Deviant Swedes in the Global Covid-19 Media Environment

ABSTRACT During the early COVID-19 pandemic, many countries took compulsory measures to combating the virus’s spread, while Sweden took a more voluntary approach. This led to polarized reactions in the international media, with some praising it and others proclaiming it disastrous. Using the concept of “moral panic” I examine how actors within the global media environment portrayed Swedes as a deviant population, using persuasive language, exaggeration, and selective reporting, and how an amplification of media attention served to solidify the deviant label. I also argue that Sweden was made deviant partly to justify restrictive measures in other countries.

Although Sweden was reported as "refusing" to lockdown, in fact many measures were put in place to limit the spread of disease. In comparison, however, many other counties' restrictions during the early pandemic were further reaching than those taken in Sweden. These "lockdowns"often unprecedented restrictions on civil liberties and livelihood -needed to be justified as the new norm, through framing them as based on science and as moral practices taken to protect the population from an emerging virus. Sweden posed a moral threat to the newly established norm of "locking down," and this deviance needed to be highlighted and criticized to further justify lockdowns in other countries.
Sweden could also be framed as deviant because it is both familiar and unknown. Sweden is a complex country, with challenges and social problems like any other place. Yet, those without a direct connection to the country often have a one-dimensional view of it based on success, as the country performs very well on nearly all global rankings: quality of life, competitiveness, transparency, governance, education and other measures (Rapacioli 2018). Because of "the world's preconceptions about [Sweden] as some sort of paradise" it is easy to set up stories about how the world's perfect society has collapsed, often with "great dollops of schadenfreude (Rapacioli 2018: Chapter 1)." These kinds of "how-the mighty-have-fallen" make for good news stories (Rapacioli 2018;Hall et al. 1980;Ericson et al. 1987). This relative unfamiliarity with Sweden also meant that the international media could portray a simplified, and sometimes factually inaccurate, version of the Swedish approach because their audiences were not familiar enough with the country to contextualize or question the reporting. This only served to solidify the deviant label.
In the sections that follow I discuss how deviance is constructed in the media and my methods for studying the global media environment. I then describe the Swedish approach and how it was juxtaposed against "lockdowns," serving not only to frame the Swedes as deviant but also to further establish "lockdowns" as best practice. I then discuss the dominant media framings: good Swedes and deviant -or exotic, stupid, naïve, and arrogant -Swedes, and the way in which the media used selective and exaggerated reporting to "punish" Sweden, thereby solidifying the deviant label.

Deviants and the news media
In the study of infectious disease, anthropologists have long been interested in how dominant and alternative narratives, including "rumors, conspiracy theories and propaganda" are created and circulate, and how different actors interpret information from various media sources (Ali 2020:2;Briggs and Nichter 2009). For instance, Merrill Singer noted how geographies of blame were constructed through media headlines and statements about "illegal immigrants" bringing H1N1 to the United States (Singer 2009). Using the example of SARS in 2003, sociologist Stephen L. Muzzatti examined how the media manufactures public health threats, often drawing upon cultural myths of dangerous others (Muzzatti 2005). Similarly, Briggs and Nichter highlight the example of Mexico during the H1N1 outbreak to note how the media "evoked cultural accounts" to explain the case fatality ratio, in which "self-medication and delaying visits to appropriate health providers" became a "fault of national character" (Briggs and Nichter 2009: 192). In these examples, the notion of deviance is explicit, or at least implicit.
Dictionary definitions of deviance focus on non-conformity to an accepted standard. Although not inherently negative, the term has become associated with behaviors, beliefs, appearances, and forms of expression that violate social norms (Jensen 2011). In examining "deviant Swedes" and the media, I draw on the work of sociologists Erich Goode and Nachman Ben-Yehuda who, in turn, base their research on sociologist Stanley Cohen's theories around deviance and moral panics.
Cohen, writing in the context of youth riots in the UK in 1960s, examined how the media exaggerated the extent of the violence both causing and reflecting a wider "moral panic." To Cohen, a moral panic is a scare over a real or imagined threat from "folk devils," or deviants who supposedly pose a threat to societal values and morals. Within this framework deviance is constructed: "any claim that some act or condition is deviant depends upon a set of shared, subjective understandings about the nature of the social order" (Best 2011: 17). Deviance is relational as it is defined in contrast to that which is normal. It is subject to penalty or punishment because it poses a threat to the established social and moral order (Henry 2009:3;Goode and Ben-Yehuda 2009;Cohen 1972).
