Dealing With Covid‐19 in Casual Democracies

This article reports findings from an in‐depth, autoethnographic study of local communities in Denmark and England left behind by local journalism. The study was conducted during—and is thematically framed by—the Covid‐19 pandemic


Introduction
Recent decades have seen a steep decline in local journalism across Europe, North America, and other parts of the world (e.g., Hayes, 2021;Nielsen, 2015). Local news media have been forced to shut down due to loss of readership and lack of sustainable revenue models, creating so-called "news deserts," communities with no journalistic outlet reporting on local affairs, and "ghost newspapers," outlets so deprived of resources that they no longer can engage in proper reporting (Abernathy, 2020). Scholars argue that the collapse in local news outlets represents "the greatest challenge facing journalism today" (Wahl-Jorgensen, 2019, p. 163) and a major threat to the well-being of local democracies (Miller, 2018). Studies indicate that citizen engagement drops significantly when local journalism vanishes (Hayes & Lawless, 2015), voting polarisation increases with newspaper closures (Darr et al., 2018), and communities without journalistic outlets have higher degrees of corruption and other forms of misgoverning (Adserà et al., 2003).
Yet, in-depth research on how local democracy develops in places left behind by local journalism is scarce and often based on the assumption that local journalism is a prerequisite for civic engagement and informed citizenship (Usher, 2023). This article presents a study where the link between (a lack of) local journalism and civic engagement, and informed citizenship is not assumed but empirically investigated. The article builds on immersive, autoethnographic data from two communities left behind by local journalism. The study aimed to empirically investigate the opportunities and challenges for informed citizenship in these communities where local journalism plays an insignificant role and, consequently, what characterises local democracies in such communities. The study was conducted during the peak of the Covid-19 pandemic in 2020 and 2021, providing knowledge on how local democracies without local journalism developed in a time of crisis. This is important because people's need for accurate information increases during a crisis (Westlund & Ghersetti, 2015). This was also the case during the Covid-19 pandemic (Bento et al., 2020) when several journalistic media institutions across Europe experienced a massive increase in audience reach (Van Aelst et al., 2021). At the same time, the crisis hit the news media hard financially, with revenues from advertisements dropping dramatically, especially for local news media (Olsen et al., 2020). The important question is, therefore, how people in local communities managed to satisfy their need for accurate information about the local handling of the pandemic when local, legacy news media could no longer provide it.
For obvious reasons, the pandemic made ethnographic research difficult, especially comparative ethnographies across countries. This study is, therefore, unique as it presents first-hand autoethnographic data from two different countries, England and Denmark, during a period when multiple restrictions were in place, especially related to travel and social mobility. The two cases are similar in representing places left behind by local journalism. They are different in that they represent social, political, and cultural differences between the UK and Denmark, and in the sense that one is urban (Frederiksberg in Copenhagen, Denmark) and the other rural (Ringwood in Hampshire, UK). After reviewing relevant research and literature, the two cases, the methodological procedures, and the findings will be presented in detail.
Ultimately, the article argues that places left behind by local journalism are not necessarily news deserts but instead resemble what I label "casual democracies," in which the constitution of "news" is fluid and partly randomised, conditions for freedom of speech and whom and what to trust are disguised by platform infrastructure and undisclosed interests, and where the local public sphere is operated in part by random individuals. There is a potential risk that the combination of platform infrastructure and individual agency can create new, unstable hierarchies of power with unclear democratic anchoring.

