Audience Engagement with COVID-19 News: The Impact of Lockdown and Live Coverage, and the Role of Polarization

ABSTRACT Existing research on media and the COVID-19 pandemic is largely based on quantitative data, focused on digital media, limited to single-country studies, and often West-centred. As such, it has limited capacity to provide a holistic account of the causes and consequences of audience engagement with COVID-19 news, or to consider the impact of systemic political and media factors. To compensate for that, we examine a large set of qualitative interviews and media diaries collected in four eastern European countries during the first wave of the pandemic. We show that changes in news consumption—including the resurgence of television and decline of print consumption—were not driven solely by audience demand for up-to-date information, but also by practical constrains of home-bound life in lockdown, and the introduction of live briefings. Our findings underscore disruption and uncertainty as key elements of audience experiences and highlight the markedly privatized and depoliticized nature of public debate in the early phase of the pandemic. We argue that the pandemic was an unpredictable, open-ended, and exhausting media event with high potential for divisiveness and polarization, especially in contexts marked by low levels of media freedom, declining democratic standards, and elite-led politicization of the crisis.

instance, several studies used large-scale datasets drawn from digital platforms to investigated the prevalence and diffusion of misinformation on digital platforms (e.g., Ferrara 2020; Gruzd and Mai 2020;Islam et al. 2020) or used surveys to examine links between sources of information and susceptibility to misinformation (e.g., Allington et al. 2020;Fridman et al. 2020;Motta, Stecula, and Farhart 2020;Nielsen et al. 2020).
While existing research on media and COVID-19 generated valuable knowledge, it also has some notable gaps. First, the overwhelming concern with the digitally enabled "infodemic" of misinformation often went at the expense of a more holistic understanding of audience engagement. Even though several studies have noted the resurgence of legacy media, there has been little consideration of how audiences combined different digital and non-digital sources in their information-seeking and sense-making practices. Second, existing work is overwhelmingly based on quantitative methods, either population surveys or large-scale digital data sets. As a result, existing research offers plenty of information on broad patterns in media use and news consumption, and on how these correlated with attitudes and behavior, but tells us little about citizens' experiences, responses and motivations-for instance, about what led citizens to prioritize particular sources of information over others, how and why they experienced and responded to the information received, or whether and how they shared or discussed this information with others. Without such qualitative insights it is difficult to understand the causes and consequences of citizens' engagement with COVID-19 news. Third, while the number of comparative studies is growing, single-country studies are by far the most prevalent, and most of existing research is focused on the West. This makes it difficult to appreciate the impact of systemic political and media differences, including factors such as the availability of a trusted public service broadcaster, or the extent of politicization of public health measures.
Our article seeks to address these gaps by drawing on a large qualitative data set comprising two waves of semi-structured interviews and media diaries with 120 participants from four eastern European countries: Czech Republic, Hungary, Poland and Serbia. While the first wave of interviews was conducted during a relatively uneventful period in February 2020, the media diary phase coincided with the rapid escalation of the crisis and the start of lockdown in March. The second wave of interviews was conducted in April and offered an opportunity to discuss news consumption since March and comment on specific trends reflected in diaries. The systemic differences between the four countries also allow us to investigate the role of political and media contexts in shaping audience engagement with pandemic news.
In what follows, we first review existing literature on media and COVID-19, focusing on news consumption patterns and the impact of political polarization and partisanship, and noting relevant similarities and differences between COVID-19 and earlier pandemics. This is followed by an outline of similarities and difference between the four countries, the methodology, and the presentation of key results. We conclude with a consideration of key contributions to existing knowledge on media and COVID-19, while also highlighting contributions to wider theoretical debates on "media events".

Consuming Pandemic News
In existing research on media and COVID-19, news consumption rarely constitutes the central object of analysis. Typically, it features merely as one of several factors that can help explain differences in levels of knowledge about COVID-19, susceptibility to misinformation, or attitudes to public health measures (e.g., Allington et al. 2020;Fridman et al. 2020). Nonetheless, work that focuses more centrally on news consumption (e.g., Casero-Ripollés 2020;Van Aelst et al. 2021;Nielsen et al. 2020;Ofcom 2020) is sufficiently extensive to enable us to reconstruct some of the main trends during the early stages of the pandemic across.
