Examining Legitimacy in Government Agencies’ Crisis Communication

ABSTRACT

Governments exercise power over the public in many ways. One clear example is during societal crises when government agencies engage in purposeful strategic communication, issue restrictions and recommendations, and rely on people’s compliance to reach response objectives and meet needs in society. Here, issues of authority and legitimacy are crucial. In this study, we are interested in how this power is configured, and we analyze different dimensions of legitimacy in the strategic communication employed by government agencies during the Covid-19 pandemic. Drawing on empirical data from Q&A sessions of daily press briefings, we examine how the pragmatic, moral, and cognitive legitimacy was challenged and defended in the interactive process between government agencies and journalists. Our analysis shows how all three dimensions of legitimacy co-exist and reinforce one another. However, considerable weight can be ascribed to the pragmatic dimension in this process of legitimation, in which government agencies rely on knowledge-based descriptions (there is) combined with imperatives regarding what needs to be done (we must). This empirical contribution improves our understanding of both the legitimacy concept and how strategic communication is central to the exercise of contemporary government power, which relies on processes of legitimation.

Introduction

Every day, government agencies exercise power in a multitude of ways in their efforts to influence the population’s behavior to achieve desirable outcomes. During crises, when societal order and security are threatened, this power relation can become evident as government agencies publicly urge people to comply with their directives. While academia, following scholars like Michel Foucault or Pierre Bourdieu, has paid significant attention to understanding implicit or cryptic forms of power, this kind of power is “apparent” (Hearn, 2012, p. 21). In situations of high stakes and uncertainty, government agencies’ purposeful strategic communication, along with their ‘toolbox’ of warnings, restrictions, and recommendations, is of central importance in efforts to reach response objectives and meet needs in society. Therefore, societal crises – like a global pandemic – provide good opportunities for understanding how power is configured in a society by directing attention to how government agencies attempt to influence the public and how they continuously legitimize their positions, decisions, actions, and regulations in this process (Deephouse & Suchman, 2008; Hearn, 2011). In such contexts, strategic communication clearly is about power, dominance, and ethical choices, which is an issue that has often been neglected in traditional research on the subject (Falkheimer & Heide, 2018; Heide et al., 2018). This contribution seeks to address this gap.

Strategic communication can be viewed as a means for organizations to reach their objectives (Van Ruler, 2018), and the centrality of strategic communication in crisis management is often highlighted in research (Chandrasekar et al., 2021; Chen, 2009; Fredriksson, 2014; Merkelsen, 2013). Several studies also connect crisis communication to legitimacy (Bowen & Heath, 2007; Falkheimer, 2021; Merkelsen, 2013; Stark, 2010). Yet, while legitimacy has long been a central concept in organization studies (Deephouse & Suchman, 2008; Suchman, 1995),¹ several scholars agree that more research is needed to fully understand the role of legitimacy in crisis management (Christensen et al., 2019; Demiroz & Unlu, 2018). Christensen et al. (2016) emphasize how crises are managed in the interface between responding agencies and the public, where legitimacy is a key feature of that relation. In addition, they stress the need to examine how government agencies achieve and maintain legitimacy in crises since it is critical for governance capacity. Without legitimacy, agencies will struggle to gain compliance and acceptance of their interventions in people’s lives. Therefore, legitimacy becomes a bridging concept in efforts to understand strategic communication as an exercise of power, and the process of legitimation becomes essential to understand.

During the response to the Covid-19 pandemic, the reach of government power has been extensive, comprehensive, and intensive (cf., Wrong, 1980), with far-reaching consequences for personal freedoms. Power has been exercised throughout the societal system, by a multitude of actors, in ongoing processes of legitimation (Hearn, 2011). The main aim of this study is to give an empirical account of one such process; the Swedish strategic crisis communication conducted by national government agencies in response to the pandemic. In our approach, we have been less concerned with government capacity, questions of effective crisis management, or the roles and practices of “professional communicators” (Falkheimer & Heide, 2018, p. 74). Instead, we examine legitimacy by focusing on interaction where legitimacy is challenged, using journalists’ critical questions that seek to challenge legitimacy and government agencies’ answers that seek to defend it. We draw our empirical data from the question and answer (Q&A) sessions in the agencies’ public press briefings from March to June 2020. Legitimacy serves as our conceptual tool. To operationalize the concept of legitimacy and categorize the data, we apply three dimensions of legitimacy defined by Suchman (1995) and developed by Alexiou and Wiggins (2019).

Power, authority, and legitimacy

There is a ‘modern’ kind of governing that goes on in western, liberal, democratic societies (Hearn, 2012), where legitimate authorities seek to influence the attitudes and behavior of people by providing them with knowledge, recommendations, and good reasons to comply. This influence is achieved by processes of legitimation in dispersed power relations throughout the societal system (Hearn, 2011). The Swedish response to the Covid-19 pandemic is a good example of how modern institutions try to compel subjects to conform and strengthen their individual agency by claims of superior knowledge. It did not rely on enforcing strict rules, coercive sanctions, or lockdowns to control the spread of infection. Instead, the response has emphasized the individual’s autonomy, responsibility, and competence in the Swedish strategy to handle the pandemic.

Legitimacy scholars suggest that we can understand this modern form of power by investigating how the authority of government agencies is constituted and justified when they strive to achieve voluntary cooperation and community engagement (Tyler & Jackson, 2014) and how their authority is recognized by the public as legitimate (Boyd, 2000). Hence, key features of this modern power are the public acceptance of government agencies as legitimate and that their advice is considered worth conforming to. People are “influenced by others because they believe that the decisions made and rules enacted by others are in some way right or proper and ought to be followed” (Tyler, 2006, p. 376). Such influence can be based on different kinds of authority. In some cases, it is based on a formal, often legal, right to issue commands or operate in specific ways (Hearn, 2012; Wrong, 1980). In other situations, it relies on expert knowledge and epistemic authority (Gieryn, 1999; Heinzel & Liese, 2021; Lidskog et al., 2020; Wrong, 1980). This capacity of epistemic authority to exercise power relies heavily on knowledge-based descriptions of a situation or phenomenon (there is), combined with imperatives regarding what needs to be done (we must) (cf., Desrosières, 1998). Power, therefore, relies on a degree of acceptance of this rationale among the public. Hence, authority needs legitimacy (Hearn, 2012).