News Deserts, Platformisation, and Informed Citizenship
Journalism and democracy have for centuries been viewed as inseparable concepts by scholars and professionals alike (Ryfe, 2019). One central idea behind the journalism/democracy nexus is that journalism's core mission is to enable informed citizenship. The premise behind what Schudson (1999) labelled the "informed citizen model of democracy" is that in a well-functioning democracy, people have a need to be informed about politics and the deliberations and decisions of those who are elected to govern. Ideally, journalism fills such a role, both in local and national contexts. In practice, however, local journalism has not necessarily met this expectation (Nielsen, 2015), partly because it often lacks the resources necessary to scrutinise local authorities properly. As revenue streams have decreased, resources for many local newspapers have declined to such an extent that they have become "ghosts" of their former selves, implying that the "quality, quantity and scope of their editorial content are significantly diminished" (Abernathy, 2020, p. 24). There are two paths to becoming ghost newspapers, according to Abernathy. The first is that a local newspaper is acquired by a larger daily and slowly fades away as its news-gathering operations are merged with the larger paper's. The second path is when a local paper's newsroom staffing is cut so dramatically that the remaining reporters cannot adequately cover their communities. It is well documented that many local newspapers have followed one of these paths in recent decades, not only in the US, both also in many European countries (Hayes, 2021;Nielsen, 2015). None of these paths leads to a good place-they tend to end in "news deserts," places where no local news is reported by journalists at all.
However, communities left behind by local journalism are not necessarily deprived of news or mediated public spaces facilitating informed citizenship (Collier & Graham, 2022). Local public spheres, understood in a broad sense as the metaphorical space where people find out what's happening in their community, the social, cultural, and political issues that face them, and where they also can engage with these issues and have their voices heard (McKee, 2005, p. 5), have in large parts been re-configured around digital platforms. A far-reaching process of platformisation has transformed not only how people stay informed about and connected to their community, but has also re-shaped almost all aspects of what constitutes society (Poell et al., 2019). A premise for the study to be presented in this article is, therefore, that communities left behind by local journalism are not necessarily news deserts, understood as places devoid of news-they might be places where local news finds alternative forms and alternative means of production, distribution, and consumption.
Local social media groups facilitate the distribution and discussion of local news (Swart et al., 2019), even in places where no traditional local news outlets exist. However, social media news consumption is quite different from the consumption of news produced and distributed by traditional local news outlets operated by journalists and editors. First, it is facilitated and restricted by technological affordances that regulate what can be communicated on a given social media platform, how such communication is facilitated, and the algorithms that control its diffusion patterns-all of which are controlled remotely and not, for instance, by a local editor; instead, they involve black-box algorithmic decisions that are inaccessible to users (Cotter, 2021). Second, social media news consumption is incidental and heavily dependent on opinion leaders (Bergstrom & Belfrage, 2018). It does, therefore, not adhere to the routines associated with journalistic news production and consumption, nor to the judgments of locally based professional journalists and editors. Third, social media news consumption is enmeshed with other kinds of information sharing and consumption from both private and public spheres, thereby blurring the boundaries between various producers and types of news and information. The consequence is that in social media news consumption, news has become separated from journalism (Steensen & Westlund, 2021).
I, therefore, argue that the idea of "news deserts" confuses "news" with "journalism" as part of the "descriptive fuzziness" (Usher, 2023, p. 239) that makes the concept problematic. Places without traditional local newspapers might lack journalism, but they don't lack news. However, what kinds of local news exist, who distributes, consumes, and discusses it, and how trustworthy such news is depends on the technological affordances of social media platforms and other factors related to local cultural norms and values within the social fabric of the community in question. Such factors include what is considered news-and trustworthy by whom in a given community, how boundaries between the private and the public are drawn, what is considered proper means of public communication locally, local expectations towards volunteerism, degrees of social control, and opportunities and traditions for individual agency in the community. Many of these factors are related to local news production being de-professionalised and instead left to any community member with the time and capacity to produce and share information that might be perceived as news.
The disconnection of news from journalism also means that the idea of informed citizenship needs to be de-connected from journalism. As argued by Deuze (2008, p. 848), the old notion of informed citizenship is "a thing of the past-a prescriptive and rather elitist notion of both how people should make up their minds and what (political) representation means to them." The many technological, cultural, and social factors that shape platformised news production, distribution, and consumption also shape what informed citizenship means in a community without local journalism. One cannot assume that because there is no local journalism, possibilities for informed citizenship do not exist. Rather, how and to what degree people living in places left behind by local journalism are informed about and connected to their community is an empirical question. Finding answers to that question requires looking beyond traditional newsrooms and instead focusing on whatever information networks constitute the local public spheres-who, how, and what shape them. This will be investigated in this article concerning a crisis, the Covid-19 pandemic. The aim is not only to find out what kinds of news circulate in communities left behind by local journalism but also to discover how local democracy develops when local journalism is no longer a significant part of it. Specifically, the article addresses three empirical research questions, the first being: • RQ1: What possibilities did the people of Ringwood and Frederiksberg have to stay informed about the local handling of the Covid-19 pandemic?

