Institutional trust and emergency preparedness: Perceptions of Covid 19 crisis management in Sweden

Abstract

The Covid 19 pandemic has put the issue of public trust at the centre of political analysis. This article inquires into the level of public trust in Sweden concerning current crisis management as well as preparedness for future crises. The empirical basis for this study consists of unique data generated through two surveys on the Covid 19 virus conducted during 2020 in Sweden. Respondents were asked to assess their trust in different actors' crisis management. In addition, the first survey asked about how the actions of different actors impact on the forward-oriented trust held in Swedish emergency preparedness for a future severe crisis. Four key findings stand out. First, the results display rather weak levels of interpersonal trust. Second, a surprisingly high level of institutional trust can be found during the initial phase of the pandemic. Simultaneously we see somewhat decreasing levels of trust later during the pandemic. The results include interesting variations across age, gender and education. Third, relating the results to earlier research on trust in Sweden, the results contain indications of stable trust levels despite the all but stable contextual conditions. Fourth, the analysis shows important variations in trust levels among different institutions and layers of government.

1 INTRODUCTION

The Covid 19 pandemic has become a defining feature of contemporary political life. Characterized by complexity, uncertainty and acuteness, the crisis has brought unprecedented actions by governments, in turn challenging both domestic policy consistency and international collaboration. In addition—and in focus in this article—it has come to imply a novel situation regarding public trust in government.

In the league of public trust Sweden stands out as one of the most high-trusting countries, both in terms of interpersonal trust and institutional trust. As we will show Sweden often comes out at the top regarding countries with the highest ranking in different trust barometers. Sweden thus entered the Covid 19 pandemic from a favourable position, where its citizens were expected to trust the public authorities and their measures on how to deal with the pandemic.

However, this expectation is challenged by the Swedish crisis management strategy that stands out in various respects. Criticized by many but at the same time enjoying considerable support by the Swedish people (European Commission, 2021; Kantar Sifo, 2021), the strategy has been more voluntary and less pervasive than in most countries (Hale et al., 2021; also Johansson et al., 2021; Kuhlmann et al., 2021; Pierre, 2020). Simultaneously, the number of deaths as a proportion of the population is sadly enough comparatively high, at least in the first wave of the pandemic (Johns Hopkins, 2021). The Swedish case thus leaves us with the puzzle how public trust is affected when government authorities in a typically high-trusting country chooses a crisis management strategy that is drastically different compared to other/neighbouring countries and is highly questioned (Nygren & Olofsson, 2020).

Against this complex background, we depart from the aim to explore how the level of trust that Swedes hold in different actors' crisis management was affected by the first wave of the Covid 19 pandemic. Based on our findings we also want to inquire into the link between trust and emergency preparedness for a future severe crisis or even war. While of immediate relevance from a crisis management perspective, the Covid 19 context also allows for reflections on the division of labour that results from the Swedish organization of public administration and from policy development in recent decades in the direction of decentralization and privatization (on the latter point see Hall et al., 2016).

The empirical basis for this article consists of data generated through two surveys on the management of the Covid 19 pandemic, conducted during April–June and September–December 2020 in Sweden. In the surveys the respondents were asked to assess their trust in different actors' management of the crisis thus far. In addition, the first survey included a future-oriented question on how different actors' management of the crisis have impacted on the trust held in Swedish emergency preparedness for a severe crisis or war.

We now turn first to our theoretical points of departure regarding the perceptions and dynamics of institutional trust. We then provide an empirical background on trust in Sweden to establish a baseline. After outlining our methodological tools, we turn to our empirical analysis of trust in Swedish civil emergency preparedness during the pandemic. We conclude our article by returning to our aim and discuss our conclusions on how trust has been affected in Sweden and how the level of trust in high-trusting countries more generally can be affected by external shocks like the pandemic.

2 INSTITUTIONAL TRUST: PERCEPTIONS AND DYNAMICS

As a key social phenomenon, trust continues to attract considerable scholarly attention.

Research on trust generally departs from the distinction between generalized/interpersonal trust and institutional (also referred to as systemic or political) trust (see Fitzgerald & Wolak 2016 for discussion) but also inquires into how the two dimensions are connected to each other (Keele, 2007; Spadaro et al., 2020)—it is a common assumption in the literature that the two dimensions are related, but it remains unclear how strong the link between two is and what the causal relationship looks like (Johansson et al., 2021, p. 343). While institutional trust is in focus in this article, our data also allows for a discussion on interpersonal trust in a crisis context.

Why does institutional or political trust matter? Public trust in the government, government agencies and other collective actors is important not only for the legitimacy of the actors and their performance per se (related to among other things the degree of compliance by the public) and, relatedly, for the functionality of society, but also for the willingness of the public to take on responsibilities and make sacrifices for society and specifically societal security (Esaiasson et al., 2016, pp. 9–10; Spadaro et al., 2020, p. 2). Not least in a crisis context is trust important as it relates to the inclination of citizens to comply with government recommendations (Johansson, 2021, pp. 337–338). Pierre (2020, p. 480) argues: ‘The degree to which institutions can shape social behaviour depends to a large extent on the level of institutional trust in society’.

Conceptually, trust can be understood in a number of ways, reflecting the distinction above. A prominent understanding denotes trust as an individual trait, a ‘deeply embedded disposition’ as stated by Fitzgerald and Wolak (2016, p. 131), in which case a person simply is trusting or distrusting by nature. Alternatively, trust can be thought of as a reflection of the environment in which an individual exists and of the interactions in which the person engages. In this light, experiences (own and others') and social connections play key roles for determining levels of trust and distrust. This understanding is at the heart of the literature on social capital (Putnam, 1993; Rothstein, 2005). Finally, trust can be conceptualized as an individual's subjective response—attitude—towards the performance of institutional actors (the government, agencies or other collective actors); that is, a perception about the outcome of an implicit or explicit exchange (about welfare production; security and defence provision etc), and in which case different objects/entities of trust may enjoy different levels of trust by the same respondent (Christensen & Lægreid, 2005, pp. 487–490; Fitzgerald & Wolak, 2016, p. 132; also Berg & Johansson, 2020, pp. 291–293). While it is reasonable to assume that these different dynamics of trust relate to each other, issues of causality, drivers etc are matters of continued scholarly dispute (Fitzgerald & Wolak, 2016, pp. 131–132, 142; Van Ryzin, 2007, p. 527).

As noted above, trust can thus be formed by personal experience or by information about performance and/or capacity of a certain actor, but is nonetheless a subjective phenomenon based on perceptions, that is, an attitude towards some other actor/entity. As such, interpersonal as well as organization/institutional relationships can be thought of as psychological contracts, which in contrast to formal contracts are subjective to the effect that ‘one party's understanding of the contract may not be shared by the other’ (OECD, 2017, p. 16). For instance, there may be different views on responsibility and division of labour between say the public and various government agencies, or discrepancies in ideas about personal responsibility in a political context, such as the role of the individual in a welfare state context. In essence, it is the perception of government performance (rather than actual outcomes) that forms the basis for the degree of trust. Hence there is a link between prior experiences and perceived performance (rather than the idea that citizens respond directly to the outcomes of government) (Van Ryzin, 2007, pp. 528–529, 532).