Legitimacy can be defined as a belief or acceptance that an institution has formal authority; a right to govern (Falkheimer, 2021; Tyler, 2006), which emanates from social roles and institutions (Tyler & Jackson, 2014). Hearn argues that legitimacy refers to how power and authority are reproduced in processes of legitimation (Hearn, 2011). Such processes rely heavily on communication between an organization and its various audiences, especially during crises when legitimation processes may be played out in the public forum (Falkheimer, 2021; Suchman, 1995). Indeed, crises are “dynamic forces in ongoing processes of legitimization, delegitimization, and relegitimization” (Hart & Boin, 2001, p. 31). Yet, the concept can be challenging to grasp. To investigate how legitimacy is constituted, we first need a better understanding of how it can be operationalized.

Operationalizing legitimacy

Legitimacy is a “pivotal but often confusing construct” (Suddaby et al., 2017, p. 451) in research on organizations and management. For academics, it is a multi-layered and diverse concept where scholars have identified over 18 types (Alexiou & Wiggins, 2019). In addition, legitimacy often becomes conflated or substituted with a range of other concepts, like trust (e.g., Bargain & Aminjonov, 2020; Bengtsson & Brommesson, 2022; Demiroz & Unlu, 2018), reputation (Coombs, 2007; Merkelsen, 2013), image and prestige (Falkheimer, 2021; Wæraas & Maor, 2014), or status (Deephouse & Suchman, 2008). Christensen et al. (2016) illustrate this conceptual dependence: “When government authorities enjoy the trust, confidence and satisfaction of other stakeholders and citizens, they generally have a good reputation and their legitimacy is supposedly high” (p. 889).

Tyler (2006) claims that government agencies find it easier to govern when they are perceived as legitimate authorities to control the population. Hence, legitimacy becomes an inherent capacity to govern. It functions almost as ‘oil in the machinery’ for governing processes when carried out without significant resistance from the governed subjects (Christensen et al., 2019). Importantly, we tend to take it for granted unless there is noticeable friction in a system or when authority is openly challenged. When things run smoothly, legitimacy disappears from view (Hearn, 2011, 2012).

Empirical studies usually only cover limited aspects of the phenomenon and from different epistemological perspectives (Suchman, 1995). In their comprehensive review, Suddaby et al. (2017) identified three such perspectives: legitimacy as a property, a process, or a perception. Many studies only study its effects, for example, peoples’ compliance to rules or decisions made by others, without attention to how it is achieved (Tyler, 2006). Our study aligns with the view that legitimacy can be examined as an interactive process of ongoing justification of decisions, actions, and regulations in specific situations (Hearn, 2011; cf., Zelditch, 2001). The co-construction of legitimacy takes place in an interactive, non-static process of communication and ongoing social negotiation between actors. Therefore, the unit of analysis is the process, not the outcome. The positioning and agency of the engaged actors are key features of this communication process of persuasion and influence (Suddaby et al., 2017).

Suchman (1995) suggests there are three main dimensions of legitimacy: cognitive legitimacy, pragmatic legitimacy, and moral legitimacy. These dimensions have been further operationalized by Alexiou and Wiggins (2019, p. 471). While Alexiou and Wiggins focus on psychometric measures of the perception of legitimacy, they claim that the dimensions could be useful also for comprehending legitimacy as a process. The three dimensions “capture the majority of the conceptual domain of the construct and represent the most frequently measured dimensions of legitimacy” (Alexiou & Wiggins, 2019, p. 472). Suchman (1995) strongly argues for such integrative efforts to keep the scholarly discourse of legitimacy coherent and accessible to practitioners. Hence, the dimension set provides a relevant framework to analyze the process in which the legitimacy of the agencies managing the Swedish response to the Covid-19 pandemic is challenged and defended.

The cognitive dimension of legitimacy rests on whether an authority is understood to have a comprehensible and formally acknowledged role. This is an implicit and very impactful source of legitimacy since challenges become impossible if alternatives become unthinkable (Suchman, 1995). It gives actors a natural or logical position of power, which is taken for granted, and seen as inevitable or necessary (Alexiou & Wiggins, 2019). In managing a crisis such as the Covid-19 pandemic, this can be exemplified by asking if the correct government agencies are handling the situation, if they provide essential functions, and if they have the right competence, policies, or mandates. Thus, it is reasonable to assume that a crisis response system needs to be accepted as an appropriate societal structure in a “shared cultural-cognitive schema” (Alexiou & Wiggins, 2019, p. 473) if it is to be viewed as legitimate. Importantly, the construction of the cognitive dimension rests on passive judgments of legitimacy, which are subtle but very powerful sources of legitimacy, as opposed to the more active judgments made for pragmatic and moral dimensions (Alexiou & Wiggins, 2019; Aldrich & Fiol, 1994).

The pragmatic dimension becomes visible when actors are perceived to be doing the right things and appear knowledgeable and beneficial to their stakeholders. Hence, it can be thought to build upon an expected value of organizations (Alexiou & Wiggins, 2019) and the practical consequences of their activities (Suchman, 1995). Similarly, Boyd (2000) contrasts what he calls actional legitimation, where organizations need to demonstrate the legitimacy of specific regulations or actions, with institutional legitimation, where organizations must defend their entire enterprise. For example, the pragmatic legitimacy of government agencies involved in response activities and policymaking during the pandemic depends on their ability to actively account for the benefits of their work and the underlying rationales for their measures and recommendations. Importantly, an agency needs to be seen as responsive to the public interests since a display of responsiveness may be more important than producing immediate results (Suchman, 1995).