Covid-19 and the Infodemic
The Covid-19 pandemic has been portrayed as a crisis with two dimensions: a public health dimension and a public information dimension. Concerning the latter, the World Health Organization (WHO) labelled the crisis an "infodemic," meaning "an overabundance of information-some accurate and some not-that makes it hard for people to find trustworthy sources and reliable guidance when they need it" (WHO, 2020, p. 2).
Concerns about false and misleading information related to the pandemic are largely connected with social media platforms. Several studies have documented the spread of Covid-19 related conspiracy theories and other forms of misinformation on social media platforms (e.g., Islam et al., 2020), while others have documented that such misinformation could cause psychological disorders and panic, fear, depression, and fatigue (Rocha et al., 2021). Such concerns led authorities, social media platforms, and news and fact-checking organisations to fight against Covid-19 misinformation actively, for instance, by filtering out or flagging false content (Niemiec, 2020). This, in turn, created concerns over what the infodemic did to the conditions for free expression. Reflecting such concerns, a Council of Europe report stated that "free and pluralistic public debate is crucial for the public's understanding of the situation and for their ability to make informed decisions, limit rumours, recognise disinformation and foster solidarity and trust in measures taken to address the crisis" (Noorlander, 2020). Studying pandemic-related news and information in communities that are left behind by local journalism and dominated by social media platforms is therefore important since such communities might have been more vulnerable to misinformation while relying on platform algorithms and administrators of local social media groups to moderate free speech. It is well documented that the algorithmic moderation of social media platforms represents a limitation on any potential threat to users' right to free speech (Dias Oliva, 2020). This potential tension between the right to free speech and a need to fight misinformation related to the pandemic will be investigated in this article as a critical incident for the state of local democracy in places left behind by local journalism. Specifically, the following two empirical research questions will be addressed: • RQ2: To what degree was Covid-19 related misinformation a concern in Ringwood and Frederiksberg? • RQ3: How was the tension between censoring misinformation and the right to free expression negotiated in the local public spheres of Ringwood and Frederiksberg?

The Cases
Most studies of places left behind by local journalism are conducted in rural areas marked by socio-demographic problems related to unemployment and poverty. It is, therefore, difficult to assess how the decline of local journalism has affected these communities and their local democracies since there are so many other issues that are causing problems simultaneously. This is one of the reasons for choosing Frederiksberg and Ringwood as cases for this study, as both these places are quite prosperous and marked by relatively few sociodemographical problems compared to the average communities in Denmark and the UK. Moreover, as countries, Denmark and the UK are highly developed democracies with relatively few socio-demographical problems compared to many other democracies (United Nations Development Programme, 2020, p. 343).
Frederiksberg is a densely populated municipality encapsulated within the city of Copenhagen. A population of 105,000 live there, squeezed into nine square kilometres. There used to be two local newspapers in Frederiksberg. One closed in 2018, while the other, Frederiksberg Bladet (a weekly, free, ad-financed newspaper), had to lay off all but one journalist during the spring of 2020 due to a massive loss of advertising revenue. In late August 2020, the Danish trade press publication Journalisten.dk ran a story titled, "The Biggest News Desert in Denmark? You'll Never Guess Where it Is" (Albrecth, 2020). The answer was Frederiksberg.
