds referring to their opponents' data. As we can see, aside from quantification, there is a space for contextualisation: a proper understanding of the context behind numbers. Accounting for context, the prime minister demonstrated his familiarity with the field by showing that he knows what every expert knows; he also challenged the credibility of his opponent's knowledge. When someone accounts for contributory knowledge, he or she must be able to place his or her expertise in a more considerable discipline.
Experience-based contributory expertise 34 Experience-based expertise was predominantly associated with professional medical voices. Elite doctors were regularly invited to debates, and their experiences counted as valuable. One of them was introduced by the TV moderator, Václav Moravec, in the following way: He treated the Czech president. He is a medical authority to the majority of the public. He has been involved in healthcare for long enough to describe the main problems of Czech healthcare and to draw up their solutions. These are the opinions of Professor Pavel Pafko. (5 September 2005) 35 When someone's personal experience is culturally recognised as valid, he or she does not need to have numbers, graphs or specific knowledge to gain legitimacy for his or her claims. In this case, personal experience accounts for the speaker's credibility in commenting on issues that significantly transcended his medical field. His medical expertise allowed him to make ideological statements taken seriously and without discussion. He was the first guest in the discussion, with no opponent.
Professor Pavel PAFKO: Czech society cannot generate financial sources, and political representation must say whether the state puts additional money into the system unsystematically or whether it makes cuts. The second problem is that Czech citizens overuse healthcare. … When I discussed this with my colleague from England, he explained that people who have money to corrupt doctors do not do it because they can buy private care. The problem with Czech healthcare is that people who own financial sources cannot spend them legally in health care facilities. (Otázky Václava Moravce, the Czech TV, 5 September 2005) medical chamber, David Rath, did this when he criticised the financial results of hospitals run by regional governments led by the opposition party.
David RATH, minister of health: My colleagues, who work there, told me that it was the first time in our lives, when the regional government led us, that our wages decreased. They have been working in healthcare for more than 30 years. They also told me that it was the plundering of the healthcare system that was going on there. (Sedmička, TV NOVA, 30 October 2005) 37 Referring to experience-based expertise about truth claims might challenge contributory expertise. Experience-based expertise might further contextualise graphs and numbers and reveal what is behind them.
38 On the other hand, it relies heavily on pre-existing cultural patterns and on policy actors' established positions. The experienced-based expertise of other actors, such as nurses or patients, was missing. There were no direct voices of nurses and patients, and speakers did not refer to their expertise, as Rath did about doctors from regional hospitals. The Czech discussion on health policy proved the cultural significance of the medical profession, which penetrates the entire Western world (see, e.g., Lupton, 2003). However, TV discussions reinforced their positions and contributed to the further sedimentation of this cultural pattern.

Institutionally Recognised Interactional Expertise
39 During the first period of the crisis, the minister did not directly oppose the statements of General Insurance Fund representatives, nor those of the representatives of medical professionals. Instead, she pinpointed the partiality of their contributory knowledge and its shallowness and insisted on the need for a deeper analysis of the situation. The prime minister, Milada Emmerová, did not posit the Ministry itself as an owner of knowledge; however, she insisted that the Ministry was a neutral institution that could guarantee to get appropriate knowledge, as an honest broker (see, e.g., Pielke, 2007). The Ministry did not hold the knowledge itself; however, it had privileged access for seeking the truth.
Milada EMMEROVÁ (ME), the Ministry of Health: To do some deep analysis, for example, and not only the study of the controls in hospitals that are directly managed by the state, it is also necessary to focus on the management of insurance companies. (Otázky Václava Moravce, the Czech TV,5 September 2005) 40 While the General Health Insurance Fund and representatives of medical professionals argued that the crises were caused by the imbalance between their actual expenses and the funds that the state contributed, the minister of health called for a more sophisticated view on costs and expenditure. This was part of the construction as an impartial actor. In this case, the rhetorical struggle over facticity was not just a binary conflict between truth and fiction; it was also a struggle over the scope of the fact and the relevance of the epistemic stance. The institutionally recognised expertise was based on procedures for knowledge generation rather than on a particular knowledge owner, proposing that seeking the truth is a never-ending process.