The moral dimension of legitimacy can be distinguished when an actor is perceived as doing the right thing according to moral standards. The distinction between pragmatic and moral legitimacy lies in the difference between doing the correct or proper thing and doing the right thing, as in the most ethical, good, or fair. In the context of a societal crisis, this can be illustrated by how government agencies strive to meet victims’ physical and psychological needs (Falkheimer, 2021). Moral judgments can also “reflect beliefs about whether the activity effectively promotes societal welfare, as defined by the audience’s socially constructed value system” (Suchman, 1995, p. 579). Thus, it rests on an active assessment of the degree to which an agency adheres to social norms and shared ideals and if its practices are aligned with what is considered moral and ethical standards (Alexiou & Wiggins, 2019). Tyler (2006) describes how people follow rules when they think that these rules are aligned with their moral values. Moreover, when they perceive agencies as legitimate, they authorize them to decide what is right and wrong and tend to behave accordingly. At the same time, moral legitimacy judgments can function as a check against following immoral orders given by legitimate authorities (Tyler, 2006).

In a context like the Covid-19 pandemic, where issues such as public health, life, and death are deemed at risk, moral concerns should be expected. Assessments of moral legitimacy are active and ongoing. While the perceived outcome is central to such judgments, the evaluation of response features, like measures, strategies, organizational structures, and leadership, may be equally important. As Suchman notes, “a hospital is unlikely to lose legitimacy simply because some patients die” (Suchman, 1995, p. 580).

Governing the pandemic in the Swedish context

As Falkheimer emphasizes, the pandemic “may without doubt be a fundamental and global turning point of legitimacy for governments all over the world” (Falkheimer, 2021, p. 3). It has also been called a crisis for liberal democracy due to the implementation of strict regulations (Alteri et al., 2021). However, society’s response to a crisis is highly contextual (Lægreid & Rykkja, 2019). In the international arena, Sweden’s crisis management strategy in response to the pandemic has often been seen as exceptional, as it has not included lockdown measures (Andersson & Aylott, 2020; Giritli Nygren & Olofsson, 2020; Petridou, 2020). Sweden, with its extensive, comprehensive governance structure, with high reliance on independent bureaucratic administration and sectoral division of responsibilities (Öberg & Wockelberg, 2015), is a good example of the modern governing of interest in our study. Furthermore, the country’s general crisis management policies (e.g., MSB, 2018) mirror its governance model, which emphasizes local responsibility and decision-making autonomy, and how response operations should rely on the best available scientific knowledge and advisory functions in the shape of specialized national expertise. Importantly, government authorities are autonomous from the political government (Førde et al., 2019).

In this context, it is unsurprising that the political government relied on expert authorities to manage the pandemic. Three government agencies have been at the forefront; the Civil Contingencies Agency, the National Board of Health and Welfare (NBHW), and the Public Health Agency of Sweden (PHA). This study focuses on the two latter. The NBHW works to ensure access to high-quality health and social care, with a central task to produce statistics, regulations, guidelines, and knowledge for actors in the health and social care sector, including emergency preparedness issues. The PHA is responsible for public health issues and works to ensure good public health, including detection and protection against communicable diseases and other health threats. On their official websites, they present themselves as ‘knowledge authorities’, stating that their work revolves around collecting information and scientific evidence, which is then transformed into policies, such as recommendations, guidelines, and regulations, aimed at local and regional authorities, as well as the public. Together with the political government, these have been two main actors in the Swedish response (Ludvigsson, 2020).

Public press briefings as a legitimation process

Crisis communication is a social and interactive activity (Fredriksson, 2014). In line with the aim of this study – to examine how the legitimacy of government agencies is challenged and defended in an interactive process – we focus our analysis on national public press briefings. We see press briefings as a specific arena for interaction (Hafner & Sun, 2021; Maros & Nasharudin, 2016) and a site where communication actually happens (Falkheimer & Heide, 2018). Press briefings have long been a central platform for communication in the Swedish government’s public relations (Larsson, 2012). During the pandemic, press briefings were a key element in the strategic crisis communication, resembling a “campaign perspective” (Falkheimer & Heide, 2018, p. 56). The agencies communicated up-to-date situation reports, addressed current topics, reiterated the national strategy, and issued general advice. Importantly, they provided listeners with the underlying justifications for their decisions, measures, and recommendations.

While press briefings are usually an opportunity for accredited journalists to pose questions to experts in a semi-private setting (Eriksson, 2011), the Swedish press briefings held during the pandemic reached an audience of millions (Dahlgren, 2021). They were broadcast live on national television, radio, news websites, and YouTube. Watching them has been referred to as “gather[ing] around the bonfire” (Östlund, 2021, p. 126). However, as the public is not present, interactions between journalists and agency representatives can be seen as mediating a proxy dialogue, where the mass media can be seen to represent and cater to the interests of actors in the public sphere (Bhatia, 2006; Eriksson, 2011; Falkheimer & Heide, 2018).

Swedish media is characterized by liberal democratic ideals about the media’s dual role in informing and educating the public while at the same time monitoring those in power and holding them accountable for their decisions and actions (Syvertsen et al., 2014). In general, Swedish journalists² share ideals about scrutiny, objectivity, and neutrality (Wiik, 2014). Still, during crises, there is a mutual codependence since the mass media are critical channels for government agencies in their attempts to reach the public (Johansson & Odén, 2018).

Traditional media hold a prominent position in Sweden. Main newspapers maintain a high readership, also online, and public service holds a strong position in radio, television, and online (Syvertsen et al., 2014; Weibull & Jönsson, 2007), with small differences in media consumption among societal groups (Syvertsen et al., 2014). Some scholars have shown how the notion of this egalitarian and democratic Swedish media landscape has been threatened by digitalization and increased fragmentation of online media sources (e.g., Ohlsson et al., 2017). However, empirical studies of news consumption patterns during the pandemic confirm that established media platforms, like public service and traditional newspapers, held a very strong position as primary sources of information for the Swedish public (Bohlin et al., 2021), with an increased consumption among all age groups during the period of interest in this study (Andersson, 2021; Bohlin et al., 2021). This confirms the media’s dual role as a mediator between the government agencies and the public and the importance of including them in research on both legitimacy (Deephouse & Suchman, 2008) and crisis communication (Johansson & Odén, 2018).

A press briefing is a performative setting; it serves as a stage for mass education and reflects the ongoing drama of crisis. At the same time, it is a ritualized and highly formal setting with a structured participation framework featuring an agenda, a moderator, and a turn-taking system (Clayman & Heritage, 2021). The commonly included Q&A sessions provide an opportunity for co-orientation, clarifying the situation, and establishing a consensus on current events among responding organizations and the public (Botan & Penchalapadu, 2009). Ideally, communication then becomes a joint effort, a dialogue between participants, but in practice, power relations come into play and shape the interaction, mirroring negotiations or interrogations (Ekström, 2009). Clayman (2017) refers to press briefings as a process of legitimation-in-action.