In contrast to Frederiksberg, Ringwood is a small, rural market town in southern England, just north of the city of Bournemouth. The town is situated within Ringwood town parish (a parish being the lowest tier of governance in England), in which around 15,000 people live, mostly in villages outside the town centre. There is no traditional local newspaper in Ringwood, but the town occasionally receives some coverage from the local papers in the nearby, bigger cities of Bournemouth and Salisbury. These newspapers used to have local offices in Ringwood, but those were closed several decades ago. There is, however, an untraditional, non-journalistic newspaper in town, the Ringwood and Fordingbridge News. This newspaper is published bi-weekly and distributed to almost all households for free. The publisher is a local politician, and the content is a mixture of local ads and articles submitted by mostly local charities and other NGOs. No one gets paid to write for the paper, according to the publisher (Interview 17), who established it when the regional papers closed their satellite offices in Ringwood in the 1990s.
Given this lack of coverage from local journalism, various social media platforms became the main channels of news, information, and public deliberation on matters of local interest in both Frederiksberg and Ringwood. Facebook was the most significant in both communities, with several local groups and pages playing important roles in the local public spheres. In addition, the location-based social media platform Nextdoor played a role in Ringwood.
During the study period, August 2020 to July 2021, many important events concerning the pandemic happened, which had important ramifications for people in the two local communities. Figure 1 displays a timeline of infection rates and the main events of the pandemic in Denmark and the UK during this period.

Materials and Methods
The study is based on an analysis of three datasets collected from the two case studies, Frederiksberg and Ringwood, as part of a larger research project on local public spheres. Dataset 1 contains field notes from 10 months of autoethnographic observations and reflections (five months in each community) and transcribed interviews with "key information brokers." Dataset 2 contains data from local Facebook groups and pages, while Dataset 3 contains a content analysis of the two local newspapers, Frederiksberg Bladet and Ringwood and Fordingbridge News. Since the data material contains personal information, approval to collect, store, and analyse such data was acquired by the Norwegian Agency for Shared Services in Education and Research.

Dataset 1: Field Notes and Interviews
Autoethnography implies that the researcher writes about their own experiences as part of a culture (Pitard, 2017). It differs from other ethnographic approaches as the researcher's own experiences are part of the analysis (hence the prefix "auto"). I lived in Frederiksberg from August to December 2020 and in Ringwood from February to July 2021, trying to become as integrated into these communities as possible. The field notes contained impressions and reflections on experiences with living in the communities, notes from encounters with people and meetings I attended, and reflections and descriptions of information and news I came across, both mediated and unmediated, offline and online.
The field notes thereby also included elements of netnography, an adaptation of ethnography for virtual spaces (Kozinets, 2020). Netnography involves studying online cultures and communities by being part of them over time. I became a member of several public and private social media groups, predominantly on Facebook but also on Nextdoor (in Ringwood). Notes from the netnographies were included in the same notebook as other field notes.
Interview subjects were identified through the autoethnographic experience. I identified "key information brokers" in each community, i.e., people who played important roles in the local public spheres. These included local politicians, representatives of local authorities and NGOs, administrators of local social media groups and pages, publishers of important news outlets, and some journalists and editors. These interviews were semi-structured and included some generic questions on the interview subjects' reflections on local democracy, conditions for free speech, the local media scene, the availability of trustworthy news and information,  and specific questions related to their roles in the community. Interviews lasted between 45 minutes and 2 hours. Most were conducted face-to-face, but some were conducted over Zoom due to Covid-19 restrictions. Thirty-two people were interviewed, 16 in each community, all giving informed consent to participate as informants. All interviews were recorded and transcribed by a research assistant. The field notes and interview transcripts were imported to Nvivo and coded with three broad categories of codes: A topics category, out of which the covid topic with several sub-topics will be analysed in this article; a themes category, with codes such as "democracy" (with sub-codes), "disinformation," "trust," "moderation," "free speech," etc; and a media category, with codes reflecting the names of mentioned media and platforms. The codes were mostly inductively developed, as topics, themes, and media were mentioned in the data. This coding structure allowed me to utilise the analytical power of Nvivo to find the themes that corresponded with various aspects of the Covid-19 topic.