Deviance also requires an audience (Henry 2009;Goode 2014;Goode and Ben-Yehuda 2009). Someone must observe or be told about a deviant act, and then determine whether or not the behavior is deviant. It is often through the media that we become aware of a deviant population and in this way the media acts as a mediator between "the deviants" and their audience. The media utilize a number of strategies: exaggeration, disproportional attention to the "moral panic" in relation to the harm it causes, and the use of sensationalist and melodramatic language. Overall, the media serves to "visualize" deviance, to draw attention to harm and, through framing, create a certain way of thinking about the deviant that becomes taken-for-granted and the normal state of things (Goode and Ben-Yehuda 2009: 89-95;Chermak 2002).

Ethnography in the covid-19 global media environment
In March 2020 I began writing a diary as part of the Corona Diaries Project coordinated by the German medical anthropology journal Curare. In it, I documented my daily life in the pandemic, much of which revolved around consuming news and social media about COVID-19. From this perspective, I found that the international media's portrayal of the "Swedish strategy" was very different than my lived experience in Sweden. I then decided to explore the question of why there was so much attention paid to, and at times misleading information about, Sweden and COVID-19.
I initially intended an ethnography of a news cycle: how Sweden's approach to COVID-19 came to the attention of international media during the first month of the pandemic and the interplay between the national debates and international news coverage. However, the news cycle never ended. Instead, the international media followed every twist and turn, making stories out of minor or off-hand statements by Swedish officials. This is similar to Carlo Caduff's discussion of Mary Ann Doane's discursive problematic: the news needs "to be continuously created and maintain a sense of newness." (Caduff 2018: 88). In this way it "promotes its own consumption" and is "incapable of closure" (Caduff 2018: 89); this is even more the case when a deviant phenomenon develops in the media (Garland 2008). Although the media coverage continued throughout the pandemic, for practical reasons, I ended data collection in mid-July 2020, four months into the pandemic and roughly corresponding with the end of Sweden's "first wave." I approached the field with inspiration from John Postill and Sara Pink's social media ethnography. This involves five overlapping sub-practices or routines: catching up, sharing, exploring, interacting, and archiving (Postill and Pink 2012). Extending this to global media environment at large, every morning I caught up with Swedish and international news and Twitter. I then explored by following the links I found in tweets or on news websites which took me to YouTube videos, blogs, other international media outlets, and websites. Although I do not present the Twitter material here, I mention it because I used Twitter to identify which articles were being shared. I very rarely am active on social media, so I did not share much content, but I occasionally forwarded articles via e-mail. I interacted, periodically corresponding with journalists and coauthoring an opinion piece in a Swedish newspaper during the first wave; I also discussed my research informally with colleagues and neighbors. Finally, I archived by creating a folder with news and social media. This included 333 screen shots and 922 news articles. I also had a 33-page fieldnote document (8000 words) with links and transcriptions of news and the aforementioned diary of about 10,000 words.
I started my day with the Swedish morning papers: Dagens Nyheter, Svenska Dagbladet, Sydsvenskan and the afternoon tabloids Expressen and Aftonbladet. Dagens Nyheter, Sydsvenskan are both owned by the Bonnier News and are considered independent liberal; Dagens Nyheter is the largest paper in Sweden while Sydsvenskan is the main paper for southern and central Skåne, Sweden's southernmost region which includes the third largest city of Malmö. Dagens Nyheter has largely taken a stance critical of the government and the Public Health Agency's handling of the pandemic and has itself been criticized for publishing factually inaccurate and speculative articles and opinion pieces (Von Schreeb 2020). Svenska Dagbladet is a national paper which leans toward the moderate right. Aftonbladet is independent but moderate-left leaning, whereas Expressen is independent liberal.
In terms of the international press, I skimmed through the Guardian (UK) and CNN.com, both mainstream left, and BBC News (politically neutral). I then used Twitter to identify which articles were being shared: this included a range of primarily English-language articles from outlets with a global audience, including Al-Jazzera, L'Agence France-Press (AFP), Bloomberg, Deutsche Welle, The New York Times, Vox and others which I present in the results section. I regularly listened to news programs on BBC Radio 4 and Sweden's public service radio, Sveriges Radio (politically neutral). However, what is considered "left" and "right" is largely dependent on country, context, and time period, and public service broadcasting is often critiqued for real and imagined bias. I assign political categories to these news outlets as only as a general guide to orient the reader.