Institutional or political trust can be understood as the degree to which citizens appraise public institutions and the government (and leaders/representatives of these bodies)—‘the extent to which individuals accept and perceive institutions as benevolent, competent, reliable, and responsible toward citizens’ (Spadaro et al., 2020, p. 3). In the literature there are competing ideas of what drives institutional trust. Often related to welfare service production, research highlights features such as efficiency and good governance, more specifically fairness, transparency and honesty. This falls back on two different dimensions for understanding the nature and development of trust, namely competence (efficiency and capacity) and values (norms and principles that make up the foundation for performance) (OECD, 2017, p. 21). Regarding the former, a conventional assumption is that government performance will impact citizen satisfaction, which in turn will affect the level of trust of the government/government agencies. At the same time, it is fully reasonable to think in a causally opposite direction that previous experiences of, images of and information about others' experiences/functionality of the government determine current levels of trust, which in turn influence perceptions of current government performance (Van de Walle & Bouckaert, 2003, p. 892f). According to this line of reasoning one can expect age to influence trust since with increasing age one has had more experience of government functionality and performance; age is in this light positively correlated with the level of trust. A competing perspective on how age could impact trust is specifically related to the pandemic, as older people are generally more vulnerable and prone to be negatively affected by a Covid 19 infection (Jørgensen et al., 2021).

Education is another variable that can have an effect based on knowledge about government functionality; higher levels of education is likely to be positively correlated with knowledge about the political system and the administrative organization of the public sector. Education (as well as income) could also be explanations based on socioeconomic standing, where higher standing is likely to create more trust (Jørgensen et al., 2021). It also follows from the argument on values having an effect on trust that a citizen's political values may influence his or her trust in the government. Someone standing to the left being in favour of a big government is more likely to trust a big government. We also know from previous research that such values vary between men and women, gender is thus another variable with the potential to influence levels of trust, to the effect that women have higher trust in the government than men do (Christensen & Lægreid, 2005, p. 495f; regarding Sweden, Falk & Andersson, 2021, pp. 47–48). Applied to the field of crisis management, the distinction above between competence and values corresponds to governance capacity and governance legitimacy, respectively (Christensen & Lægreid, 2020, pp. 713–714).

Modern states are characterized by complex modes of organization. While different states naturally are organized in different ways, it is a generic phenomenon that ‘the state’, from the perspective of citizens, appears as a set of entities situated on different levels of the political system. It is necessary to differentiate between levels of government because the distance between the citizen and the government varies depending on level, as does the intensity of contact. There is reason to expect, not least in modern welfare states, that local government is enjoying higher degree of trust because of closer distance between the citizen and the government and because contemporary welfare production primarily takes place at the local level, which in itself is a reflection of strong trends in recent decades towards decentralization (Fitzgerald & Wolak, 2016, pp. 133, 135–136; Van de Walle & Bouckaert, 2003, p. 904). Still, as Fitzgerald and Wolak argue, there may be reason to problematize the differentiation of government levels. Although in a general sense it seems clear that local government in many European countries have gained increasing influence, responsibilities and resources, it may still be difficult for citizens to separate different levels of government; in consequence, national party politics and policy development may impact on people's attitudes towards not only national but also local government (Fitzgerald & Wolak, 2016, p. 131). There may also be reason to question the higher degree of trust at the local level since research shows that the local level is not only associated with a closeness to the citizens, but also with tendency towards higher risk of corruption (Erlingsson et al., 2008). The issue of closeness also relates to the differences between citizens in urban areas and citizens in rural areas. Citizens living in rural areas are likely to find the central government more remote and distant, while the local government may be perceived as closer, since rural municipalities often are smaller and personal contacts with decision-makers are easier to establish.

From an organizational perspective, the complexity of the modern state may thus be hypothesized to challenge public trust in government, as it becomes harder for the single individual to acquire a correct understanding of distribution of resources, loci of responsibility, chains of delegation and ultimately possibilities for demanding accountability.

A severe crisis such as the Covid 19 crisis can be hypothesized to challenge trust in government further. A crisis can in general terms be understood as an unexpectedly erupting and rapidly escalating process characterized by uncertainty, extraordinary procedures and limited time for decision-making, limited availability of information and (risk for) contradictory communication (Lipscy, 2020; Pearson & Clair, 1998; Welch, 2021). Due to their acuteness and disruptive nature, crises can hence in general terms be expected to exacerbate the challenges referred to above. In addition to this general hypothesis, other research shows the importance of efficient public communication during a crisis. The degree of success in public communication can according to this study help us determine if the general hypothesis of trust being challenged during a crisis, will be strengthened or weakened (Lee, 2009; Wang et al., 2020). On the contrary, it can be hypothesized that a crisis potentially strengthens citizens' trust in the government. Kritzinger et al. (2021, pp. 1207–1211) point to two related factors that speak in this direction—that citizens turn to the government as a way of reducing the uncertainty that a crisis produces (i.e., trust as a cognitive mechanism for reducing uncertainty), and that citizens trust the government to have the capacity to protect them and as a crisis unfolds, enactment of political initiatives translates to enhanced levels of trust. Having said that, they conclude that the early empirical record in the context of the Covid 19 crisis is mixed (see also Devine et al., 2021 for a review of studies in this context).

As a final point of discussion, again of enhanced relevance in advanced welfare states, one may inquire into the relationship between interpersonal trust and institutions. In essence, the issue revolves around whether institutions crowd out interpersonal trust (the latter being less necessary once well-functioning institutions built on accepted rules and shared norms of behaviour are in place), or whether also basic institutions rather form the foundations for the enhancement of interpersonal trust (Spadaro et al., 2020, pp. 2–3). In turn, this is linked to issues of citizens' personal responsibilities, for instance in times of crisis, and their willingness to contribute to society and societal security, but also their capability to do so (Sandstig & Ivanoff, 2020).

This relates to the issue of determinants of interpersonal trust. Generally speaking, research gives rise to two lines of thought regarding factors affecting the dynamics of interpersonal trust—an individually-oriented approach focused on issues such as morality, responsibility, and personal experiences, and a society-based approach focused on societal traits and dynamics (for distinction see Kwon, 2019). While both are relevant in our context, the society-based approach (which points to determinants such as history, equality and fairness in society, and national culture) is of special importance as it explicitly addresses the importance of the quality of formal government institutions and governance frameworks; that is, linking interpersonal and institutional trust through the dimension of personal experiences of government functions (Kwon, 2019, chapter 1; specifically on the importance of government performance for interpersonal trust see Van de Walle & Bouckaert, 2003; van Ryzin 2007).