Datasets 2 and 3: Media Content
Part of the autoethnographic experience was identifying media and platforms relevant to the local public spheres. Facebook groups and pages played a major role in both places, and I set up lists with the biggest local Facebook groups and pages in CrowdTangle, a public insights tool owned and operated by Facebook. CrowdTangle only provides access to public groups and pages, meaning that some significant private groups had to be analysed manually through netnography (see Section 5.1). All posts from all the local Facebook groups and pages published between August and December 2020 (Frederiksberg) and January and July 2021 (Ringwood) were downloaded as CSV files, which were imported to Excel for analysis.
Dataset 3 is the content of the print editions of the two local newspapers, Frederiksberg Bladet and Ringwood and Fordingbridge News. I was able to access and download all issues of Frederiksberg Bladet from August to December 2020 as PDF files. These issues contained 625 stories. As for the Ringwood and Fordingbridge News, I collected all but one issue published between February and July 2021. These issues contained 210 stories. Two research assistants assisted with the content analysis of these stories. The content analysis was designed based on variables used by Williams et al.'s (2015) study of hyperlocal news sites in the UK, meaning that the main topic of each story was identified mostly through a predefined list of topics (with the addition of Covid-19 as a new, separate topic). Other variables of relevance for this study were author, role of author, and genre. I set up a detailed coding manual in cooperation with the research assistants, and intercoder reliability tests (Krippendorff's alpha) showed satisfactory results (ranging from = 0.80 for topic to = 0.97 for author role).

Findings
The findings will be presented case by case, starting with Frederiksberg. First comes findings from the content analysis, which mostly addresses RQ1. Then follows findings from the autoethnographic field notes and interviews, which are relevant mostly concerning RQs 2 and 3. However, as the discussion in the next sections will reveal, all three datasets have relevance for all three RQs.

Diffusion of Covid-19 Information in Frederiksberg Media
The content analysis of Frederiksberg Bladet (the print edition) during the fall of 2020 shows that the Covid-19 pandemic was not a topic that dominated this newspaper. The share of the content published from August to December that dealt with the pandemic was 4.5% (28 stories), making it the 10th most popular topic. The share is even lower if we look at news stories only (3.7%), meaning that most of the Covid-19-related content was letters to the editors. The biggest group of Covid-19 letter writers was local politicians, who were mostly calling for support of local businesses and cultural institutions during the lockdown. The second biggest group of letter writers was citizens who discussed various types of restrictions (like if masks are useful or not, and if they should be allowed to run/exercise in parks).
The content of the Frederiksberg Facebook groups and pages shows a different picture. Ten percent of all posts from the local groups and pages analysed were related to the pandemic. Figure 2 shows the number of posts and the share of Covid-19-related posts per group and various page categories. Posts in the groups make up more than half of all the posts in the dataset. But Covid-19 related posts were more common in the pages than in the groups. The official page of the municipality had the highest share of Covid-19-related posts (17%), implying that the municipality found it important to provide information about the pandemic. Local politicians and community organisations were also quite active in their Covid-19-related Facebook communication, as was Frederiksberg Bladet, which had a significantly higher share of Covid-19-related posts on its Facebook page than in the newspaper.
Covid-19-related posts published by the municipality on its official Facebook page and by local politicians on their Facebook pages were shared more than twice as often as other posts from the same pages. In contrast, Covid-19-related posts published on Frederiksberg Bladet's Facebook page were shared much more rarely than other posts. In other words, people found Covid-19related information published by the municipality and local politicians far more shareable, and thereby perhaps more trustworthy, than Covid-19-related information published by Frederiksberg Bladet.
This picture becomes even clearer when we look at the posts with the most interactions, meaning posts  None of the most interacted with Covid-19 posts was published by Frederiksberg Bladet or any other news outlet for that matter.