During the first month I watched the WHO's press briefings every other day and COVID-19-related news programs on Sweden's public access channels, using the on-demand service SVTplay.se. This included daily briefings with the Public Health Agency of Sweden (Folkhälsomyndigheten, FHM). The daily briefings at 14.00 became a focal point for Swedish society, described as a "ritual and a gathering point, a campfire" during which representatives from FHM presented the day's statistics and explained the principles of epidemiology, followed by statements from other related agencies, a question-andanswer session, and individual interviews with media outlets (Johansson in Nordensson 2020). One journalist went so far to describe it as the Swedish version of mass, with the state epidemiologist, Anders Tegnell, as our "new high priest" (Mellin 2020).
I analyzed the material using mixed approaches from media studies, grounded theory, thematic analysis, and rhetorical analysis (Floersch et al. 2010;Braun and Clarke 2006;Orgad 2012;Winton 2012) Casting a wide net, I started with mapping how the Swedish approach was described and immersed myself within this environment. I identified the main themes in the material, which was largely polarized into articles in favor of or against the "Swedish approach," and the sub-themes which articulated why the approach was good or bad. In this step, the veracity of the reporting was also a factor: was an article against or in favor of the "Swedish approach" or what they thought the "Swedish approach" was, and which aspects of everyday Swedish life was highlighted, exaggerated, or ignored? For example, most articles were accompanied with photos of people in cafes or on pedestrian streets, not photos of people working from home or isolating.
From rhetorical analysis I was interested in the intended audience. For example, left-wing outlets tended to focus on moral failings of Sweden to prevent deaths, while right-wing outlets were more concerned with restrictions to civil liberties. I also analyzed the use of language and argument to persuade the reader, as well as the use of specific rhetorical devices, such as headlines posing questions with implied answers such as: Could Sweden's risky coronavirus strategy work? (Al-Jazeera, 5 th May) Is Sweden's approach to covid-19 wise or reckless? (The Economist, 17 th May) At this point in the analysis, I started to think about "deviance" and the applicability of Stanley Cohen's "moral panics" became apparent: the use of sensationalized vocabulary, exaggeration and selective reporting are key methods of constructing deviance in the media (cf. Cohen 2002: xix;Goode and Ben-Yehuda 2009: 30-31).

The Swedish approach
Sweden is everyone's favorite example of a country that got it wrong. Consensus is building that its laissez-faireherd immunity strategy has led to unnecessary levels of illness and death among vulnerable populations (Loder 2020, Editorial in the British Medical Journal) During spring and early summer 2020, voluntary measures in Sweden, directed at individuals, focused on staying home if sick, handwashing, working from home, avoiding public transport if possible, and generally avoiding other people. These measures applied to everyone, but especially to those over 70. Compulsory and semi-compulsory measures were directed at businesses and public institutions. Senior high schools and universities were recommended to move to distance learning, public events with more than 50 people were banned and further guidance, including binding recommendations, were laid out for shops and public transportation operators to minimize crowding. In this context the word "recommendation" requires nuance: while universities were "recommended" to move to distance learning, it was not an acceptable option for universities to continue large-scale in-person learning. Similarly, a "recommendation" directed at individuals were not really voluntary and its violation was subject to social punishment.
Much of the reporting stated or assumed that that Sweden had a "herd immunity strategy," based on the premise that Sweden was staying open so that as many people as possible would be infected, thus leading to immunity at the population level. The "evidence" for this strategy during the first wave came from a Facebook post by the previous state epidemiologist and a YouTube video posted by a Swedish mathematician, with the latter claiming it was the Swedish strategy. Other critics of the government published a series of opinion pieces in Dagens Nyheter further forwarding this theory without evidence. At the same time, Swedish officials discussed publicly the role population immunity could play in slowing infections, and later e-mails showed the concept has been discussed internally as well (Karlsten 2020). However, at the daily press conference on the 23 rd of April, Deputy State Epidemiologist Anders Wallsteen had warned that there was much we did not know about immunity, not least how long it would last. He also warned that achieving herd immunity would not necessarily change the need for physical distancing measures, because of the uncertainties around testing and interpreting immunological tests.