To sum up, the Covid 19 pandemic has put the issue of trust at the centre of political analysis. This short review underlines the importance of looking into both (1) interpersonal and (2) institutional trust. Regarding the latter, the complex system of the modern state, which is accentuated in the case of crisis management, also underlines the importance of (3) differentiating between levels of analysis.

Based on previous research we expect the levels of institutional trust during the pandemic to vary in different ways:

• We have competing expectations when it comes to the effect of age on institutional trust. On the one hand older people have had more encounters with different institutions; that is, experience of welfare service production, which can be hypothesized to be positively correlated with trust. On the other hand they are more likely to be negatively affected by the pandemic.

• Education can be expected to have an effect on institutional trust since higher education leads to more knowledge about the functionality of the government. If functionality is good, then trust will be more positive. If not, then trust will be more negative.

• We expect higher income (and education) to have a positive effect on institutional trust since high income (and high education) is associated with higher socioeconomic standing and higher socioeconomic standing is associated with higher trust.

• Previous research leads us to expect gender to be of importance for institutional trust, to the effect that women have higher trust in the government than men do.

• We expect citizens standing to the left of the political spectrum having higher institutional trust, because of their ideological bias towards a big state and strong institutions.

• We furthermore expect people living in rural areas to have lower levels of trust in the central government than people in urban areas, due to the perceived greater distance to the central government. Conversely, we expect people in rural areas to have higher trust in the local government than people in urban areas, due to the relative proximity to municipal representatives.

• We expect trust levels to vary from the most dramatic part of the pandemic to the time when the situation has stabilized and the pandemic has become the ‘new normal’. During the ‘new normal’ we expect trust levels to return to levels closer to the situation before the pandemic.

These expectations inform our study about the variations in institutional trust. Referring to our discussion above about the inconclusive nature of current research on the links between institutional and interpersonal trust, we also have reasons to explore that issue further.

Before turning to our own empirical findings let us first recapitulate findings from previous studies on trust in Sweden.

3 PUBLIC TRUST IN SWEDEN

A note on the Swedish model of government is appropriate to contextualize various empirical findings on public trust in Sweden. Administratively the organization of government in Sweden spans three levels: central, regional and local. Reforms in recent decades have transferred important areas of responsibility (such as primary education and elderly care) to the local (municipal) level, which enjoys far-reaching, constitutionally based autonomy from the central government. Also, substantial parts of responsibilities for and execution of emergency preparedness, crisis management and societal/civil security reside with the 290 municipalities at the local level. The regional level is primarily concerned with medical and health care and the 21 different regions have thus come to occupy key positions in the Covid 19 crisis management (Pierre, 2020, p. 481). The central government level is characterized by small government ministries and large and relatively independent government agencies. In contrast to most other countries, policy implementation as well as expert knowledge reside in government authorities outside of rather than inside the government ministries (specifically in relation to Covid 19, see Johansson et al., 2021, p. 339; Kuhlmann, 2021, p. 560). This model of government yields particular precondition for steering (primarily through annual directives, budgetary resources and priorities, and appointment of heads of agencies, but prohibition of ministerial rule) (Hall, 2016; Jacobsson & Sundström, 2016).

In conclusion, the Swedish model of government comprises a decentralized system of governance with an organizational division of labour across levels of government, implying a constant need for policy coordination across levels. Moreover, it is a dual system of government with quite large, independent and strong government agencies. As Pierre (2020, p. 481) elaborates: ‘With autonomous agencies, regions and local authorities, central government policy-making institutions find themselves managing and coordinating a highly decentralized system’. From the perspective of public trust, this complex model can be expected to impact negatively on levels of public trust due to lack of clarity regarding execution of government tasks and a resulting expectations-responsibilities gap. Furthermore, this arrangement implies a potential for differentiated levels of trust vis-à-vis different public (and private) actors.

Along with the other Nordic countries, Sweden stands out as a high-trust society (Ortiz-Ospina & Roser, 2021). This holds in international comparison, but a more nuanced picture is warranted. Drawing on the most recent OECD data on ‘trust in government’ (the share of people reporting having confidence in their government), the figure for Sweden is 67.1%, in the vicinity of Denmark (71.6%), but notably lower than Finland (80.9%) and Norway (82.9%—the corresponding figure for Iceland is 59.2%) (OECD, 2021). Also, in an EU context we find that the Nordic societies have high levels of trust in their government, albeit with important differences. The most recent data shows that Denmark has the highest level of public trust in the government among EU Member states (78% say they trust their national government) while the corresponding figures for Finland and Sweden are 63% and 62% respectively (fourth and fifth place). The EU average stands at 40% (European Commission, 2020). In a temporal perspective, moreover, the Nordics are moving in different directions. Since 2006 (base year for the OCED surveys), trust in government in Norway has increased, with a significant rise taking place during the early years of the financial crisis. The other Nordic countries display a more complex picture. Finland and Denmark follow similar patterns (although trust levels in Finland initially were much higher than in Denmark)—sharp decreases in trust during the initial phase of the financial crisis, a further decrease around 2013 and then recovery in recent years. Sweden returns a combination of these patterns: increasing trust as the financial crisis started to take on, but a substantial decrease in trust reported in 2013 and then further decreases in recent years. Iceland, finally, holds a different picture: starting from rather low levels in 2010 (the first year of reporting), OECD data shows a dramatic increase in 2013 and then falling back in recent years. However, all five countries display significant increases in the most recent (2020) data (OECD, 2021).1

Looking specifically at Sweden, data on institutional trust in different societal institutions yields a rather complex picture of continuity and change (in part mirrored in the OECD data), but importantly also displaying significant differentiation among different societal institutions. Drawing on data from the SOM Institute (University of Gothenburg), since the 1980s trust in the central government has seen significant fluctuations over time, with an all-time high in 2010 when 52% of respondents reported very high or rather high trust in the central government; the latest available figure (2020) stands at 43%. In contrast, and an important observation in our context, is that the trust placed in the local government is substantially lower (in 2020, only 24% say they have high or rather high trust in the local government (Falk & Andersson, 2021, p. 46). It can be added that for the vast majority of years, the trust net balance for the local government is substantially negative (more answering that they place low or very low trust in the local government compared to those answering that they have high/very high trust in the local government). For the central government, the net balance varies considerably throughout the period but remains more positive than for the local government at all times except in 2005 (Martinsson & Andersson, 2021, p. 8).2

In one of the first studies on institutional trust in Sweden during the Covid 19 pandemic Peter Esaiasson with co-authors show how the general level of institutional trust in some agencies relevant for the handling of the pandemic actually increased. As the crisis took on, the Swedish citizens seemed to unite in their support for the government and their agencies (Esaiasson et al., 2020). Such a ‘rally around the flag’ moment has been observed elsewhere but is not a necessary crisis outcome, as Kritzinger et al. (2021) show. Subsequent studies on Sweden confirm the picture that despite a rather negative trend in public trust before the pandemic, during the Covid 19 pandemic institutional trust seemed to increase in Sweden. Falk. and Andersson (2021) show how public trust in the various levels of government and in a range of government authorities have remained or increased somewhat during the first two waves of the pandemic. Johansson et al. (2021) draw on the high levels of trust among Swedes in relation to government authorities to infer that this fundamental feature helps explain why the Swedes followed the advice of the government to such a high degree (see also Pierre, 2020). Kuhlmann et al develop this aspect further in their comparative study of French, German and Swedish crisis responses, concluding that ‘high levels of trust in Sweden regarding the relation between citizens and public authorities provide an important explanation for the fact that compliance was predominantly assured by recommendations’ (Kuhlmann et al., 2021, p. 561).