The first of these trends is a sharp rise in overall news consumption, with particularly marked increases in online and television news consumption (Casero-Ripollés 2020;Van Aelst et al. 2021;Nielsen et al. 2020;Ofcom 2020). However, the increase differed considerably across countries; according to a study covering seventeen European countries (Van Aelst et al. 2021), the lowest increase occurred in Poland, Romania and Austria, while the sharpest increase was seen in Germany, Belgium and Sweden. Countries also differed in the relative importance of online and television news; while online news use exceeded television news consumption in Argentina, South Korea, Spain, UK and the US, the opposite was the case in Germany (Nielsen et al. 2020).
Despite the wide reach of online news, legacy media and especially television were clearly in the lead as the most important and/or trustworthy sources of information on COVID-19. In the USA, television coverage was most positively assessed, with national free-to-air channels reaching the highest score of all platforms, and legacy media on the whole outperforming digital outlets (Casero-Ripollés 2020). In the UK, citizens were likewise relying primarily on legacy media sources, in both broadcast/print and online form, and less on social media or digital-only news sources (Ofcom 2020). These tendencies were common across Europe and beyond, with levels of trust in mainstream news media as sources of information on COVID-19 considerably higher than levels of trust in online sources (Nielsen et al. 2020).
In many European countries public service media outperformed commercial competitors. Public service media performed well across much of the continent, saw higher levels of increase in usage than commercial competitors, and also enjoyed a more pronounced increase in trust (Van Aelst et al. 2021). There were, however, some notable exceptions: Israeli, Polish and Norwegian public service broadcasters actually saw a decline in use, while Israeli, Polish and Spanish public broadcasters also experienced a considerable decline in trust (Van Aelst et al. 2021).
Another trend noticed by some studies was the decline in print media consumption during the early stage of the pandemic, tentatively explained in relation to the disruption caused by lockdown restrictions (Newman et al. 2020, 9). Existing research offers some support for this explanation. A study based on US data found that print news initially saw an increase before lockdown (Casero-Ripollés 2020), while a comparative study covering seventeen European countries indicated that the decline in print consumption was more pronounced in the tabloid sector, and almost non-existent in the broadsheet section, with some variations across countries (Van Aelst et al. 2021). That said, the latter study examined consumption by brands, encompassing both print and online versions, and it is therefore possible that the difference between tabloids and broadsheets arose simply because broadsheets were able to compensate losses in print by increases in online consumption.
It is instructive to compare these trends with those seen in earlier epidemics, although it should be noted that none of these earlier events constituted a global crisis of a magnitude comparable to the COVID-19 pandemic. As one might expect, online sources and especially social media came to play and increasingly important role over time. During the 2003-2004 in China and the bird flu (H5N1) outbreak in 2006, audiences were still turning primarily to television and newspapers as the main sources of information (Gaglia et al. 2008;Roy et al. 2020). Social media started playing a more prominent role during the 2009 swine flu (H1N1) pandemic (Chew and Eysenbach 2010) and especially during the Ebola outbreak in 2014-2015 and the 2015-2016 Zika epidemic, when several scholars drew attention to the diffusion of misinformation online (Basch et al. 2017;Sharma et al. 2017). In the context of the growing role of online communication in recent years, the centrality of television and continued importance of legacy media during the COVID-19 pandemic came as a surprise. Yet so far, there has been little attempt to explain these patterns, or investigate how the use of legacy media interacted with the use of online sources.

Public Health Crises, Media, and Political Polarization
A shared concern evident in a significant portion of existing research on media and COVID-19 is the impact of political polarization, which is seen as an important obstacle to successful implementation of preventative measures in public health crises, particularly if combined with a polarized news diet (Van Bavel et al. 2020, 464). Existing research based on US data confirms the existence of polarization both online and offline and lends support to concerns that selective media exposure is linked to disparate attitudes to public health measures. People consuming right-leaning TV channels such as Fox News were more likely to endorse misinformation, or believed that health authorities exaggerated the risks (Motta, Stecula, and Farhart 2020), and greater reliance on Fox News was also associated with reduced compliance with the stay at home message (Simonov et al. 2020). An analysis of online chatter on COVID-19 on Twitter among USA-based users likewise confirmed the debate was polarized, and showed correlations between partisanship, attitudes to government measures, and propensity for sharing public health messages (Jiang et al. 2020).