Herd immunity became a red herring which deflected from more concrete reasons for Sweden's approach, such as its legal framework. Swedish law provides the government with very few possibilities to suspend civil rights and freedoms (Jonung 2020;Klamberg 2020;Nordberg and Mattson 2020). Constitutionally, Swedish citizens have an absolute right to enter and leave the country, and to travel around the country so they could not legally be prevented from traveling, although during the first wave people were asked to avoid unnecessary journeys. Moreover, the Foreign Ministry advised against unnecessary foreign travel, a recommendation which invalidates many travel insurance policies (and many other countries' borders were closed anyhow). This stands in stark contrast to, for example, the United Kingdom where people were banned (with the threat of fines) from traveling both inside and outside the country without a "valid reason" during much of the pandemic. Also enshrined in the Swedish constitution is the right of public access (Allemansrätten) which ensures access to nature, meaning it is extremely difficult to limit access to a beach or nature reserve. (A pandemic law passed in January 2021 implies that a municipality can shut down a large gathering at a public beach, but people still have the constitutional right to walk across the beach).
Under the Communicable Diseases Act, individuals or small areas (i.e., a street or a few blocks) can be placed under quarantine but the law is not intended for large scale societal closures. Beyond the constitution and the Communicable Diseases Act there are many other seemingly unrelated, but applicable laws. For example, in December 2020 the government tried to shut local libraries, but under the Library Act (2013), libraries are classed as "essential" services and cannot be closed. In practice, the running of libraries is largely up to local government and most had been running a limited or adapted service already.
This said, the government could have taken more measures during the spring, although the extent and nature of these hypothetical measures is debatable (Orange 2020). Under chapter 2, paragraph 24 of the Swedish constitution (Regeringsformen) the freedom of assembly can be limited "to combat an epidemic" and the Public Order Act could have been used to further limit gatherings, as it was later in Autumn 2020 when public events of more than eight were banned and limits were placed on alcohol sales in restaurants after 8pm.
Other contextual aspects were used by the government and civil servants to explain the use of voluntary measures. FHM and its representatives discussed tradition: "Swedish public health work builds on a strong tradition of voluntary [measures] with emphasis on the individual's responsibility (FHM 2020)." Sweden also took a whole-of-society approach. That is, there was a concern over the damage to public health of closing society, such as domestic abuse, loneliness, and the health impact of mass unemployment (FHM 2020; Liman and Rolander 2020). The strategy also aimed for sustainability, with the government and authorities taking the perspective that this was "not a sprint, but a marathon" and that the Swedish people needed measures that could be accepted for the long haul (SVT 2020a(SVT , 2020b.
Four months into the pandemic, Sweden had 523 deaths per million, significantly more than other Scandinavian countries -Denmark was 104, Norway 47 -but fewer than Belgium, the United Kingdom, Italy and Spain, according to figures collated by Johns Hopkins University.
While the international media focused on Sweden as a whole, the disease was not spread evenly across the country. Skåne, the region which includes Malmö and borders Copenhagen, had significantly fewer deaths than Stockholm, which was the epicenter of Sweden's epidemic during the first wave. According to figures from Denmark's Statens Serum Institute and FHM, by mid-July 2020 Skåne had 18.7 deaths per 100,000, similar to Copenhagen's 19, and much less than Stockholm's 99.
There were many theories as to why Sweden had more deaths than other Nordic countries. For instance, schools in Skåne had had their winter break during the week before Stockholm; this mean that Stockholmers had traveled to the Alps while there was more virus spreading there. Oslo and Copenhagen also had their winter breaks the same week as Skåne, lending further credence to the theory (Falk et al. 2020). There were also differences in demographics, preparedness, and health care organization which impacted the numbers, the discussion of which are outside the scope of this article. The point here, which was often missed in the media, is that Scandinavian countries are unique jurisdictions and that analyzing and evaluating policy is a complex task, not a media soundbite.