To conclude, Sweden has seen a negative trend in institutional trust during the last years. In contrary to what some previous research has implied, citizens' trust in institutions at the local level is significantly lower, compared to the trust in national institutions. Still, there are signs of an increase in institutional trust during the pandemic. These general trends should be kept in mind when we turn to our surveys on the level of trust that Swedish citizens hold of different actors' management of the Covid 19 crisis and how trust impacts on perceptions of Sweden's emergency preparedness for a future severe crisis or even war.

4 A NOTE ON METHODS

The empirical results reported here are based on the SOM survey on the Covid 19 virus and the Covid pandemic caused by the virus. The first survey was conducted during the spring of 2020 and a representative sample of Swedish citizens answered a wide range of questions related to political and societal aspects of the pandemic. In three waves during April–June a total of 6000 individuals were targeted, 44% of which responded to the survey (n = 2549).3 The rather low response rate is explained by the context of an acute pandemic. This context motivated a shorter field period since it was important to collect data during the same contextual circumstances. The downside of this is of course that a shorter time in the field results in a lower response rate. Still, the technical report shows that the sample is in general representative, but with a slight overweight for women, people in their mid-age and older people, and also people born in Sweden (Sandelin, 2020). This is a minor problem given our aim. Our statistical analysis takes the differences in response rate between different groups into account, meaning that the statistically significant differences should be seen as representative for the greater population.

The results reported draw on two specific questions asked within the SOM survey on the Covid 19 virus. In the first question, we asked the respondents: ‘To what extent do you trust the following actors' management of the Covid 19 pandemic?’ Here the respondents were asked to state the level of trust in a number of Swedish actors, ranging from citizens and civil society to authorities on local, regional, national and international levels. The respondents could indicate their trust on a scale with five degrees, from ‘very limited trust’, to ‘very high trust’. This construction of the question makes it possible to measure both the interpersonal trust as well as the institutional trust. It also gives us the opportunity to measure the differences in institutional trust between different levels of government.

In the second question, we asked the respondents: ‘How has the actions of the following actors during the Covid 19 pandemic affected your trust in Swedish civil emergency preparedness in the context of a severe crisis or war?’. Here the respondents could indicate the degree to which their trust was affected, from strongly negatively affected to strongly positively affected. The second question is thus forward looking in nature, giving us an opportunity to estimate how the experiences during the pandemic has affected the immediate trust in civil emergency preparedness in relation to future crises.

The results from the SOM survey on the Covid 19 crisis, conducted in the spring of 2020, are interesting in themselves. However, to evaluate the results, we need to compare the results to see if they are stable over time or only an expression of a specific effect in the midst of the outbreak of the pandemic. Therefore, we first include results from the annual standard SOM survey conducted during the fall of the same year (2020). This survey included a question identical to the one we had employed in the spring—‘To what extent do you trust the following actors' management of the Covid 19 pandemic?’—which actually gives us the opportunity to compare the levels of trust during the pandemic. The second survey was conducted in two waves during the period September–December 2020. 7500 people were targeted, of which 49% answered (net selection; n = 3534–3569) (see further Falk & Andersson, 2021).4

Secondly, we also include data from 2018 when we asked a representative sample of citizens in a panel study about their level of trust in different actors' civil emergency preparedness (Brommesson & Bengtsson, 2019). This survey did not include exactly the same question, but it still provides us with the opportunity to put the results from 2020 in perspective and to get a basic understanding of what we can understand as a normal level of trust in this case.

5 TRUST IN SWEDISH CIVIL EMERGENCY PREPAREDNESS DURING THE COVID 19 PANDEMIC

The combination of government complexity in general, Sweden's dual model of administration in particular, and the crisis context of Covid 19 all point in the same direction of public trust in government being challenged. What, then, characterized the nature of public trust in different actors’ crisis management in Sweden during the first wave of the pandemic and how does that relate to trust in emergency preparedness for a future severe crisis or even war?

The issue of institutional trust on different levels is particularly important in the case of emergency preparedness and crisis management in Sweden, since these are tasks situated with a number of different authorities and societal actors – subnational and national but also international in nature. For this reason, we asked the respondents about degrees of trust in relation to municipal, regional and central government as well as civil society, the corporate sector and the EU. It should be noted that regarding the different layers of government, we refer to the government at each level as an entity and do not differentiate between political executives and civil servants, nor between different government agencies (for further elaboration see the Discussion section below). To capture interpersonal trust, we also asked about trust in (fellow) citizens regarding their part of/role in crisis management. As is evident in Table 1, notable variation in trust levels feature in Swedish society, both during the outbreak of the pandemic during the spring of 2020 and later on during the fall of 2020. The actor category that enjoys the lowest degree of trust is the individual citizen herself (only 22.9% of respondents have high or very high trust in the individual citizen for managing the Covid 19 crisis during the spring of 2020, and similar levels of trust during the fall of 2020). This observation becomes principally relevant in light of Sweden's crisis management strategy, which emphasises recommendations rather than prohibition, and hence a belief in the responsibility and capacity of the individual.

Table 1. Trust in Swedish actors' management of the Covid 19 pandemic 2020 (percent)

	April –June		September–December
Actor	High trust	Net balance	High trust	Net balance
Fellow citizen	23	−3	23	−9
Municipality	39	+22	38	+21
Region	56	+44	54	+42
Central government	58	+47	49	+34
EU	31	+11	25	+4
Private businesses	30	+12	32	+15

• Note: The question posed read To what extent do you trust the following actors' management of the Covid 19 pandemic? Answers: Very high trust/high trust/neither high nor low trust/low trust/very low trust. In the table ‘high trust’ comprises alternatives ‘very high’ and ‘high’ trust. The net balance is the difference in percentage points between very high/high trust and very low/low trust. N = 2549 (April–June), 3534–3569 (September–December) (data regarding September-December retrieved from Falk & Andersson, 2021).

At the other end of the spectrum, among the actor categories under investigation here, the central government enjoys the highest degree of trust at the outbreak of the pandemic (57%), but decreasing to a second place (49%) during the fall. In itself, this may not come as a surprise, given the media exposure of ministers, officials and government agencies, among other things. However, the decreasing level of trust in the central government is interesting and important to note. As discussed above, some of the literature suggest we should expect a ‘rally around the flag effect’ in the beginning of the crisis, and the results reported here confirm this expectation.