At first sight, marked levels of polarization and politicization may seem connected to the growing role of social media during pandemics. As previously noted, the 2014 Ebola outbreak was the first major health crisis during which social media paid a central and often negative role. Much of the Ebola coverage was also marked by politicization; news media were mobilizing the disease "as a frame through which domestic politics could be discussed" (Abeysinghe 2016), while social media communication was characterized by a high proportion of politicized posts, othering and blaming (Roy et al. 2020;Sell, Hosangadi, and Trotochaud 2020). In contrast, the media coverage of the swine flu (H1N1) in 2009, when social media communication was more limited, was far less politicized. In the US, news reporting was marked by "a kind of fusion of science, the state, and media, a largely harmonious collaboration between health officials and mainstream journalists" (Briggs and Hallin 2017, 134). Similar patterns were identified in Argentina and Venezuela, where the coverage was likewise largely devoid of disagreement and dominated by public health authorities and experts as key sources (Hallin et al. 2020). The prevalence of the consensual mode of reporting suggests that the 2009 swine flu pandemic was treated in a similar way as national security crises such as wars or terrorist attacks, during which "political divisions are overshadowed to a significant degree by the feeling of threat to the community" (Hallin et al. 2020, 19). To put it differently, the 2009 swine flu pandemic fell into the "sphere of consensus" and was seen as being beyond the realm of legitimate political debate and disagreement, triggering a shift towards a consensus mode of reporting, which was evident in the fact that health authorities were presented in a largely positive or neutral manner (Hallin et al. 2020, 14).
It may be tempting to suggest that such consensual nature of mediated pandemics may now be a thing of the past, impossible to achieve in a fragmented, high-choice information environment of the present. However, existing comparative research on media and COVID-19 suggests a more cautious approach. A comparative study covering Argentina, Germany, South Korea, Spain, UK and USA showed that a measure of polarization existed everywhere, but levels in the USA exceeded those in all other countries, with Argentina scoring lowest on this dimension (Nielsen et al. 2020, 17-20). Furthermore, although the study found correlations between political orientations and levels of knowledge about COVID-19 in Germany, South Korea, UK and USA, the correlation was absent in Argentina and Spain (Nielsen et al. 2020, 27). These differences further underscore the importance of comparative research on media and COVID-19.
With these considerations in mind, our analysis was focused on the following questions: RQ1: How can we explain the key quantitative changes in news consumption noticed in existing research, including the resurgence of TV and the decline of print? RQ2: How did audiences experience, and engage with, pandemic news? RQ3: How did the political and media context shape audience experiences and engagement with COVID-19 news?

Methodology
Our study involved 120 participants, 30 from each of the four countries. A combination of purposive and quota sampling was used, with the sample limited to participants who were at least somewhat engaged with news consumption and who followed politics on a regular basis (minimum weekly). Quota sampling was used to ensure the sample was sufficiently diverse with regards to age, gender, location and political orientation. As a result, each country sample was roughly equally balanced in terms of age (18-34, 35-59, 60+), gender (male, female) and location (urban, rural), and political orientation. The research design was modeled broadly on Couldry, Livingstone, and Markham's (2006) study of public connection, adapted in line with the study's research aims. Each participant was asked to participate in two interview sessions-one face-to-face in February and one remotely in April-as well as to keep a diary for three weeks from 9 to 29 March. In their diaries, participants were asked to include descriptions of all encounters with information, broadly defined, regardless of whether these encounters involved media or not. For each encounter, they were asked to record the time, location, news sources and device, followed by a brief description of the news, why they used the particular news source or device, whether they had any thoughts on, or reactions to, the news, and if they did anything in response (e.g., shared news on social media). The first interview lasted an hour on average, and covered everyday routines, political engagement, the media environment, and news consumption routines and preferences. The second interview lasted 40-45 min on average and covered recollections of news consumption during the diary period, participants' views on various issues concerning the pandemic, and follow-up questions on selected aspects of the participant's diary.