Constructing deviancy
In the week that followed the WHO's decision to call Covid-19 a pandemic (March 11 th ), Nordic media, followed by German media began to take an interest in Sweden. The following week the Guardian (UK) published the article: "Swedish PM warned over 'Russian roulette-style' Covid-19 strategy: Health experts 'deeply concerned' about Sweden's response." The piece discussed a group of experts based in Sweden who disagreed with the government's response, and was a catalyst for a media storm which, in turn fed back into the national debate. In the international media the "Swedish approach" drew headlines such as: The headlines, as well as the content of the articles, juxtaposed Sweden's approach with that of other countries: "lockdown" as normal had to be constructed in relation to something else, and that something else was Sweden. It was implied and assumed that legislation was stronger and better, while voluntary measures were weaker, with the latter referred to as "lax" and "laissez-faire." Some articles also assumed that voluntary measures meant Swedes were not social distancing, most accompanied with photos of Swedes "acting as normal." The coverage implied that the Swedish authorities were not taking it seriously. The headlines posed and answered rhetorical questions, passing judgment on the response: Not locking down became "risky" and a "gamble" and was a "challenge to the scientific mainstream." However, "lockdowns" were neither established best practice nor was there scientific consensus on their effectiveness. In press briefings during the first month, WHO representatives were clear that lockdown measures would not end the pandemic but would buy countries time to prepare the health system. Representatives also noted that "lockdowns" could disproportionally affect the poor and that these types of restrictions may simply not be "practical" in all settings. Moreover, it is difficult to come out of lockdown and there is the risk of lifting lockdown too soon, experiencing a resurgence and getting into an endless cycle of lockdowns (WHO 2020a(WHO , 2020b. Certainly, there are aspects of "lockdown" measures that work in reducing the spread of disease, but there are also negative consequences, and it can be difficult to safely transition to an open society again. Because many types of measures were applied at the same time, it is also difficult to assess which specific measures work and under what circumstances (ECDC 2020). In some cases, certain aspects of "lockdown" were performances -a form of "hygiene theatre" -like spraying chlorine on the street (Ohlsson 2020; Thompson 2020), but it was effective performance through which best practice was constructed and interpreted as authoritative knowledge. Moreover, as Carlo Caduff has noted, the same media outlets which condemned China's handling of the early pandemic as a "controversial" threat to civil rights and liberties, later turned to promoting "lockdowns" as the moral and responsible way to prevent deaths from Covid-19 (Caduff 2020).
Having established "lockdowns" as normal, two themes emerged in the reporting: Good Swedes who follow recommendations and Deviant Swedes who are arrogant and irresponsible for not taking the pandemic seriously.

Good Swedes
Swedes have above-average confidence in their politicians, who in turn, trust citizens to follow their advice (Time, April 9 th ) Swedish authorities have allowed a large amount of personal freedom unlike other European countries. Sweden's actions are about encouraging and recommending, not compulsion. (CNN, April 10 th ) In the beginning of the pandemic, Swedish Prime Minister Stefan Löfven used speeches and press conferences to appeal to solidarity and a sense of responsibility, duty, and sacrifice. The message was clear: by protecting yourself, you are protecting others and society (SVT 2020c, 2020d) There was also a focus on voluntary measures and following recommendations. Speaking at a daily press briefing State Epidemiologist Anders Tegnell noted that "people are very responsible; people are taking correct decisions most of the time (SVT 2020e)." He also discussed the "long tradition in Sweden of [people following] the advice of authorities," using the example of vaccination: "We don't need legal advice to get people to come and vaccinate their children for example. We get 98% of parents to come anyway (SVT 2020a)." In the international media, historian Lars Trädgårdh was quoted regularly on trust, describing three aspects to the Nordic high-trust society: First, citizens tend to place a lot of faith in public agencies and trust that they act in the public interest. Second, the authorities on their part trust citizens to heed their advice. Third, there is a high level of interpersonal trust where Swedes trust one other to act responsibly . . . Therefore, you don't need to micromanage or control behavior at a detailed level through prohibitions or threat of sanctions or fines or imprisonment (Trädgårdh in Anderson and Libell 2020; Rothschild 2020). Some national and international media focused on this Good Swede narrative which included various descriptions of national character: Swedes are calm, level-headed, and emotionally prepared for a national crisis (Trägårdh in Anderson 2020; Braw 2020). They are responsible and trust each other and the government. It would be "impossible to legislate and ban everything" so people needed to take responsibility for themselves (SVT 2020f). In the Swedish context recommendations were not actually voluntary, and, in any case, Swedes were used to following the recommendations of authorities (Lindberg 2020;SVT 2020g).