Prior research on press briefings has often explored their discursive features and focused on political leaders such as presidents (Bhatia, 2006; Clayman et al., 2006; Ekström, 2009) or other politicians (Eriksson, 2011; Hafner & Sun, 2021), with an emphasis on political communication, negotiation, and ideological language. These studies show how the setting can “provide excellent data to study how ideologies are discussed and negotiated, how power relations are asserted, and how political differences on difficult issues are discussed and communicated in a positive way” (Bhatia, 2006, p. 174). However, and this is important to note, the press briefings of interest in our study did not feature any formal political actors. Instead, government agency representatives were at the forefront of the response. This emphasis on scientific, ‘neutral’ expertise is a feature that was shared with other countries (Gesser-Edelsburg et al., 2021).

Materials and method

The present study draws upon empirical data to investigate how legitimacy was challenged and defended in the interactive process between the media and government agencies during the Covid-19 pandemic press briefings. The analysis is based on Q&A sessions from 70 press briefings, spanning the period from March 6, 2020 (the first briefing) to June 30, 2020. This period was chosen as it was a time of high uncertainty when the crisis response was being shaped. The analysis only considers questions that were seen as openly critical and the responses. Critical questions reveal potential issues of legitimacy in the way they highlight tensions and conflicts, which the answers from agency representatives then attempt to resolve by defending their arguments and justifying their intended and implemented actions and regulations (cf., Hafner & Sun, 2021). We begin by presenting a brief overview of our material.

A daily press briefing was held each workday between March 6 and June 9, 2020. After June 9, they were held twice a week. Each ended with a Q&A session, where journalists were invited to pose questions to agency representatives. Each session took up approximately 50% of the time of the full briefing (see Appendix A). While several government agencies were present at the press briefings, the PHA and the NBHW have been the most prominent in the empirical data. Most of the questions were posed to their representatives; therefore, this analysis focuses on them. The PHA made all briefings available on YouTube. We downloaded them and coded their full content using NVivo software. Coding was broken down into four steps:

First, all of the material was coded according to the identified topics. For example, they included: ‘what is being done’, ‘flatten the curve’, and ‘Swedish strategy’. The Q&A section was coded as ‘Q&A’ and not analyzed further.

Second, we selected all of the questions and answers and coded them according to the topics that had already been identified in the first step, noting whether they were openly critical or not. These themes are presented in Appendix B. The coding of critical questions was inspired by a study on journalistic aggressiveness by Ekström et al. (2013). Questions were coded as critical if they were negatively framed, such as indicating an oppositional stance by questioning interpretations of facts or probing into disagreements, contradictions, and alternative explanations of events. In addition, questions were coded as critical if they suggested failure or inefficiency in response, requested justifications for decisions, actions, measures, or recommendations, or probed for accountability.

Third, we created a new coding scheme for the answers given in the Q&A session, focusing on the type of argument used to answer the question. This is also presented in Appendix B. Up to this point, the coding frame was jointly developed during regular meetings that were held to ensure inter-rater reliability.

Fourth, we adapted Alexiou and Wiggins’ (2019) operationalization of the three dimensions to our material (Appendix C). In this step, all critical questions and their answers were coded according to the three dimensions. This step was done jointly by the authors.

Previous research has expressed the number of questions (or statements) and answers as the number of exchanges between journalists and agency representatives (Dahlgren, 2021). However, this approach does not capture how one journalist can pose several unrelated questions in a single exchange or that one question can be divided into several probing queries on the same topic. Therefore, we counted each unique question/ statement as one exchange in this analysis. Consequently, some exchanges were divided into several unrelated questions posed by the same journalist, while other questions/ statements were clustered into one exchange when they related to the same topic. This approach resulted in a total of 805 exchanges. As we only investigated critical questions, the final sample comprised 212 exchanges (26% of the Q&A sessions).

The following sections present the outcomes of our analyses. It should be noted that the coding process had significant implications for these results. In particular, most questions and answers were coded as only one of the three dimensions of legitimacy (e.g., either pragmatic or moral), while in practice, they were multifaceted and could contain elements of one, two, or all three dimensions. For clarity and to highlight the interactive nature of the process, our results are organized into cognitive, pragmatic, and moral dimensions of legitimacy.

Results

Our results distinguish between the three dimensions of legitimacy in Q&A sessions. However, some initial general observations can be made, as particular framings and topics recurred across all three dimensions.

First, a prevalent theme was framing the Swedish response as exceptional from an international perspective; for example, many questions were framed as comparisons to other countries. Secondly, another recurring theme was the timing of the Swedish response, which was criticized for being introduced too late and not achieving the proper effects. This was especially prominent in discussions about specific measures, notably why facemasks were not recommended. Thirdly, many questions were framed with a hindsight perspective. For example, journalists used the past tense to ask what could have been done differently and what the outcome would have been. Finally, another common theme that touched upon all three dimensions was questions about disagreement within the scientific community, particularly discrepancies between experts or contradictory data.

The answers given also shared characteristics across all three dimensions of legitimacy. First, they were very pedagogical; agency representatives went to great lengths to explain epidemiological technicalities and provide different perspectives on the questions. The declared Swedish strategy of flattening the curve and ensuring the healthcare system was not overburdened was constantly and patiently repeated, day after day, week after week. In general, answers were delivered in a friendly and courteous tone, the journalist was given credit for asking relevant questions, and concerns raised were acknowledged. Finally, there was no attribution of blame to other actors in the answers.

In the following sub-sections, we take a deeper look into the process of challenging and defending legitimacy, drawing on the three dimensions.