The difference in trust levels concerning the central government reflects one kind of variation – temporal variation in trust regarding one and the same actor. Another variation, and a key result of our investigation, is the variation in trust among different public actors. Whereas the ‘regions’ (regional level of organization, responsible for medical care) are more or less on par with the central government (56% during the spring, and 54% during the fall), the municipalities hold considerably less trust (39% during the spring, and 38% during the fall). These findings concerning the municipalities are interesting and thought-provoking against the background of these entities carrying primary responsibility for the practical capacity and implementation of emergency preparedness, but also for the care of elderly people. The care of the elderly was heavily debated in Sweden during the first phase, since the pandemic spread within the nursing homes resulting in high mortality. Trust levels regarding the EU are also relatively modest (31%, decreasing to an even lower 25% in the fall), which can be hypothesized to reflect both an appreciation of the limited formal competence and capacity (beyond economic support) that the EU has regarding public health and crisis management, and a ‘national reflex’ regarding crisis management as a responsibility of the state (also evident in the unilateral and at times opposing crisis management strategies of different countries, in the EU and even in the Nordic context).

If we compare the results from the two 2020 surveys from the beginning and the midst of the pandemic with trust levels under more normal circumstances, we find similarities as well as differences. To begin with, comparing the data regarding the question we asked in 2020—To what extent do you trust the following actors' management of the Covid 19 pandemic?—with those regarding a question we posed in a different survey in 2018 (To what extent do you trust the following actors' civil emergency preparedness?) we find only small variations (Brommesson & Bengtsson, 2019). As a general conclusion, there is limited variation when comparing the data from a time of normality with the data collected in the midst of the pandemic. The strongest variation is seen in relation to the EU, with substantially higher level of trust in 2020. In all other cases the trust in 2020 is comparable to the trust in 2018. While the fact that different questions were asked in these surveys calls for cautiousness when interpreting the results, the relevance of the results are strengthened by similar positive conclusions regarding stable levels of trust, drawn by Esaiasson et al. (2020).

Moving on from these aggregate figures we find interesting variations based on gender and age (as shown in Table 2), but also income levels and ideological orientation. Generally speaking, women hold higher degrees of trust than men do, and trust increases with age. Furthermore, people leaning towards the left of the political spectrum and people with higher income are generally more trustful. Table 2 shows the results of our bivariate regression analysis of these and other variables. These findings are largely confirmed by the study by Falk and Andersson (2021) drawing on the SOM survey data from fall 2020. In parallel to our study, they find that there seems to be a modest age factor to the effect that older people (aged 65–85) are somewhat more trusting than younger groups. Also, people to the left of the political spectrum generally have a higher degree of trust than people to the right; the only exception regards trust in private companies, where people to the right, as can be expected, have higher degrees of trust. The authors did not find any substantive gender differences (see further Falk & Andersson, 2021, pp. 52–55).

Table 2. Trust in Swedish actors' management of the Covid 19 pandemic, April–June 2020

	Fellow citizens	Civil society organizations	Private businesses	Municipality	Region	County board	Government	EU
Gender	−0.014	−0.026	−0.09* *p < .05; **p < .01; ***p < .001.	−0.093* *p < .05; **p < .01; ***p < .001.	−0.190*** *p < .05; **p < .01; ***p < .001.	+0.185*** *p < .05; **p < .01; ***p < .001.	−0.238*** *p < .05; **p < .01; ***p < .001.	−0.338*** *p < .05; **p < .01; ***p < .001.
Age	+0.168*** *p < .05; **p < .01; ***p < .001.	+0.105*** *p < .05; **p < .01; ***p < .001.	+0.20	+0.02	+0.038	+0.043* *p < .05; **p < .01; ***p < .001.	+0.077*** *p < .05; **p < .01; ***p < .001.	−0.052* *p < .05; **p < .01; ***p < .001.
Education	+0.067*** *p < .05; **p < .01; ***p < .001.	+0.043*	−0.037	+0.023	+0.029	+0.009	+0.038	−0.034
Income	+0.066* *p < .05; **p < .01; ***p < .001.	+0.116*** *p < .05; **p < .01; ***p < .001.	+0.088** *p < .05; **p < .01; ***p < .001.	+0.057	+0.152*** *p < .05; **p < .01; ***p < .001.	+0.129*** *p < .05; **p < .01; ***p < .001.	+0.226*** *p < .05; **p < .01; ***p < .001.	+0.085* *p < .05; **p < .01; ***p < .001.
Urban/Rural	+0.031	+0.023	+0.009	+0.007	−0.012	−0.002	+0.063** *p < .05; **p < .01; ***p < .001.	+0.51
Left/Right	−0.007	−0.065*** *p < .05; **p < .01; ***p < .001.	+0.134*** *p < .05; **p < .01; ***p < .001.	−0.088*** *p < .05; **p < .01; ***p < .001.	−0.094*** *p < .05; **p < .01; ***p < .001.	−0.108*** *p < .05; **p < .01; ***p < .001.	−0.171*** *p < .05; **p < .01; ***p < .001.	−0.063** *p < .05; **p < .01; ***p < .001.

• Note: For the question posed, see Table 1. The table reports the unstandardized B-coefficient from the bivariate regression. Gender: Woman = 0, Man = 1. Age; four groups, 16–29 = 0, 30–49 = 1, 50–64 = 2, 65–85 = 3. Urban/Rural: Country side = 0, Smaller town (‘tätort’) = 1, Town or city = 2, Major city (Stockholm/Gothenburg/Malmö) = 3. Education: Low = 0, Rather low = 1, Rather high = 2, High = 3. Income: Maximum 300,000 SEK/year = 0; 301,000–700,000 = 1; more than 700,000 = 2. Left/Right: Clearly left = 0, Somewhat left = 1, Neither or = 2, Somewhat right = 3, Clearly right = 4. N = 2549.
• *p < .05; **p < .01; ***p < .001.

Moving on to our second question in the survey—how different actors' Covid 19 crisis management actions impact the trust placed in Sweden's emergency preparedness capacity—we find that the results from the spring of 2020 largely reflect the patterns of variations described above (see Table 3). Actors who enjoy the lowest levels of trust in the contemporary crisis are also the ones with very little or no positive effect on trust regarding future crisis management. The net balance (strengthened forward-looking trust vs weakened forward-looking trust) is positive for most actors, but only marginally so for the citizen actor category (+3) and, interestingly, virtually no effect regarding local government (+0,7). Here should also be noted the rather drastic negative net balance for the EU (−7,7); in sharp contrast to the positive effects for the central government (+20.6) and the regional government (+17.3). Notably, civil society enjoys the largest net balance of all actors at +27.1. Although the net balance is only marginally positive for some of these actors, we argue that the results could still be seen as supporting the rather positive results from our first question. Given the extreme situation during which the survey was conducted and the very hard criticism the Swedish Covid 19 strategy faced, we find the results more positive than what could be expected beforehand.