Deviant -or stupid, naïve, evil and arrogant -Swedes
This initial attention to the narratives of "Swedes trust authorities" and "the authorities trust Swedes to be responsible" was neutral, even curious. But curiosity turned to exotification, and deviance: . . . the government is relying upon Swedish character and traditions to see it through the pandemic. But behind this exceptionalism lies a worrying social compact between state and citizen . . . For trust in government to be a good thing, citizens need to keep doubting, examining, questioning, judging governments and their choices. In an organicist society however, tests on trust cease. And, with blind trust in the government, people are relieved from the responsibility for what happens around them -even to them-to their fellow citizens and to their family. (Boston Review, May 8 th ) Rather than telling people what to do -or, worse, telling them off -the Government is asking Swedes to do the right thing, and giving them the liberty to prove they are responsible citizens. . . . But that's a lot of trust to put in your people at a time of global pandemic, and a gamble that could have catastrophic consequences. (ABC News (Australia), April 6 th ) The questions posed by these writers were: can you really trust the government, and can you really trust your fellow Swedes? For example, on March 27 th Bloomberg ran the sub-headline: "Do you trust millions of your fellow citizens to do the right thing in a pandemic?" In an invective, the Observer (UK) columnist Nick Cohen wrote that "The tragedy of the Swedish outbreak is that it is a warning of what happens to countries that trust too much (Cohen2020)." There was also the idea that Swedes do not express dissenting views or complain about the government's handling, and that there was no critical coverage in national media or that criticism was silenced (Palm 2020a(Palm , 2020bCohen 2020). Cohen wrote that "Sweden sounds like an authoritarian technocracy intolerant of opinions and individuals that don't fit in (Cohen 2020)." In this way the media promoted the idea that Swedes blindly trust the government, are captors in a country-wide example of Stockholm syndrome, and need to be saved by international reporters who will tell the Swedish people "the truth" about COVID-19.
However, criticism of the government's whole strategy or parts of it featured every day in the mainstream Swedish media, not least in Dagens Nyheter which regularly ran opinion pieces by government critics. Especially in the early stages, as other countries shut down, there were constant questions of why Sweden was not doing likewise (SVT 2020h). Polls were also largely positive; even if four months into the pandemic confidence had declined from its peak: in late June, Kantar Sifo, a polling company, reported that 68% of survey respondents had a fairly or very large confidence in the Public Health (Kantar Sifo 2020).
Swedes were also framed as arrogant. Former Prime Minister, Carl Bildt, noted in an interview that there was an international perception that by not "locking down" Swedes were "signaling that they were a little better and more mature than other countries" (Larsson Hultin 2020). Indeed, the narrative that "Swedes follow voluntary recommendations" was provocative, as it was interpreted as an implied indictment of other countries.
Sweden is a sort of regional hegemon, and, its critics say, given to a certain arrogance and exceptionalism that can be grating (New York Times, June 22 nd ) Sweden's corona approach was driven by exceptionalism, overconfidence and a belief that Sweden is different and better than other countries (Financial Times,June 16 th ) In the beginning of the pandemic Swedish representatives felt that the media were focused on the differences rather than the similarities of the strategy, the idea that all countries shared the same goal to "flatten the curve" but that Sweden was using voluntary measures. For instance, during the daily press briefings, Anders Tegnell stressed that other countries have different contexts and ways of handling things: . . . we have taken different measures for different reasons. That one has different legal backgrounds, traditions about what types of measures one takes. (March 23 rd ) While in other countries you need a legal obligation and people are still protesting and going out and doing things. All of these things you need to put into context. I think it's very important for each country to do what fits into their context. If you need legal obligations to do things that's fine, if you can do it a voluntary way that's fine (March 27 th ).
As the media criticism of Sweden grew, representatives of FHM gradually become more negative toward other countries' handling of COVID-19, noting that the experiment was not staying open but "locking in a whole population for 4 or 5 months" or that in other countries one does not have a "trusting and understanding population." (Carlson 2020). In July, Anders Tegnell spoke of "lockdowns" in a popular radio program: "it was like the world had gone crazy . . . Country, after country closed down . . . in part because of political pressure" (Tegnell 2020).
However, the media also amplified the arrogant Sweden narrative which further solidified the deviant. That is, Swedes were forced into justifying or defending their strategy and then mocked for thinking that they were "morally superior" (Palm 2020b). For example, Sweden saw an increase of cases in June, which led the WHO's Office for Europe to put out a warning over the country (BBC 2020). FHM representatives clarified that this was due almost entirely to increased testing and that most of the cases were mild. Helen Clark, the former Prime Minister of New Zealand, compared Anders Tegnell to Donald Trump, tweeting: Like USA, #Sweden reported its highest daily number of #COVID19 cases this week. Its chief epidemiologist now sounds like POTUS as he complains about @WHO reporting of the situation saying their numbers are high because they are testing more. Hello! However, deviants are not deviant to everyone, and ambiguity is a characteristic of deviance (Downes and Rock 2003;Goode and Ben-Yehuda 2009). Sweden, traditionally praised by the left, became used as an example of the right and libertarian groups, particularly in the US and UK. Protestors in Michigan shouted "be like Sweden." The libertarian think tank, The Ron Paul Institute reprinted an article on "Why Sweden has Already Won the debate on COVID 'Lockdown' Policy" and Libertarian Senator Rand Paul touted the Swedish model to congress (Tobin and McLaren 2020). While often based on a misunderstanding or simplification of the Swedish approach, the basic arguments from these groups were that Sweden represented common sense, it had not panicked, the evidence for locking down was not that strong, social isolation has downsides, and it preserved civil liberties and did not destroy the economy (Fund and Hay 2020;Carlson 2020;Navarro-Genie 2020;Erixon 2020).