All interviews were transcribed, and paper-based diaries were digitized. All material was analyzed in NVivo, using thematic analysis (Braun and Clarke 2006). In the first step, individual profiles were generated for each participant based on the first interview, covering news preferences and attitudes. These profiles were then compared with the information from diaries and second interviews to identify key changes in news habits during the pandemic. In the second step, we focused on diaries and second interviews and examined three key themes: (a) participants' descriptions and explanations of changes in news routines (e.g., having more time for news due to lockdown); (b) their general experiences of COVID-19 news (e.g., disruption, frustration, etc.), and (c) their modes of engagement with news (e.g., discussing news with friends, family or colleagues, sharing news on social media etc.).
We should note that our original plan for data collection did not foresee the arrival of a pandemic; our intention was rather to use the combination of interview and diary data to examine news consumption during a relatively uneventful period. As it became obvious that the period covered by our diaries will be anything but uneventful, we adapted our plans to enable us to examine the changing news consumption during an unprecedented health crisis.

Political and Media Context
While all four countries share the experience of the communist past and geographical proximity, there are considerable differences between their political and media systems, reflecting divergent trajectories of their post-communist transition and contemporary political developments. In the Czech Republic, Hungary and Poland, democratic consolidation and establishment of a pluralistic media system, including the institutionalization of public service broadcasting, were largely completed within the first decade of the fall of communism (Jakubowicz and Sukosd 2008;Downey and Mihelj 2012), and the three countries have enjoyed a period of relative political stability and economic growth since then. However, recent years have seen significant democratic backsliding in Poland and especially in Hungary, with governments criticized for eroding the rule of law and actively undermining democratic institutions, especially the judiciary and independent media (Vachudova 2020;Surowiec and Štětka 2020).
While the process of democratic backsliding has been affecting the Czech Republic as well (Hanley and Vachudova 2018), the country still fares considerably better than Poland and Hungary in terms of both the quality of democracy and media freedom. This is particularly visible with regards to its public service broadcasting channels, which are not only more politically independent than their Hungarian and Polish counterparts (Brogi et al. 2020), but are also significantly more trusted by the public (Newman et al. 2020). Serbia, on the other hand, followed a rather different path, having undergone a "delayed transition" to democracy following the end of Slobodan Milošević's autocratic rule in 1999, and never fully completing the process of democratic consolidation (Greenberg 2010). The state of media freedom has worsened again in recent years under the populist right-wing government and subsequent presidency of Aleksandar Vučić, with rising attacks on journalists, significant collusion between politicians and private media, and government control of public service broadcasting.
The systemic differences between the four countries are to an extent reflected in different government responses to the COVID-19 pandemic. While all four countries reacted rapidly and introduced a state of emergency around the same time in mid-March, the pacing of preventative measures and levels of policing differed, and so did the extent of politicization of the pandemic. In the Czech Republic, government acted swiftly, mandating a two-week self-quarantine in early March, followed by the closure of all educational institutions, borders, and universal mask wearing by mid-March (Government of the Czech Republic 2020). The level of politicization was relatively low, and although President Zeman caused a stir with his promotion of aid received from the Chinese government (Stojanović 2020), the government's handling of the pandemic remained largely uncontested. In Poland, the introduction of more stringent measures took longer, as school closures and a national lockdown were introduced only on 21st March, followed by the closure non-essential services in April (Watoła 2020). The Polish right-wing government also caused significant controversy with its handling of the upcoming national election, originally scheduled for 10th May 2020. In early April it pushed through a controversial bill allowing for the elections to take place fully by postal voting, prompting accusations of narrow political self-interest from the opposition (Vashchanke 2020, 7).
As in Poland, the introduction of more stringent measures in Hungary lagged somewhat behind Czech Republic, with a full lockdown introduced only on 27th March. However, Hungarian authorities went considerably further than their Polish counterparts in abusing the pandemic to extend their powers. On 30th March the Hungarian parliament voted on the new "Coronavirus Law", giving Prime Minister Orbán the power to rule by decree and imposing further limitations on media freedom in the country (Walker and Rankin 2020). The following day, the Hungarian government banned gender change in the country, and announced that disseminating "fake news" about the pandemic was punishable by up to five years in prison (Beauchamp 2020).