Table 3. Actors' management of the Covid 19 pandemic and the effect on the trust in Swedish civil emergency preparedness, April–June 2020

	Fellow citizens	Civil society organizations	Private businesses	Municipality	Region	County board	Government	EU
Gender	−0.086* *p < .05; **p < .01; ***p < .001.	−0.164*** *p < .05; **p < .01; ***p < .001.	−0,149*** *p < .05; **p < .01; ***p < .001.	−0.144*** *p < .05; **p < .01; ***p < .001.	−0.270*** *p < .05; **p < .01; ***p < .001.	−0.185*** *p < .05; **p < .01; ***p < .001.	−0.280*** *p < .05; **p < .01; ***p < .001.	−0.296*** *p < .05; **p < .01; ***p < .001.
Age	+0.122*** *p < .05; **p < .01; ***p < .001.	+0.075*** *p < .05; **p < .01; ***p < .001.	+0.034	+0.043* *p < .05; **p < .01; ***p < .001.	+0.044* *p < .05; **p < .01; ***p < .001.	+0.039	+0.112*** *p < .05; **p < .01; ***p < .001.	−0.032
Education	+0.55** *p < .05; **p < .01; ***p < .001.	+0.013	−0.005	−0.008	−0.003	−0.004	+0.002	−0.059
Income	+0.047	+0.102*** *p < .05; **p < .01; ***p < .001.	+0.077** *p < .05; **p < .01; ***p < .001.	+0.032	+0.091** *p < .05; **p < .01; ***p < .001.	+0.069* *p < .05; **p < .01; ***p < .001.	+0.180*** *p < .05; **p < .01; ***p < .001.	+0.031
Urban/Rural	−0.010	+0.015	−0.017	−0.004	+0.002	−0.013	+0.043	+0.017
Left/Right	−0.033*	−0.073*** *p < .05; **p < .01; ***p < .001.	+0.086*** *p < .05; **p < .01; ***p < .001.	−0.076*** *p < .05; **p < .01; ***p < .001.	−0.068*** *p < .05; **p < .01; ***p < .001.	−0.068*** *p < .05; **p < .01; ***p < .001.	−0.197*** *p < .05; **p < .01; ***p < .001.	−0.048* *p < .05; **p < .01; ***p < .001.

• Note: The question posed read How has the actions of the following actors during the Covid 19 pandemic affected your trust in Swedish civil emergency preparedness in the context of a severe crisis or war? The table reports the unstandardized B-coefficient from the bivariate regression. Gender: Woman = 0, Man = 1. Age; four groups, 16–29 = 0, 30–49 = 1, 50–64 = 2, 65–85 = 3. Urban/Rural: Country side = 0, Smaller town (‘tätort’) = 1, Town or city = 2, Major city (Stockholm/Gothenburg/Malmö) = 3. Education: Low = 0, Rather low = 1, Rather high = 2, High = 3. Income: Maximum 300,000 SEK/year = 0; 301 000–700,000 = 1; more than 700,000 = 2. Left/Right: Clearly left = 0, Somewhat left = 1, Neither or = 2, Somewhat right = 3, Clearly right = 4. N = 2549.
• *p < .05; **p < .01; ***p < .001.

Variations among different kinds of respondents return also here. Women's trust is developing positively more than for men. In a sense most striking, men's trust in Sweden's capacity for future crisis management has been negatively affected by the EU's handling of the crisis (net balance: −17.4), whereas women display no such EU effect (net balance: +0.4). Also, age is a key variable in forward-looking trust—the youngest age category is the one displaying the largest decrease in trust as a result of the management of the current crisis. Notably, this holds for all actors with one exception—the EU.

In Table 3 we present our regression analysis for the second survey question. As can be noted, we find significant effects of gender, age, income, and to some extent also ideological orientation, in relation to the impact of current crisis management on future emergency preparedness. If we were to combine all the variables having significant effect on the trust in Swedish emergency preparedness capacity, the ideal type of such a person would be that of an older well-educated woman, standing to the political left while still having a high income.

6 DISCUSSION

The preceding analysis yields four key findings. First we see rather weak levels of interpersonal trust. As discussed above, Sweden is often seen as a country characterized by a high degree of trust in general. However, during the spring of 2020 when trust was put to test with regard to how fellow citizens dealt with the Covid 19 pandemic, the trust in fellow citizens' management of the pandemic is notably weak, also compared to the trust held in not always very popular institutions. This somewhat unexpected outcome, which runs counter to that of Esaiasson et al. (2020) at the very early stage of the pandemic, speaks to the discussion above about factors influencing interpersonal trust. As for individually-oriented explanations, the outcome may be explained by personal experiences of others' behaviour as well as expectations about to what degree others will take responsibility in times of crisis. Bringing the discussion into society-based explanations, existing research points to the link between interpersonal and institutional trust in which case the performance of public institutions come into focus, pointing towards issues of fairness and impartiality in personal experiences of government performance (see Kwon, 2019 for elaboration). Some of the variation that we have established, for instance regarding how trust relates to age, may be understood in this light. The findings, however, also begs a more general question, not least in welfare states with a big government such as Sweden, of how people perceive personal responsibility in relation to the responsibilities of the state. Not being used to think in terms of personal responsibility and sacrifice may arguably correspond to low trust in others' willingness and capacity to contribute to crisis management.

This leads us to our second key finding, a surprisingly strong institutional trust. Despite the very troubling circumstances during the first wave of the pandemic, with a widespread virus, high infection rates, high number of citizens in intensive care and tragically enough also a high mortality, the institutional trust does not seem to have been negatively affected to any greater extent. This is true also in the forward-looking perspective, based on our question about how the management of the pandemic has affected the trust in the management of future crises. These results must be seen as an indication of a relatively strong institutional trust, compared to the situation under more normal circumstances. Previous results on public trust in Sweden during the pandemic (Esaiasson et al., 2020) is in this part confirmed. This finding relates to the scholarly discussion referred above regarding the relationship between crisis and trust. In short, our observations can be interpreted as trust not being challenged by the crisis but rather reinforced in times of crisis (see Kritzinger et al., 2021 for a similar conclusion regarding Austria).

As a third finding, our results display considerable variations between different groups. Here, we have shown how female gender, higher age, political positions to the left and high income (and to some extent high education) have a positive effect on the levels of trust. In general, our expectations based on a review of existing research (as summarized in theory section above) are confirmed. The observed variation holds also for our forward-looking question where we see the same groups having higher levels of trust in the crisis management preparedness for future crises. These findings are of special relevance in the Swedish case, due to the ongoing investments in a modern civil defence in Sweden, based on the existing crisis management capacity.