Cognitive legitimacy – the way things are

The cognitive dimension of legitimacy emerged when taken-for-granted features of the response were called into question. It was reflected in Q&A that drew attention to the structural and cultural circumstances of the response. Such questions addressed the societal systems, assumptions, and ideological ideas in which the response management was embedded. They made up a relatively small percentage (13%) of the total number of critical questions. Many questions in this category related to the general crisis management system, including unclear legal mandates and tensions between national, regional, and local levels. These questions challenged agencies’ legitimacy by probing into the allocation of responsibility and a perceived lack of clarity or efficiency in the response. Several questions addressed flaws in preparedness to handle the pandemic, such as a lack of organizational capacity to implement specific measures such as mass testing or more generally critical of the country’s level of crisis preparedness. Such questions framed Sweden as a country inexperienced in handling crises, with a neglected healthcare system and a vulnerable just-in-time supply system. This can be illustrated with the following quote:

Journalist: In an interview/…/[you are] critical that we do not have any emergency stocks and/…/that the hospitals have just-in-time deliveries. And you call it the IKEA syndrome. And it makes one wonder, first of all, is it correct that there are no emergency stocks, and in that case, when and how and why did we end up in this situation? (2020-03-12)

The above question asked for an explanation regarding a previously-identified lack of preparedness but, with its reference to IKEA, also touched upon a particular Swedish notion of efficiency and its drawbacks. Interestingly, the critique against the supply system originated from the agency representative, and, in the answer, the lack of resources is acknowledged as a known problem.

NBHW: We who work with preparedness have said for a long time that supply chain continuity and security are important and will be critical in all kinds of societal development or disruptions. (2020-03-12)

This can be interpreted as an implicit criticism of the Swedish government for not acting on expert advice before the pandemic. It is also an example of an interaction in which the authority agrees with the journalist’s framing while at the same time reproducing their legitimacy as knowledgeable experts.

Typically, questions in the cognitive dimension received cognitive answers. Agency representatives retained their legitimacy by confirming their position in the crisis management system. The roles, duties, and legal mandates of different actors were thoroughly clarified, often with lengthy explanations of ‘how the system works’ and what the agencies can and cannot do within the scope of their responsibilities. The national government agencies often positioned themselves as a central actor in a network, as a coordinator or provider of expert knowledge. Other actors were described as responsible for specific decisions, measures, and operational activities. Interestingly, these clarifications of who was responsible for what were not presented as attributions of blame. Instead, they were factual descriptions of given circumstances – ‘the way things are’ – reaffirming the cognitive dimension of legitimacy.

In some cases, the explanations of the system were intertwined with references to the current development of the pandemic, described in terms of statistical trends and epidemiological modeling. The agencies portrayed themselves as proactive and transparent and underlined Sweden’s longstanding competence built upon the country’s past experience of handling pandemics. In such accounts, agencies relied on their pragmatic legitimacy to manifest their expert authority. However, it is also interesting to note that some of these answers revealed the underlying assumptions that guided the Swedish approach. For example, claims were made that the strategy rested on established, culturally-specific public health management practices, with voluntariness and public education as key features. In this context, the choice of words matters:

Journalist: Can you clarify what you mean when you say ‘recommend’?

PHA: We are very careful in the Swedish language with what words we use when there is a legal obligation to do things, a legal obligation to achieve something, or just general advice to do something. I don’t think that that is what we are talking about here. I think what we are talking about here is the sort of Swedish culture, how they interpret recommendations from authorities./…/I think there is a long tradition in Sweden to follow the advice from authorities. We don’t need legal advice to get people to come and vaccinate their children, for example. We get 98% of parents coming anyway, while in other countries, you need legal obligations to do things, and I think there is a cultural difference./…/all these things, you need to put into context, and I think it is very important for each country to do what fits their context. If you need legal obligations to do things, that is fine. If we can do it the voluntary way, that’s fine. (2020-03-27)

References to cultural circumstances, as in the above quote, constituted an important legitimation for the contextual – and perhaps at this time unique – Swedish strategy. Ultimately, these descriptions of ‘how things are’ reproduced the “shared cultural-cognitive schema” that legitimacy judgments in the cognitive dimension rely on (Alexiou & Wiggins, 2019, p. 473). In theory, such statements are an opportunity for this passive, subtle form of legitimacy to be openly scrutinized. However, in practice, none of the questions coded in the cognitive dimension went beyond requesting clarifications.

Pragmatic legitimacy – the proper response

The pragmatic dimension of legitimacy could be analytically distinguished in questions that expressed doubt about response activities and their alleged benefits. Such questions aimed to investigate the underlying knowledge base and the rationale for measures, recommendations, and the overall strategy. Another central theme was the appropriateness and efficiency of measures and regulations and whether enough was being done to tackle the pandemic. Potential consequences were also addressed, both in terms of the risks for those infected and the impacts of strict measures. Most of the critical questions (68%) fell into this category.

The presented knowledge basis for the government’s strategy and the measures that were taken generated many questions about the data and scientific evidence. These questions, addressing presented facts and their interpretation, clearly aimed to challenge the foundations underlying expert judgments. For example, direct questions were asked about the accuracy of numbers and statistics, their validity (collection and analysis methods), and their reliability (can the data be trusted). The validity of estimative mathematical models was called into question, as was the credibility of the agencies (with references to specific invalid data, calculation errors, or failed predictions).

Many questions aimed to scrutinize the rationale underlying decisions and measures, including why certain measures had been introduced, why others had not, and the appropriateness of enforcing or lifting specific restrictions. A frequent topic was the decision not to recommend a general use of facemasks. These questions were often framed around measures taken in other countries or referring to statements from other experts. Questions like these captured how legitimacy was challenged with comparisons: If measures are based on the best scientific knowledge, then why does everyone else seem to be doing things differently?

Questions about the rationale driving the overall strategy also fell into this category. They often focused on expected outcomes or negative consequences, and the death toll was used to probe for self-critical and evaluative statements. Occasionally, the basic assumptions of the strategy were questioned; Is the overall aim herd immunity? Why is the goal not complete suppression of the virus? Why is lockdown not the appropriate response? Will voluntary compliance with government recommendations really work?