In contrast to the other three countries, Serbia adopted considerably more drastic measures, and policed them more heavily. Following the introduction of the State of Emergency, the army was dispatched to guard the borders and important buildings, and several hundreds of people were arrested for flouting the rules. A controversial police curfew was introduced in late March, followed by weekend-long curfews until early May. Much like Poland, Serbia was also gearing up for parliamentary elections, and although these were postponed swiftly, the government nonetheless attracted criticism for abusing the pandemic for political gain, with opposition parties calling for civil disobedience in face of harsh curfew measures. Media freedom was curtailed as well, with a journalist arrested for publishing a story about the lack of PPE, but later released (Stojanović 2020). As we shall see, audience engagement with pandemic news was to an important extent shaped by these contextual differences.

Results and Discussion
The analysis confirmed that the first wave of the pandemic was accompanied by a major disruption of existing news consumption routines. A large majority of our participants reported consuming news considerably more often than they normally would, and most also mentioned watching more television and relying on public service broadcasting to a considerably greater degree, with some mentioning less print media consumption. While these results largely echo the findings of existing research of news consumption during the pandemic (Casero-Ripollés 2020; Van Aelst et al. 2021;Nielsen et al. 2020;Ofcom 2020) our analysis goes a step further by examining (a) why these quantitative changes occurred, (b) how they were experienced by citizens, and (c) what forms of engagement they were associated with. The three sections that follow focus on each of these in turn, drawing on one of the three key issues examined in our thematic analysis, namely explanations, experiences, and forms of engagement.

Explaining Changes in News Consumption During the Pandemic: The Impact of Home-bound Life and Liveness
It is tempting to assume that changes in news consumption resulted primarily from public demand for up-to-date knowledge in the context of heightened uncertainty. Indeed, one of our Polish participants did explain his changing news habits in these terms, explaining that the newspaper she normally reads "couldn't keep up with the changing news and lagged behind" (Pol-04, female, 62, rural). Yet, the inability of print media to provide the kind of timely, live coverage of events available through broadcast and digital media was clearly a secondary reason for out participants' reliance on television. Rather, our analysis suggests that the increase in television viewing, and to some extent the increase in news consumption more generally, was driven primarily by practical constrains of life in lockdown, linked to the fact that people were spending considerably more time at home. This home-bound nature of daily life impacted on media use in three interconnected ways: it enabled people to spend more time on consuming news, it encouraged greater reliance on domestic media technologies, disrupted access to news sources outside of the home, and increased the influence of family members on one's media routines.
The reliance on domestic media appliances was particularly common among workingage participants who were either furloughed or suddenly found themselves working from home, but also among older, retired participants who were prevented from engaging in activities outside of their homes. The following excerpt, taken from the Czech Republic, is illustrative of this wider trend: If I had gone to work, I certainly would not have followed news so much because I wouldn't have had enough time for it. But as I was at home more or less the whole [diary] period I had so much time, but on the other hand I didn't have enough time to read anything because children were at home and I had some duties to attend to, so I listened to a lot. male,44,rural) The homebound nature of daily life during lockdown also disrupted access to news sources outside of the home. As our data suggests, these practical obstacles were the key reason for the decline in press consumption, confirming the tentative explanation offered in existing survey-based research (Newman et al. 2020). As one of our Polish participants explained, "press is hard to come by these days", as many salespoints he would normally use were shut male,23,rural), while a Serbian participant noted that he stopped reading the newspaper because it was no longer delivered, and he did not want to pay for it twice by buying another copy in the shop male,26,urban,media diary 25 March). For one of our Hungarian participants, these practical constraints combined with worries about newspapers inadvertently transmitting the virus: I try to minimise buying things that other people have touched. I know that the virus only survives for a short time, for 24 hoursaccording to the latest researchon a newspaper, but no, no, I go crazy from these things. female,59,rural) There was no indication of differences between tabloids and broadsheets in our data; both seemed equally affected by practical constraints. This confirms that differences between tabloid and broadsheet consumption noted in prior research (Van Aelst et al. 2021) likely arose due to the fact that broadsheet brands were able to compensate for any losses in print copy sales through an increase in the consumption of online versions.