Misinformation and Free Speech in Frederiksberg
The Nvivo coding matrix reveals that thematic codes related to moderation, free speech, and misinformation often co-occurred with field notes and snippets of interview transcripts coded with various Covid-19 topical codes. This was particularly related to restrictions such as the mandatory use of face masks. Discussions on face masks, other restrictions, and the pandemic, in general, could become quite heated and confusing, often with an unclear factual basis. One example was when someone shared a graph in one of the Facebook groups showing the number of deaths due to Covid-19 compared to other deceases, the point being that Covid-19 was much less deadly. This created a heated debate and lots of confusion, but it was impossible to determine the graph's source and if the numbers were correct. Nevertheless, many people took the graph at face value and used it to argue against the restrictions (field notes, week 18, 2020).
The two biggest Facebook groups dealt with discussions on restrictions in very different ways. The biggest group, which was set as private and which therefore is not part of the CrowdTangle data presented above, banned discussions on face masks because such discussions tended to violate this group's policy, which stated that political content and hateful rhetoric were not allowed (field notes, week 7, 2020). On the other hand, discussions on Covid-19 restrictions were common in the second biggest Facebook group, and the municipality would use this group to post updates on restrictions and inform people about how they were to behave locally. However, the group administrator would often question such posts, as he, for example, did when the vice-mayor posted a video on the importance of wearing face masks (field notes, week 7, 2020). In fact, the group admin would often post things himself or comment on others' posts with arguments against face masks. In the interview, he said: "There are just many things about this pandemic where the information simply isn't on top. I am one of those who believe that masks do not have the desired effect" (Interview 2). This admin's views became quite widespread in the Facebook group, and debates could be intense. Quite often, people would be accused of being "fact resistant" or similar things when they expressed views opposing restrictions or other official positions regarding the pandemic. In other words, there was tension between the right to free speech and the need to avoid misinformation and confusion in Frederiksberg. Some of the people I interviewed felt that free speech and, thereby, democracy was under threat: You get called tinfoil hats or conspiracy theorists. Just because you have a simple and critical question about face masks, or something else….We should accept that there is a minority that believes something other than the huge majority. They should be heard. They should always. But that has stopped with this coronavirus. I think it is a democratic problem. (Interview 3) Many people I interviewed and talked to expressed anger and disappointment with the recent developments of Frederiksberg Bladet and its lack of engagement with local politics and other important matters due to the newsroom cutbacks. On several occasions, I wrote reflections in my field notes on the local news I would have liked to have been informed about, but for which I could not find any information in the local public sphere. This is related, for instance, to the consequences of the pandemic for local businesses. Frederiksberg is famous for its many small shops, cafes, and restaurants. How did the lockdown affect them? Who survived, and who had been forced to shut down permanently?

Diffusion of Covid-19 Information in Ringwood Media
The content analysis of Ringwood and Fordingbridge News shows that 18 out of the 210 stories (8.5%) published during the spring of 2021 in the nine editions analysed were related to the Covid-19 pandemic. This made the pandemic the fourth most popular topic in the newspaper. However, most of these stories (14) were letters written by citizens. Only one Covid-19 story had been written by the publisher himself. Local politicians or public servants did not use the newspaper to convey any Covid-19-related information.
Regarding the local Facebook groups and pages, Covid-19-related posts comprised 3.2% of the content, far less than in Frederiksberg. Groups were much more dominant in the Ringwood Facebook domain than in Frederiksberg: Posts in groups accounted for 77% of all posts in Ringwood, compared to 53% in Frederiksberg. However, discussions in the groups revolved around the pandemic to a very limited extent (only 2.1% of all posts, see Figure 3). The town council's Facebook page was dominated by Covid-19-related posts, implying that this local authority found it important to inform about the pandemic. There were no significant differences between Covid-19-related posts and other posts regarding types of interactions, apart from one thing: When local businesses posted something Covid-19-related, these posts were more than three times as likely to be liked than other types of their content. None of the 12 most engaging Covid-19 posts (regarding likes, shares, and comments) were posted by local authorities or the Ringwood and Fordingbridge News. The most interacted with of all Covid-19-related posts was published by a local business, a wine bar, which on 28 July 2021 announced on its Facebook page that it had won a travellers choice award and thanked everyone for their support during the pandemic.