In their answers, agency representatives relied on their expertise, giving lengthy and pedagogical explanations of epidemiological data. They explained its inherent uncertainty, the complexities of science, and the need for a knowledge-based strategy. Quantified information was a central feature of answers in this category, along with the contextualization of numerical facts, such as the methods used in particular studies. Representatives described how information and epidemiological evidence had been gathered and analyzed and how there are different types of data, tests, and measurements, which all need to be compared to normal circumstances. They often provided general clarifications of measures and recommendations, also when answering questions that did not specifically ask for this. There was an apparent educational attempt to explain terminology, measures, decisions, and strategy. This pedagogy was often built on chains of arguments or a detailed description of the current situation, sometimes highlighting a perceived gap between the theoretical logic and the practical feasibility of suggested measures.

A key feature of the pragmatic dimension was how agency representatives explicitly sought to establish their role as experts. They emphasized their ongoing responsibility to monitor unfolding events and analyze new data to improve the accuracy of current knowledge and epidemiological models. They often pointed out that their main role in the system was to share information and educate; they underlined that knowledge is unstable and evolving and that continuous learning was a key feature of the strategic response. By referring to these ongoing activities, they constantly re-established themselves as the most appropriate people to carry these responsibilities. Agency representatives frequently acknowledged the complexity of the situation. They referred to how the facts were uncertain, for example, pointing to discrepancies between modeling and outcomes, how measures were context-dependent, complicated, and interrelated, and how the potential effects were unpredictable or ambiguous. This point is exemplified by the following quote related to the government’s strategy:

PHA: We believe that we have the right strategy for Sweden. If the right thing would be to close completely and then slowly open, then close again, like in some countries … if that works, we’ll have to see. What we are talking about are the long-term issues. We saw in Sweden that it took 3-4 weeks after the introduction to the big outbreak in Stockholm. That is a long time. It is like steering a large ship with a time lag of 3-4 weeks. It is too early to tell. (2020-06-03)

This quote illustrates how agency representatives constantly reaffirmed their belief in the Swedish approach. In many answers, they repeated the strategy, its goals, and their recommendations to the public and explained why certain measures were deemed appropriate. They maintained that all the decisions made by their agencies were based on many deliberations and the best-available, current knowledge. Yet they did not always claim to have all the answers. Frequently, representatives avoided speculating, stated that knowledge was insufficient, or simply said “I don’t know”. When asked to assess if the response was appropriate, they argued that data were lacking, accounted for the pros and cons of measures, and claimed that any final judgments could only be made after a future evaluation.

Many answers also fell into the cognitive dimension, notably when agencies emphasized their legal responsibilities, clarified a legal term, or referred to their own tasks, formal role, and overall mission. In such accounts, national agencies positioned themselves at the heart of the response system and knowledge-based governing while at the same time emphasizing the holistic idea that all government agencies and all individuals were working together in response to the pandemic. On a few occasions, agencies gave answers that fell into the moral dimension of legitimacy, for example, when they claimed to be shaping a strategy that was beneficial for the whole of society and not just specific groups (for example, this was the rationale for keeping schools open). In general, however, moral issues were most visible when representatives expressed concern and self-critique for failing to sufficiently protect the elderly.

In conclusion, pragmatic legitimacy was challenged by concrete, critical questions directly related to data interpretation, response activities, and regulations. Journalists probed the benefits of the response and questioned whether agencies were really doing what was best for the public in the current situation. In turn, agencies defended their legitimacy when they provided clarification and explained that they were indeed doing the right thing in a multiplicity of ways, supplemented by arguments in the cognitive and moral dimensions of legitimacy.

Moral legitimacy – the right response

The moral dimension of legitimacy emerged in questions and answers that touched upon ethical issues. Such questions probed into what was good or bad, right or wrong, just or fair in the response, and whether measures and regulations promoted societal welfare and were aligned with societal values. These issues made up roughly a quarter of the critical questions.

Moral questions about response activities focused on people affected by the virus and prevention measures. Some questions concerned vulnerable groups, such as the frail and elderly, and the lack of timely measures to protect them. Others concerned potentially discriminatory measures, such as the failure to protect exposed practitioners (healthcare personnel, teachers, and restaurant staff) or minorities living in the suburbs. Another frequent topic was the priorities of the Swedish healthcare system, both in terms of the priority given to Covid-19 patients compared to those with other diseases and why not all patients with Covid-19 were placed in intensive care units.

Some questions in this category related to the comparative weighing of societal values and moral judgments, such as future effects of the pandemic in terms of expected rates of suicide, economic disruptions, or consequences of disrupted livelihoods. Interestingly, only one question in our empirical data clearly framed the Swedish response as too restrictive and unjustified in relation to expert opinions on the virus’s characteristics.

A clear majority of questions that challenged government agencies’ moral legitimacy addressed the outcome of the response. The death toll was a key theme and seemed to become the main criterion for whether the Swedish approach was a failure or a success. The high number of Covid-related deaths was compared to neighboring countries, and an increase was framed as a failure.

Journalist: German TV, I would like to once again look at other countries and Sweden. Denmark stated yesterday that if they keep everyone locked up until Easter, they can do something about their measures. The numbers are rising in Sweden. How do you view these liberal corona politics? (2020-03-31)

This quote is an example of when the appropriateness of the Swedish ‘liberal corona politics’ was questioned and whether the strategy was reasonable when so many people were dying. Similarly, the basic assumptions of the strategy were challenged, mainly with questions about whether the goal of the approach was to reach herd immunity, if this meant that it was acceptable that many people died, and if perhaps complete suppression would have been a more ethically appropriate goal.

Few answers in this category straightforwardly addressed moral issues. This does not mean there were none, but often such comments were intertwined with non-value-laden content in the form of epidemiological explanations or descriptive accounts of the current situation. Agencies frequently defended the morality of the Swedish strategy, pointing out that it was knowledge-based and that the healthcare system was coping. They provided reassurance that they were monitoring the situation closely and were willing to learn and adjust the strategy in the light of new knowledge. When the high death toll was questioned, they often suggested that it was not the best measure for judging the strategy and explained how mortality rates should be interpreted in the Swedish context, based on long-term data:

PHA: Yes, we have discussed what this death toll really means. It is people dying, not only numbers. You have to relate it to a lot of different things. The death toll per day, which varies extremely in Sweden, it ranges from 4-140, it is not a relevant number. Countries are at different stages, we see this in serological investigations, Finland has [a certain] percent and Spain another. You have to be careful with what you can know. [Death tolls] do not provide a correct answer to who is succeeding or not. (2020-05-20)

In contrast to the above example, towards the end of the studied period, agencies did acknowledge that the death toll among the elderly was too high.