The third effect of the home-bound nature of lockdown life was the influence of family members. For some young adults who moved in with their parents, the increase in television consumption was entirely down to family influence. One of our Polish participants in his mid-twenties explained that he "didn't watch television before, because I have no television set at my place", but when he moved in with his family during lockdown, he "started watching television with them" (Pol-14, male, 25, urban). Alternatively, parents' news consumption habits changed under the influence of their children's' preferences, as was the case with another of our Polish interviewees: I watched Wiadomości quite often, because this is my son's favourite news bulletin. So, I started watching more of the news bulletin on TVP1. Earlier, I didn't really watch Wiadomości because I'm not sure I can consider it reliable. I watched TVP more in that period but that was because my son would switch it on. female,47,urban) Our analysis also revealed the centrality of live coverage in people's news consumption, both in the sense of following live broadcasts of government briefings, as well as in the more general sense of keeping abreast of all developments "live", as they happened. Indeed, it was clear that the pandemic functioned in many ways as a "media event" (Dayan and Katz 1992)-an extraordinary, often history-defining event, organized outside of the media but broadcast live, which interrupts the normal flow of daily life and invites collective viewing. Across all four countries, live government briefings were a central element of news consumption, especially in the early phase of lockdown. These broadcasts typically interrupted the routine flow of the day and often involved collective viewing or listening with family members or, before lockdown, with colleagues at work. The rise in television viewing can in part be explained by the importance of live coverage, as television has traditionally functioned as the preferred medium of live connection and features prominently in Dayan and Katz's (1992) classic analysis of media events. However, as our analysis showed, a significant number of our participants-especially those in the youngest age group-preferred to engage with live briefings through digital media rather than television. The following excerpt, taken from one of our Hungarian interviews, offers a case in point: Of course, and at the beginning, 'my god' … even at work, we sometimes gathered up, 'my god, Viktor Orbán's speech is coming, let's watch the live on Facebook, what will he announce now, my god, what happened now' … And of course, a lot of people panicked about the possibility of a complete lockdown. female,28,urban) The extract also suggests that such digitally enabled participation in media events was capable of generating a similar experience as the one typically associated with media events watched on television-namely, they were marked by a sense of extraordinary occasion, and involved collective viewing, in this case with colleagues at work.
Even for those participants who did watch the briefings over television, using a traditional TV set, television was often not the only medium used. Rather, several of our participants engaged in the practice of dual screening-namely, while following the briefings on television (or on radio) they also simultaneously used other, typically digital sources, to find additional information or discuss latest developments with family and friends. The following extract, taken from the diary of one of our Serbian participants, illustrate this hybrid practice well: At 8pm I watched TV with all household members. We watched Vučić's live address, when he announced the state of emergency in the country. This led me to look for information on whether or not it is legitimate for the president to announce a state of emergency, having previously thought that only the national parliament could do that. I found relevant information online. female,28,rural,media diary 15 March) From this perspective, audience engagement with COVID-19 briefings thus resembled what Vaccari, Chadwick, andO'Loughlin (2015, 1043) called a "hybrid media event," during which audience members were able to draw on "hybrid articulations and recombinations of media", taking advantage of "a more complex mix of affordances" (see also Sumiala et al. 2018). According to Vaccari, Chadwick, and O'Loughlin (2015), the multiple affordances of a hybrid media environment has the potential to blur and complicate the relationship between active and passive practices, enabling audiences to engage in a range of activities from a relatively passive reading of information about the live broadcast as it unfolds, to more active seeking of additional information (as in the case of the two examples quoted above) or commenting on the briefing on social media. However, as shown later on, our participants largely avoided the more active forms of engagement, or chose to limit such active engagement to private communication with likeminded friends, family and colleagues.