The content of the right wing and libertarian reporting was less important than its existence: by being associated (positively) with the right, some in the left found further justification for Sweden's deviant status. That is, for some left-leaning columnists, that Sweden's "policy was a model for the right" proved that it was wrong (Palm 2020a;Cohen 2020).
The Swedish sickness is a political as well as a medical disaster. Professor Johan Giesecke, an adviser to the Swedish government alongside Tegnell, became a star of the rightwing web as he lectured other governments on the futility of their tough measures. British Conservative commentators have boomed out claims that Sweden showed there was no need to close the UK economy. Sweden had "held its nerve," they gushed, in much the same way communists once gushed about the Soviet Union (Nick Cohen in The Observer, May 23 rd )

Deviants must be punished
Swedes tend to consider themselves impeccable in matters moral and sanitary. Being treated as irresponsible carriers of disease has been a bewildering blow (The Economist, July 3 rd ) Sweden Tries Out a New Status: Pariah State (New York Times, June 22 nd ) Punishment reinforces and amplifies a groups' deviant status, and there were several examples of this in which the media drew attention to perceived failures of the country's pandemic response or gave disproportionate attention to new, but inconclusive, polling or immunity data. For instance, Sweden was castigated for not achieving herd immunity, even though it had never been the strategy. Some commentators linked the alleged "herd immunity strategy" to the economy, expressing the inaccurate belief that Sweden put the economy above public health concerns (Goodman 2020;Laterza and Romer 2020). This media preoccupation with "Sweden's herd immunity strategy" meant that every emerging study suggesting immunity was fleeting offered a new opportunity to punish Sweden, as did every negative report about its economy. Moreover, one of the reasons cited in the press for why Sweden had neither achieved herd immunity nor protected its economy was because people physically distanced too much and had not frequented shops or restaurants as normal. That is, the same media outlets that questioned the validity of voluntary physical distancing measures -accompanied by photos of Swedes drinking in crowded bars -were now suggesting that voluntary measures worked too well: Sweden is still nowhere near "herd immunity," even though it didn't go into lockdown (CNN, May 21 st ) Sweden's hope has been to achieve herd immunity, whereby enough of the population has been infected that coronavirus can no longer spread widely. Yet studies in May suggest that Sweden is nowhere near the threshold needed to realise this (The Independent, June 1 st ) Coronavirus: Sweden's economy plunging despite refusal to lockdown (AFP, June 7 th ) The media were also quick to pick up on mistakes and admissions that there had been problems with the strategy. Cohen referred to this as "community sensitization." Once a group has been made deviant, "extremely small deviations from the norm became noticed, commented on, judged, and reacted to ((Stanley) (Cohen 1967;Goode and Ben-Yehuda 2009)." For instance, as Sweden diverged with its neighbors in terms of deaths per million, headlines proclaimed: In an interview with Sveriges Radio, Anders Tegnell stated that "there was room for improvement in what we had done in Sweden" (Öhman and Rosén 2020), which was interpreted as: CNN: Sweden's top scientist admits response to virus was too lax (CNN, June 8 th ). On the main page that day, there was a link to a video with the text "Sweden makes admission about 'no lockdown' Covid-19 policy." Architect of Sweden's coronavirus strategy regrets not imposing tougher lockdown (Telegraph, June 6 th ) Sweden admits light lockdown was wrong approach to Fighting Coronavirus (Huffington Post, June 4 th ) When polls suggested that confidence in authorities were declining, headlines stated that: Swedes rapidly losing trust in Covid-19 strategy, poll finds (The Guardian, June 26 th ) Analysis: Is fragile support for Sweden's COVID-19 strategy fracturing? (Euronews, June 11 th ) Swedish Faith in Covid Strategy Plunges After Errors Revealed (Bloomberg, June 3 rd ), The URL text to the Bloomberg article was: https://www.bloomberg.com/news/articles/2020-06-03/sweden-won -t-abandon-covid-strategy-despite-admitting-to-errors Most of the reporting referred to a poll by the firm Ipsos and the newspaper Dagens Nyheter. The percentage of people who had a fairly or very large confidence in the authorities' ability to handle an increased or larger spread of COVID-19 went from 43% in March, 56% in April, 54% in May to 45% in June. The percentage with a fair or great confidence in the Public Health Agency was 69% in April, 66% in May and 67% in June; for state epidemiologist Anders Tegnell, the same numbers were 69%, 67% and 60% (Rosén 2020). The headlines suggested an extreme dissatisfaction, which was not fully supported by the data as a whole.