Misinformation and Free Speech in Ringwood
The comparatively low engagement with the pandemic in the mediated public sphere of Ringwood starkly contrasts how dominant it felt when living there. A national lockdown had just been imposed when I moved to England in January 2021. People in Ringwood moved in big circles around one another outside, and the feeling was that the pandemic posed a much bigger crisis than in Frederiksberg. It, therefore, felt strange that the mediated local public sphere did not reflect this.
In my field notes, I wrote down several reflections on the news I missed concerning the local effects of the pandemic. When restrictions were lifted in the UK, and non-essential shops, pubs, and restaurants were allowed to reopen, I was curious to know who had survived and what kinds of trouble they had experienced. And how were local schools preparing for the reopening, given the face mask policy and the testing regime they were supposed to enforce? Did parents, kids, and teachers look forward to the reopening? How had the crisis affected them all? (field notes, weeks 6 and 11, 2021). I couldn't find any such news stories in the Ringwood public sphere.
The Nvivo coding matrix of field notes and interview transcripts reveals that thematic connections were quite similar to those in Frederiksberg. Codes related to moderation, free speech, misinformation, and social control often co-occured with Covid-19 topical codes. But these themes played out differently in Ringwood, where critical views on, for instance, restrictions and vaccination were rarely seen. The most important public arena for local Covid-19 related information was a Covid-19 assistance Facebook group set up by a few local politicians at the start of the pandemic. This group became popular, with more than 3,000 members, and it played an important role in helping people in the community. The initiator and main administrator of the group explained the need to censor posts to minimise misinformation: Informant: We were very keen to focus on accurate information. And we did have misinformation. People, because of the way that Facebook group works, people joined and then suddenly started posting misinformation, and we had to remove it and kick them out. So basically, we were very actively moderating.
Interviewer: What would that type of misinformation typically be that resulted in you kicking people out?
Informant: It was everything from just speculations about things that really wasn't helpful, right up to some really quite hard anti-vax type, or some conspiracy type theories….So we had everything. And that was just dealt with and removed. Straight away. (Interview 26) Critical comments related to Covid-19 restrictions or vaccination were also removed on Nextdoor. I had set up my Nextdoor account to send me weekly updates with top new posts and discussions from Ringwood neighbourhoods, and on several occasions, I experienced that critical Covid-19 posts, which were in the email notification, had since been removed from the social media platform by the time I clicked on them.
Not everyone was happy with the strict moderation and lack of critical debate regarding the handling of Covid-19. The following interview quote, where the informant was asked about their thoughts on free speech, reflected the opinion of several people I interviewed and talked with: Before Covid, I thought it [conditions for free speech] was pretty good. Now, I think that if you exercise your free speech, or you go against the narrative, you´re labelled very quickly by a lot of different organisations, and people are being conditioned to ostracise these people now….So free speech is disappearing. (Interview 25) Several people I met felt there was no point in publicising their views, as they were certain they would be censored (field notes, week 19, 2021). One producer of a local non-journalistic media outlet answered the following when asked if he had considered giving voice to someone critical towards vaccination: "If I were to cover the Covid vaccination story…I'd love to do it, but I would lose all my friends" (Interview 25).

Discussion and Conclusion
RQ1 asked what possibilities the people of Ringwood and Frederiksberg had to stay informed about the local handling of the Covid-19 pandemic. The findings demonstrate that the two local newspapers, which both can be characterised as "ghost newspapers," were not concerned with bringing pandemic-related local news. Frederiksberg Bladet, which had followed what Abernathy (2020) identified as the second path towards becoming a ghost newspaper (implying a dramatic cut-back of editorial staff), had only one journalist left to cover the community. Coverage of the pandemic was, therefore, extremely scarce. Ringwood & Fordingbridge News, on the other hand, had followed what can be identified as a third path towards becoming a ghost newspaper, as this paper had been established as a paper without editorial staff from the very start, as a response to other, larger local papers pulling out. There was, therefore, no editorial coverage of the pandemic, or any other topic for that matter, in this local paper. Both places thereby resemble news deserts because they lacked coverage from professional journalism. However, this does not mean that Ringwood and Frederiksberg were without news, as people had other means to stay informed about the local handling of the pandemic (and other issues, for that matter), the most important being local Facebook groups and pages. In Frederiksberg, local politicians and authorities had to a large degree, professionalised their public communication, with all local councillors having Facebook pages, which they used extensively to communicate about the pandemic in addition to writing letters to the editor of Frederiksberg Bladet. This was not the case in Ringwood, where local politicians and authorities were less engaged in the mediated local public sphere.