Finally, as a fourth finding, we have found interesting variations among institutions. More distant levels of government, such as the central government and the regions, have maintained high levels of trust, while the local government within the municipalities display considerably lower levels of trust. Importantly, the proposed rural/urban differentiation yields no significant results in this context. This finding is of principal relevance as it speaks to the debate about competing expectations elaborated above regarding the importance of the distance between the individual and various government levels. As such, it also relates to the discussion above regarding our first finding, as it is primarily at the municipal level that people encounter the government (services). The finding is however also of particular empirical interest due to the Swedish model of government that comprises a decentralized system of governance (in the context of Covid 19, see Pierre, 2020). As we have argued above, this model implies a constant need for policy coordination across levels and it is obvious that this coordination did not function during the first wave of the pandemic (Brommesson, 2021). It can be debated if the municipalities are to be blamed for this, or if the failure to protect the elderly living in homes run by the municipalities was a result of a lack of support from the central national level. Still, the failure that played out on the local arena most likely affected the citizens' view of the local government capacity, while the local government legitimacy was lower already at the outset of the pandemic (as discussed above).

As a critical reflection, it should be noted in this context that our data does not allow for distinguishing between political executives and civil servants at different layers of government; i.e. we do not know if and how different parts of government at a given level are trusted to different degrees. Similarly, our data does not allow for exploration of (potential variation in) trust in individual government agencies (such as the Public Health Agency vs. the National Board of Health and Welfare or the Swedish Civil Contingencies Agency in the context of Covid 19). Recent studies (such as Falk & Andersson, 2021; Martinsson & Andersson, 2021; Medieakademin, 2021) lead us to believe that such variations may exist.

7 CONCLUDING REMARKS

In this article we set out to explore interpersonal and institutional trust in Sweden during the Covid 19 pandemic, with a special focus on the first wave (spring 2020), set in comparison to the second wave (fall 2020). As has been elaborated above, levels of interpersonal trust are distinctly lower than trust in various other actors (public and private alike). This observation warrants further research against the background of the common assumption that Sweden is high-trusting country. As for institutional trust, we take note of the considerably higher levels of trust, especially on the national level. We see a potential ‘rally around the flag’ effect. Here, we believe future research has reason to reflect on the role of the frequent press briefings, speeches by the Swedish King and the Prime Minister, and the public role of the state epidemiologist Mr Anders Tegnell who during the first wave soon became a uniting figure in Sweden, but also helped shape a discourse others have described in terms of monotony; ‘where authoritative voices offer credible reasons for one policy option only’ (Baekkeskov et al., 2021). Even if several of these authorities where criticized later on, it is apparent how they managed to secure the public's trust in the Swedish authorities, or government legitimacy despite possibly low government capacity, in the midst of a sudden and unexpected crisis, and establish the daily press briefings as a national bonfire to gather around and where people could rally around the flag. Here, Swedish citizens resorted to a form of blind trust based on a state-oriented risk culture according to which government institutions ‘administer disaster prevention, mitigate ongoing crises, and […] guide and inform citizens…’ (Johansson et al., 2021, p. 339, based on Cornia et al., 2016). This culture is obviously dependent on a high degree of trust and the state determines the relevant risks and appropriate measures. Based on such a state-oriented risk culture, the trust was apparently high enough for the government to maintain its exceptional approach (based on voluntary measures, saying no to masks, and expressing a will to keep the society open) to counter the pandemic. When the initial crisis turned into a daily routine during the second wave, the trust decreased and a more open debate concerning different measures to lower the infection rates (e.g. wearing masks, closing schools, etc.) could take place.

ACKNOWLEDGEMENT

The research reported in this article was made possible by a grant from the Swedish Civil Contingencies Agency (MSB) grant agreement 2017-99591, which is hereby gratefully acknowledged.

CONFLICT OF INTERESTS

The authors declare that there are no conflict of interests.

ENDNOTES

1 Country-specific data on trust in the government (%): Norway 2006: 61.0 (lowest)—2010: 66.0—2013: 64.5—2020: 82.9 (highest); Denmark 2006: 64.0—2010: 53.0—2013: 47.4 (lowest)—2020: 71.6 (highest); Finland 2006: 74.0—2010: 54.3—2013: 48.3 (lowest)—2020: 80.9 (highest); Sweden 2006: 52.8 (lowest)—2010: 62.3—2013: 54.7—2020: 67.1 (highest); Iceland 2010: 26.0 (lowest)—2013: 44.5—2020: 59.2 (highest) (OECD, 2021).

2 A second key data source on trust in Sweden is the ‘trust barometer’ provided by Medieakademin. Their results largely follow what has been outlined above. While unfortunately not providing data for local government, the source contains relevant data in relation to another aspect about variation in trust and differentiation among institutions—data shows great variation among different government agencies (Medieakademin, 2021).

3 For a more detailed technical report see Sandelin (2020).

4 For a more detailed technical report see Falk et al. (2021).

Open Research

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

Baekkeskov, E., Olivier, R., & Öberg, P. O. (2021). Monotonous or pluralistic public discourse? Reason-giving and dissent in Denmark's and Sweden's early 2020 COVID-19 responses. Journal of European Public Policy, 28, 1321– 1343. https://doi.org/10.1080/13501763.2021.1942158

Bengtsson, R., Brommesson, D., & Edström, Z. (2020). Civil beredskap i coronatider: Förtroendet störst för staten och för regionerna. SOM-undersökningen om coronaviruset 2020. SOM-institutet.

Berg, M., & Johansson, T. (2020). Building institutional trust through service experiences: Private versus public provision matter. Journal of public administration research and theory, 30, 290– 306. https://doi.org/10.1093/jopart/muz029

Brommesson, D. (2021). Civil beredskap och offentlig styrning under pandemin 2020-2021 [Civil preparedness and public management during the pandemic of 2020–2021]. Statsvetenskaplig tidskrift, 123, 141–158.

Brommesson, D., & Bengtsson, R. (2019). Staten som kommunikatör. En panelstudie av effekterna av kampanjen Om krisen eller kriget kommer. Kungl Krigsvetenskapsakademiens Handlingar och Tidskrift 4/2019.

Christensen, T., & Lægreid, P. (2005). Trust in government. The relative importance of service satisfaction, political factors, and demography. Public Performance and Management Review, 28, 487– 511.

Christensen, T., & Lægreid, P. (2020). The coronavirus crisis—Crisis communication, meaning-making, and reputation management. International Public Management Journal, 23, 713– 729. https://doi.org/10.1080/10967494.2020.1812455

Cornia, A., Dressel, K., & Pfeil, P. (2016). Risk cultures and dominant approaches towards disasters in seven European countries. Journal of Risk Research, 19, 288– 304. https://doi.org/10.1080/13669877.2014.961520

Devine, D., Gaskell, J., Jennings, W., & Stoker, G. (2021). Trust and the coronavirus pandemic: What are the consequences of and for trust? An early review of the literature. Political Studies Review, 19, 274– 285.