Experience-Based Interactional Expertise 44 In contrast to institutionally guaranteed interactional expertise, the knowledge in experience-based interactional expertise is derived from people's knowledge and reflexivity. The language could be described in terms of reputation, trust or personal inclination. Identity, especially its reputational facet, is a valued asset. Historical analogy is a vital tool for this fact-making. It accounts for the field's history and practical knowledge. It is based on past cases, which are examples of recent characteristics. Vlastimil Tlustý, the caucus chairman of the ODS, used this strategy when he denounced the Social Democratic Party. It is based on mobilising the past to account for the facticity of recent judgements.
Vlastimil TLUSTÝ (VT), the Caucus Chairman of the ODS: As I see it, the Social Democratic Party is using the same strategy as was used for the privatisation of IPB Bank. They spoke about the Sicilian mafia and how they would save a broken bank. And in the end, it turned out that savers of the bank picked up more money than the so-called robbers did. ( Conclusions 47 Expertise matters in political debates; however, there are different ways in which this expertise is accounted for and justified (see Table 2). Different types of expertise are associated with different kinds of actors: Institutionally recognised contributory expertise relates to external and impartial experts; institutionally recognised interactional expertise is associated with policy-makers in regular periods; experiencebased contributory expertise is associated with people with special epistemic status, which guarantees them specific access to the field (medical doctors in our case represent them); and experience-based interactive expertise relates to the historical memory of policy-makers. Furthermore, different types of expertise use different vocabulary. Contributory expertise based on recognised specific knowledge and access to evidence is legitimised by five key strategies: (1) demonstration of knowledge ownership; (2) externalisation and diversity of knowledge resources; (3) visualisation; (4) quantification; and (5) further contextualisation. Externalising and materialising knowledge seem crucial in the production of credibility and factuality. Contributory expertise is often connected to quantifying and to putting data into the proper context. Our analysis shows that contributory expertise is based on generalisation with little space for detail and everyday practice. However, there are different ways in which these truths can be pronounced. Contributory experts must distinguish their evidence from their interests. If they fail to do this, their expertise will also fail to be acknowledged. On the other hand, interactional expertise can be used as a counterstrategy. In the realm of institutionally recognised interactional expertise, the policy actor admits a lack of knowledge, which might be a precondition for unbiased judgement. The interactional expertise is associated with a dominant legitimation strategy of neutrality and impartiality and privileged access to the policy analysis. Whereas contributory expertise is always directed towards content knowledge, institutionally recognised interactional expertise is based on institutional arrangements, rights, duties, responsibilities, or access to evidence.
50 Experience-based contributory expertise is founded on broader cultural patterns of professional authority. The general cultural practices create positions filled by people who are generally knowledgeable and can comment on issues largely transcending their field of knowledge. Someone's personal experience is culturally recognised as valid and relevant, and it relies heavily on pre-existing cultural patterns and policy actors' established positions. 51 On the other hand, experience-based interactive expertise is based on memory, and people's knowledge and establishing facticity leans on three predominant strategies to (1) make historical analogies; (2) externalise judgement of others; and (3) biographisizing. This type of expertise is based on past cases, and examples of recent characteristics.
52 Our case shows that there is more than one form of valid expertise in policy controversies; different types of expertise compete against each other in the public space. Rejecting simplistic notions, we opt for a more nuanced approach to expertise that pays attention to how the actors involved argue, and to the relevant rhetorical strategies in different contexts.
accountability, credibility and truth. This paper aims to identify the discursive and rhetorical resources used by the involved parties to manage what was at stake in the debate: to maintain their credibility and legitimacy and to manage factuality. We follow Collins and Evans's (2002) approach to expertise and distinguish four types of expertise: contributory, interaction, institutionally recognised, and experience-based. These different kinds of expertise use other vehicles of corroboration and factual evaluation. We opt for a more nuanced approach to how involved actors argue and what rhetorical strategies are relevant in different contexts.