Part of the decline was probably due to the higher number of deaths in Sweden compared to neighboring countries -particularly in care homes for the elderly, and to the fact that, going into the summer holidays, many countries had closed their borders to Swedes. Also, the poll data cannot be read as an indictment of not "locking down:" while the international media focused on Sweden as a relatively open society, the response and handling was more complicated, as already noted. In fact, the poll example highlights the final way in which deviance was reinforced. By simplifying the message and consistently returning to "not locking down" as a deviant behavior, rather than looking at nuances, the deviant label could be justified: The reason for Sweden's high death rate has to do with the government's policies.
Following the advice of the country's chief epidemiologist, Anders Tegnell, the Swedish government chose not to impose strict lockdowns, curfews, or major border closings because the government felt it would hurt the economy and would only push the crisis further down the road (Vox, 28 April 28 th ) Sweden now has among the highest per capita death rates from Covid-19 in the world. Why?
The answer is simple. Sweden was lax in its implementation of protective measures in the face of the outbreak, refusing to implement broad stay at home orders for residents, or to enforce recommendations to wear masks or social distancing measures. Other than the government decision to shut down universities and high schools, compliance to public health recommendations was entirely voluntary (CNN, June 4 th ) That is, with few expectations, the global media environment attributed the "failure" of the Swedish model to the simple fact that the country did not "lockdown." This simple message ignored the actual measures taken and the extent to which Swedes had reorganized their daily lives to prevent the spread of Covid-19. It also ignored the legal frameworks and political structures which guided the response.

Discussion and conclusions
Writing during the 2009 H1N1 pandemic, Brigg and Nichter called for anthropologists "to consider how countries are rewarded in the press for being 'good global health citizens' versus being blamed as bastions of unhygienic subjects" (Briggs and Nichter 2009: 197). In this article I have looked at how, using exaggeration, sensationalist language, and selective reporting, the global media environment framed the Swedes as a deviant population.
Moral panics often serve to clarify "normative contours" and "moral boundaries" (Goode and Ben-Yehuda 2009: 29). The idea of "lockdown" quickly became legitimized and established as scientific best practice, even though there was mixed evidence for it. By not shutting down society, the Swedish approach "could indirectly cast doubt on some of the fairly drastic measures undertaken" (TT, AFP and The Local 2020). As a country it threatened the social order and needed to be punished. As a headline in the Telegraph noted: If Sweden succeeds, lockdowns will all have been for nothing" (April 25 th ).
Moreover, Swedes could become deviant because they were an unknown-familiar group. International commentators felt they knew enough about the country to pass judgment, but at the same time did not know enough to contextualize its policies, thus enabling the spread of misinformation, and disinformation to spread. In the same way, Swedish culture -or rather a caricature of it -was often invoked as an explanation for deviance. Beyond this, the stories about "perfect" Sweden's fall from grace sell.
The news frames and defines reality (Bird 2010). For instance, not even four months into the pandemic, the myth of Sweden's herd immunity strategy had been discussed as truth in peerreviewed articles, using misleading news reports as sources (Sheikh et al. 2020;Mayer and Lewis 2020). That something is widely reported does not make it correct, but through repetition, framing, rhetoric and the sheer volume of stories, misinformation, disinformation and stereotypes become truth. This "truth" has consequences. It can be used to legitimize and de-legitimize different policy choices, potentially leading to material consequences and public anxieties (Bal et al. 2020;Caduff 2020). As anthropologists we also have a responsibility to resist moral panicsin Cohen's sense of the term. Instead, we should work to nuance the media's reporting by drawing attention to the wider legal, historical, economic, and social webs in which pandemic responses are situated.

Disclosure statement
No potential conflict of interest was reported by the author.

Funding
This work was supported by the Swedish Research Council, grant number 2018-05266Vetenskapsrådet [2018-05266];

Notes on contributor
Rachel E. Irwin is a researcher in ethnology at Lund University. Her research is at the intersection of health policy and anthropology. ORCID number 0000-0002-8350-441X