One of the concerns with the rise of news deserts is that when communities are not covered by journalists, misinformation will thrive (Pickard, 2019). Places without local journalism were exposed to a double misinformation threat during the pandemic, as the "infodemic" also constituted such a threat. RQs 2 and 3 asked whether misinformation was a problem in Ringwood and Frederiksberg and if free speech had been compromised. The findings show that misinformation regarding Covid-19 was a problem, or at least was perceived as a problem, in both communities. However, the borders between misinformation, critical opinions, and adequate alternative perspectives on issues related to prevention and vaccination were blurry. Moderators of local Facebook groups were powerful actors in defining such borders. Informants in both places expressed concerns over the consequences of the way the pandemic had been handled locally for freedom of speech and, consequently, for local democracy.
This article aimed to investigate the characteristics of local democracy in places left behind by local journalism. I have argued against labelling such communities as "news deserts" because this concept confuses "news" with "journalism." Usher (2023, p. 246) argues that researchers "have a built-in preference for professional, legacy media as the providers of civic information" and that the concept of "news deserts," therefore, risks reflecting researcher bias more than the real state of local public spheres. Places without local journalism are not devoid of news. What characterises the state of local democracy in such places is, therefore, not that they are news deserts. Instead, platform-dependant local public spheres with almost no independent, professionally produced news and information resemble what I chose to label "casual democracies." The word casual has multiple meanings, many of which characterise the state of local democracy in Frederiksberg and Ringwood quite well. First, the "ghost" local newspapers had a casual, bordering on indifferent relationship with important news related to the pandemic. Of course, most news regarding Covid-19 was national in scope, so people in the two communities had access to lots of edited news about the crisis, informing them about key issues such as infection rates, death rates, hospitalisations, and restrictions. However, they did not have access to edited news about the local consequences of the pandemic. For example: How did the crisis affect local businesses and schools?
Second, the moderation of local news in local social media groups was casual, informal, fluid, and at times quite random. Some local social media administrators had formal positions of power in the communities, especially in Frederiksberg, where local politicians and authorities had professionalised their social media communication. Other administrators had achieved important positions more or less by chance, such as the administrator of the fastest-growing Facebook group in Frederiksberg.
Third, at times, the people who shared news and information with the local public spheres through the various, mostly platform-dependant channels available had a casual relationship with truth and credibility. Determining information's trustworthiness, source, and credibility could be hard. However, such concerns have also been raised about journalism, especially local journalism, which has been subject to ongoing critique by scholars and has had an ambivalent relationship with the public regarding its quality, relevance, and trustworthiness (see Nielsen, 2015). Therefore, this kind of casualness is not necessarily new but might be increasingly relevant in communities left behind by local journalism and dominated by platforms.
Finally, the public spheres of Ringwood and Frederiksberg were marked by a casual, almost ignorant attitude towards the differences between public, private, cooperate, and governmental interests. Much of the news available had been produced by people who predominantly served interests other than the public interest, for example, politicians who served the interests of their political parties and careers, officers with local authorities who served the interests of those authorities, and local business, who served their commercial interests. Of course, these actors might also have the public interest in mind, but their agendas could be difficult to spot. However, it should be noted that local journalism does not always serve the public interest and might also cater to the interests and agenda of others, so this is not necessarily a new problem but one which is enhanced when no independent source of local news is available.
This study has demonstrated the necessity of empirically investigating what local public spheres look like in places left behind by local journalism. Future research should take this into account and not assume how local news and information networks, and eventually local democracy, are affected by a lack of local reporting from professional journalists.