Erlingsson, G. Ó., Bergh, A., & Sjölin, M. (2008). Public corruption in Swedish municipalities—Trouble looming on the horizon? Local Government Studies, 34, 595– 608. https://doi.org/10.1080/03003930802413780

Esaiasson, P., Martinsson, J., & Sohlberg, J. (2016). Flyktingkrisen och medborgarnas förtroende för samhällets institutioner—en forskarrapport. Statsvetenskapliga institutionen, Göteborgs universitet.

Esaiasson, P., Sohlberg, J., Ghersetti, M., & Johansson, B. (2020). How the coronavirus crisis affects citizen trust in government institutions and in unknown others—Evidence from “the Swedish Experiment”. SocArXiv. https://doi.org/10.31235/osf.io/6yw9r

European Commission. (2020). Standard Eurobarometer 93. https://europa.eu/eurobarometer/surveys/detail/2262

European Commission. (2021). Standard Eurobarometer 94. https://europa.eu/eurobarometer/surveys/detail/2355

Falk., E., & Andersson, U. (2021). Förtroende för samhällets och pandemins aktörer. In U. Andersson, A. Carlander, M. Grusell, & P. Öhrberg (Eds.), Ingen anledning till oro (?) (pp. 43– 56). SOM-institutet.

Falk, E., Sandelin, F., & Weissenbilder, M. (2021). Den nationella SOM-undersökningen 2020—En metodöversikt. SOM-institutet.

Fitzgerald, J., & Wolak, J. (2016). The roots of trust in local government in western Europe. International Political Science Review/Revue internationale de science politique, 37, 130– 146.

Hale, T., Webster, S., Petherick, A., Phillips, T., & Kira, B. (2021). Oxford COVID-19 government response tracker. https://www.bsg.ox.ac.uk/research/research-projects/coronavirus-government-response-tracker

Hall, P. (2016). The Swedish administrative model. In J. Pierre (Ed.), The Oxford handbook of Swedish politics (pp. 299– 314). Oxford University Press.

Hall, P., Löfgren, K., & Peters, G. (2016). Greening the street-level procurer: Challenges in the strongly decentralized Swedish system. Journal of Consumer Policy, 39, 467– 483. https://doi.org/10.1007/s10603-015-9282-8

Jacobsson, B., & Sundström, G. (2016). Governing the state. In J. Pierre (Ed.), The Oxford handbook of Swedish politics (pp. 347– 362). Oxford University Press.

Johansson, B., Sohlberg, J., Esaiasson, P., & Ghersetti, M. (2021). Why Swedes don't wear face masks during the pandemic—A consequence of blindly trusting the government. Journal of International Crisis and Risk Communication Research, 4, 335– 358.

Johns Hopkins. (2021). Johns Hopkins Corona Resource Center. https://coronavirus.jhu.edu/map.html

Jørgensen, F., Bor, A., Fly Lindholt, M., & Bang Petersen, M. (2021). Public support for government responses against COVID-19: Assessing levels and predictors in eight Western democracies during 2020. West European Politics, 44, 1129– 1158. https://doi.org/10.1080/01402382.2021.1925821

Kantar Sifo. (2021). Insikter kring coronavirus. https://www.kantarsifo.se/insikter-kring-coronavirus

Keele, L. (2007). Social capital and the dynamics of trust in government. American Journal of Political Science, 51, 241– 254.

Kritzinger, S., Foucault, M., Lachat, R., Partheymüller, J., Plescia, C., & Brouard, S. (2021). Rally around the flag’: The COVID-19 crisis and trust in the national government. West European Politics, 44, 1205– 1231.

Kuhlmann, S., Hellström, M., Ramberg, U., & Reiter, R. (2021). Tracing divergence in crisis governance: Responses to the COVID-19 pandemic in France, Germany and Sweden compared. International Review of Administrative Sciences, 87, 556– 575.

Kwon, O. Y. (2019). Social trust and economic development: The case of South Korea. Edward Elgar.

Lee, K. (2009). How the Hong Kong government lost the public trust in SARS: Insights for government communication in a health crisis. Public Relations Review, 35, 74– 76.

Lipscy, P. Y. (2020). COVID-19 and the politics of crisis. International Organization, 74(Suppl.), E99– E127.

J. Martinsson, & U. Andersson (Eds.). (2021). Svenska trender 1986-2020. https://www.gu.se/sites/default/files/2021-04/1.%20Svenska%20trender%201986-2020_korrigerad%202021-04-26.pdf

Medieakademin. (2021). Förtroendebarometern 2021. Allmänhetens förtroende för institutioner, politiska partier, massmedier & företag. https://medieakademin.se/wp-content/uploads/2021/03/Presentation_fortroendebarometern_2021-1.pdf

Nygren, K. G., & Olofsson, A. (2020). Managing the Covid-19 pandemic through individual responsibility: The consequences of a world risk society and enhanced ethopolitics. Journal of Risk Research, 23, 1031– 1035. https://doi.org/10.1080/13669877.2020.1756382

OECD. (2017). Trust and public policy. How better governance can help rebuild public trust. OECD.

OECD. (2021). OECD data: Trust in government. https://data.oecd.org/gga/trust-in-government.htm

Ortiz-Ospina, E., & Roser, M. (2021). Trust. Our world in data. https://ourworldindata.org/trust

Pearson, C. M., & Clair, J. A. (1998). Reframing crisis management. The Academy of Management Review, 23, 23– 76.

Pierre, J. (2020). Nudges against pandemics: Sweden's COVID-19 containment strategy in perspective. Policy and Society, 39, 478– 493.

Putnam, R. (1993). Making democracy work: Civic traditions in modern Italy. Princeton University Press.

Rothstein, B. (2005). Social traps and the problem of trust. Cambridge University Press.

Sandelin, F. (2020). SOM-undersökningen om coronaviruset 2020—En metodöversikt. SOM-institutet.

Sandstig, G., & Dahlin Ivanoff, S. (2020). Kapabilitet under en pandemi—hur allmänheten och riskgrupper bedömer tilltron till sin egen förmåga att hantera coronavirusets konsekvenser. SOM-undersökningen om coronaviruset 2020. SOM-institutet.

Spadaro, G., Gangl, K., Van Prooijen, J.-W., Van Lange, P. A. M., & Mosso, C. O. (2020). Enhancing feelings of security: How institutional trust promotes interpersonal trust. PLoS One, 15, e0237934. https://doi.org/10.1371/journal.pone.0237934

Van de Walle, S., & Bouckaert, G. (2003). Public service performance and trust in government: The problem of causality. International Journal of Public Administration, 26, 891– 913. https://doi.org/10.1081/PAD-120019352

Van Ryzin, G. G. (2007). Pieces of a puzzle. Linking government performance, citizen satisfaction, and trust. Public Performance & Management Review, 30, 521– 535. https://doi.org/10.2753/PMR1530-9576300403

Wang, H., Cleary, P. D., Little, J., & Auffray, C. (2020). Communicating in a public health crisis. The Lancet Digital Health, 2, e503.

Welch, D. A. (2021). The ontology of international crisis. Oxford encyclopedia of politics. Oxford University Press.