Sometimes, the agency representatives responded to moral questions by discussing moral issues. For example, they accounted for the ongoing weighing of societal values in the response work and sometimes provided the logic underpinning trade-offs between risks and benefits. Often, they emphasized how the strategy was configured to benefit the wider society and showed concern for many public health concerns, not only infections. At the same time, they acknowledged that doing the right thing was complicated, given the complexity of the situation and its uncertain impacts. The available data and evidence were said to be lacking or flawed, implying that moral judgments could not be made. They referred to how it was too early for any conclusions and that evaluations and extensive research would continue for many years to come, implying that moral judgments must be made in the light of post-pandemic assessments.

Another typical response to critical moral questions was to describe how responsibility was distributed in the response system, which falls within the cognitive dimension of legitimacy. While government agencies positioned themselves as a knowledge center, they made clear that others were responsible for implementing the strategy with context-specific measures. This became a way to allocate responsibility without attributing blame or evaluating other actors’ work. Similarly, in response to criticism about how the Swedish strategy compared to other countries, agencies pointed to differing national, legal, and cultural contexts and declared such comparisons too complex.

In conclusion, government agencies relied heavily on their pragmatic and cognitive legitimacy when defending the moral dimension. In doing so, they reaffirmed their position as a center of knowledge, reassuring the public that the approach was neutral and evidence-based and that they were doing their utmost to handle the situation. They manifested an ‘evidence-based’ morality in which ethical judgments were dependent on established structures and standardized procedures, and, ultimately, they postponed the moral discussion for a distant future.

Discussion

In this study, we have examined legitimacy as an interactive process (Hearn, 2011; Suddaby et al., 2017), focusing on agencies’ strategic communication in their governing efforts during a crisis. This entails understanding legitimacy as an ongoing justification of decisions, actions, and regulations in specific situations (cf., Zelditch, 2001). The press briefings are an arena that captures this ongoing negotiation where journalists’ critical questions seek to challenge legitimacy, and government agencies’ answers seek to defend or rationalize it. The framework with three dimensions of legitimacy – cognitive, pragmatic, and moral – proved to be a straightforward and appropriate way to categorize the empirical data.

The analysis of the cognitive dimension of legitimacy shows that this was an important feature of Q&A interactions as answers to questions that referred to systems, roles, responsibilities, and descriptions of ‘how the system works’ and ‘how the Swedish people are.’ The use of such structural and cultural references reproduces the shared cultural-cognitive schema that the Swedish response strategy relied on and seemed to function as an impartial and versatile way to respond to all kinds of criticism. We consider this a central feature of the communication and one that was not extensively challenged. By not probing further into accounts of ‘the system,’ journalists seemed to accept and implicitly agree with such statements, thereby contributing to reproducing the cognitive dimension of legitimacy and reaffirming the position of the government agencies as legitimate authorities to handle the pandemic, with a right to govern (Falkheimer, 2021; Tyler, 2006).

Most questions and answers were related to the pragmatic dimension of legitimacy. This is not surprising, as the press briefings covered key aspects of the response to the pandemic and revolved around the actions that agencies had taken (or not) and the regulations they had formulated (or not). Many critical questions were focused on the expert knowledge that underpinned the response, creating a detailed, rather sophisticated debate about facts, measures, and strategy, of great significance for the process of legitimation. The pedagogic style of the representatives’ answers was a central feature of communication, as well as the subdued and neutral tone in interactions with journalists. Representatives rarely showed emotion or were dismissive. The so-called ‘battle’ between critical journalists and powerful authorities turned into a respectful conversation. The main defense strategy seemed to be to wear down any criticism with repetitive, patient reasoning. This was perhaps a way for the agencies to avoid conflict and bypass the political dimension of crisis management.

However, while portraying themselves as neutral civil servants, the representatives drew up the boundaries of the press briefings by consistently falling back on the epidemiological rationalities that shaped the Swedish response. This strategy limited what could be discussed and restricted opportunities for critique. The reliance on expertise created an asymmetrical power relation, where journalists had to be equipped with a high level of scientific literacy if they were to find flaws and inconsistencies in the presented arguments. Even if the representatives, with their pedagogical approach, appeared to be helpful and trying to balance this asymmetrical power relation, journalists were left behind. Journalists seemed to solve this problem by referring to other scientific studies, other experts, or approaches in other countries. However, most attempts to ask probing questions were effectively shut down since the agencies provided rationales for everything. When they did not have an answer, they resorted to uncertainty as a strategy. The argument that ‘perhaps no one will ever know what the right thing to do is’ becomes impenetrable. Other scholars have pointed to this particular feature of the power of epistemic authority; how experts get to describe, define, and explain ‘reality,’ with very little room left for critique or insight into the knowledge-producing practices (e.g., Gieryn, 1999; Lidskog et al., 2020; Porter, 1996).

The moral dimension of legitimacy is present in the empirical data, often in relation to death tolls. Agency representatives responded to moral questions by describing their roles and responsibilities; they assured the public that their recommendations were neutral and evidence-based and that they were doing their utmost to manage the situation. In other words, they used cognitive and pragmatic dimensions of legitimacy to answer moral questions. Any ethical judgments also became pragmatic as they are claimed to be based on standardized procedures and medical-ethical protocols. Alexiou and Wiggins (2019) argue that this alignment of ethical standards is an important element in how audiences acknowledge the moral dimension of legitimacy. Still, it is unclear if moral issues, like questions of life, death, or victims’ needs (cf., Falkheimer, 2021), were discussed in a manner that those affected might have desired.

In sum, this empirical analysis shows that the pragmatic dimension has considerable weight in the process of legitimation and can be linked to the centrality of epistemic authority. However, this study also clearly shows how the different dimensions of legitimacy co-exist and reinforce one another, as a single question and subsequent answer can touch upon all three dimensions. Deephouse and Suchman (2008) point to this mutual dependence, and this analysis has contributed with an empirical account of how the legitimacy concept is layered and dynamic.

Viewing legitimacy as a process is one way to investigate its dynamic features, but the question remains of what is actually legitimized by the government agencies. This is also a multi-faceted aspect of the communication since possible subjects of legitimation are “almost innumerable” and can encompass authority structures, organizations, regulations, procedures, and actions, ideas they rely on, as well as the system of power they are part of (Deephouse & Suchman, 2008, p. 54). In this study, government authorities can be seen to defend their position as legitimate authorities every time they enter the stage at the press briefings and expose themselves to critical questions. They also legitimize their decisions, measures, and regulations to the public by continuously describing and explaining their knowledge-based considerations. Importantly, they legitimize the particular way they have decided to exercise power, relying on their epistemic authority to govern the crisis.

Hence, the power of government agencies is not abstract, and the centrality of government agencies’ strategic communication in this legitimation process is evident. Therefore, this study can improve our understanding of legitimacy and how operational crisis communication is linked to strategic dimensions of societal governance and the exercise of government power (cf., Lægreid & Rykkja, 2019). An empirical, descriptive account of this modern, liberal power and the process of legitimation of the Swedish ‘liberal corona politics’ does not lend itself to assessments, critique, or offer alternative forms of rule. Instead, as Hearn emphasizes, it can offer an increased understanding of how power is exercised in obvious yet subtle, taken-for-granted, knowledge-based processes aimed at everyone’s engagement in self-governing (Hearn, 2011).

Many scholars point to how legitimacy ‘lies in the eyes of the beholder’ (Christensen et al., 2019; Tyler, 2006; Alexiou & Wiggins, 2019), implying that ultimately, legitimacy can be seen in the compliance of the public – something that is not included in this study. This means that whether or not the public perceived the government authorities and the strategies as legitimate or not, lays outside of our present scope. Attempts to study legitimacy as a process often fall short in this aspect (Suddaby et al., 2017), which makes a strong argument for including measurements of public acceptance and compliance in future research on crisis communication, legitimacy, and power.

Conclusions

The aim of this study has been to give an empirical account of the process of legitimation, analyzed in press briefing Q&A sessions, as a case of agencies’ strategic communication. The results and analysis contribute to our understanding of both the legitimacy concept itself and how strategic communication is linked to societal governance and the exercise of knowledge-based government power. Legitimacy serves as a bridging concept to understand this link. In Sweden, crisis communication was not framed around any references to legal obligations or citizen duties, even though the Communicable Deceases Act (2004:168) contains such applicable formulations. Instead, the governing of the pandemic was based on non-coercive power, even to the extent that disciplining language was absent. The strategy relied on descriptive, often numerical accounts of the situation, combined with rationales for why everybody must act in a specific way; the process can be captured by a chain of reasoning that mirrors the logic: there is, therefore, we must (cf. Desrosières, 1998). Therefore, a key contribution of this study is that it captures a modern form of knowledge-based government power and highlights how strategic communication is a central feature in its ongoing processes of legitimation.

Acknowledgment

We are grateful to the two reviewers who provided clear and constructive comments that improved the quality of this contribution.

Disclosure statement

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

Data availability statement

The data that support thefindings of this study are available from the corresponding author, [MS], upon reasonable request.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This research is a part of the project Power and Norms in the Work with Achieving Direction and Collaboration During Crisis (MaNISK), funded by the Swedish Civil Contingencies Agency from 2017–2022.

Notes

¹ For an overview of the legitimacy concept’s historical development in this field, see for example, contributions from Deephouse and Suchman (2008) as well as Suddaby et al. (2017).

² Journalists have their own epistemic authority (Carlson, 2020; Ekström & Westlund, 2019) and need to maintain their own legitimacy (Carlson, 2020; Ekström, 2002), but this is beyond the scope of our analysis.

Appendices Appendix A.

Date and duration of press briefings and the percentage of time dedicated to Q&A sessions

Appendix B.

Coded Q&A themes

Table

Questions:			Answers:
Basic theme:	Recurring features:	General topic:	Themes – Knowledge authority:	Themes – Structural accounts:	Recurring features:
Neutral questions: -Measures -Facts -Estimates, assessments, or prognoses Critical questions: -Measures -Facts, basis for a decision -Recommendations, restrictions -Swedish strategy	Comparison with other countries Referencing specific groups or locations Holding someone accountable Referencing current research Referencing other experts Openly addressing criticism	Number of people infected Spread of infection Death toll ICU statistics Testing Facemasks Vaccines Virus mutations Immunity Healthcare equipment Social distancing Risk groups Schools Travel Public health issues Personal integrity	Clarification of epidemiological knowledge Clarification of other facts Clarification of measures Clarification of the Swedish strategy Clarification of prognoses Describing the underlying rationality Referring to scientific uncertainty Referring to data quality Referring to broad public health issues Referring to the ongoing analysis	Referring to the agency’s mission Referring to other actors Referring to written policies Clarifying the division of responsibilities	Expressing an opinion or an evaluative statement Emphasizing individual responsibility Taking a defensive position Avoiding the question Avoiding ethical issues “I don’t know” Non-answer (nonsense)

Appendix C.

Operationalization of the concept of legitimacy.

Table

Dimension of legitimacy	Features of dimensions (topics addressed in Q&A sessions)
Cognitive	Formally-acknowledged roles Structural accounts of roles, mandates, and scope of responsibility General mission and the essential function of agencies Conceptual descriptions of the public health governance system Conceptual descriptions of the crisis management governance system General work procedures and tasks Legislative frameworks and policies References to social norms or cultural characteristics
Pragmatic	Epidemiological knowledge Facts and estimates Scientific uncertainty Data quality Measures and methods Practical consequences of response efforts Flaws/benefits of response measures Underlying rationales for measures, decisions, and strategy (pragmatic)
Moral	Weighing of societal values Societal welfare (what is best for society) Weighing of groups against groups Issues of fairness and discrimination Existential questions (life and death, right or wrong) The physical and psychological needs of victims Ethical standards for operating practices Underlying rationales for measures, decisions, and strategy (moral) Perceived outcome of the response (death toll, long-term consequences) Moral evaluation of aspects of the response such as activities, strategies, goals, organizational structure, and leadership