Experiencing Pandemic News: Disruption, Uncertainty, and Division
A shared trait of our participants' experience of news during the peak of the pandemic in March and April 2020 was a sense of disruption, uncertainty, and even division. Unlike the media events examined by Dayan and Katz (1992) in their classic study, the live briefings were evidently not planned months in advance, nor had a clear beginning and end. Rather, they were organized hastily in response to the rapid spread of the virus, and were often announced suddenly, only hours before they happened. Even once live briefings became more firmly established and turned into a regular fixture of daily life for our participants, there was little sense of how long they might continue. The sense of uncertainty and open-ended disruption extended beyond live briefings, and was abundantly present in the accounts of our participants across all four countries, with the majority of participants reporting a sense of exhaustion after the first seven to ten days of intense, marathon-like news consumption. The following excerpt, taken from an interview with of one of our Hungarian participants, echoes the sentiments of many: for a while, you can bare following the media closely from morning to evening, or listening to it or reading it … Then you need a few days … a little break, because you can't … simply can't [bare it]. female,62,urban) The pronounced sense of sudden disruption and unpredictability suggests that the coverage of the pandemic shared the traits of what Liebes called a "disaster marathon" (Liebes 1998)-that is, an extended, unpredictable, open-ended media event, involving live coverage of a catastrophic event, such as a flood, earthquake, or a terrorist attack.
In line with the hybrid affordances of the media environment, the marathon-like news consumption during the first wave of the pandemic often involved engaging with a mix of online and offline news sources. A good example is provided in the following diary excerpt from one of our Polish participants, who followed the developments of the day through the popular web portal Virtual Poland (Wirtualna Polska), through Facebook, as well as via the news-only public service TV channel TVP INFO, and commented on the activities of the parliament ("Sejm") and Jarosław Kaczyński, leader of the rightwing populist party Law and Justice ("PiS"): In the morning: An overview of events on Virtual Poland and Facebook. Great way to wait for the first remote parliamentary vote in the country's history. Many voices are saying that the system is failing, along with its authorities. Many deputies decided to appear in the Sejm in person, including Kaczyński, who is not afraid of anything. He is a deputy who represents people associated with PIS: Tracking TVP INFO and news about the situation in the country and the world. The coronavirus is continuously evolving. At noon, debates in the Sejm begin with remote talks, there are many problems, but with over an hour-long delay, he manages to start. Three hours of speeches, people from all parties. male,30,urban,media diary March 27) As this and other excerpts from diaries and interviews suggest, the pandemic functioned as a hybrid disaster marathon-an intense, prolonged disruption that led most of our participants to become more or less constantly immersed in the flow of news, using a variety of news sources, at least during the initial period of lockdown.
Another feature of audience experiences of pandemic news, and especially of live briefings, was their potential for divisiveness-although this feature was prominent in only two of our countries. From this perspective, too, a comparison between the coverage of the COVID-19 pandemic and classic media events is instructive. While media events examined by Dayan and Katz (1992) were celebratory, integrative occasions that temporarily disrupted daily routines in order to encourage societies to reconfirm their commitment to the existing social order, or celebrate consensual change, the pandemic was profoundly destabilizing, stopping society in its track without a clear sense of when, and indeed whether, a return to normal might be possible. Maintaining unity and public support in such a context was challenging, and our analysis suggests that the four countries fared rather differently in this respect. The differences were particularly clear in relation to audience responses to government briefings.
In Serbia and Hungary, government briefings provoked starkly different responses, and generated division rather than unity. The contrasting responses to briefings largely coincided with attitudes to government. In both countries, participants who were supportive of, or neutral towards the government, generally accepted what was said in the briefings and how they were conducted, while those participants who were anti-government generally tended to be irritated by the way the briefings were conducted, were distrustful of the information provided and even scared, or poked fun at the government and the experts. The two extracts below, taken from our Serbian sample, illustrate the contrast well; while the pro-government participant "completely trusted" the information provided during live briefings, the anti-government participant felt these events were used to "brainwash" people, and made him feel scared. I found those conferences at 3PM extremely significant. I was listening to what the doctors were saying, and I still check online what has happened in the last 24 hours. I completely trusted what they were saying on those conferences, even though people were saying different things about them, but I found them completely acceptable. (Srb-23, female, 60+, rural) I think that the government was using those conferences to brainwash us.
Interviewer: You think that the government was using the conferences to manipulate the people?] Yes, I do.
Interviewer: Did the presence of those doctors make you feel safer?
No, they made me feel scared, especially when they were talking about the number of the deceased, especially about the number of the deceased in the gerontological centres and in the rest of the world. I was really scared to be honest. male,48,rural) In Poland and especially in the Czech Republic, participants were far less divided in their responses to live briefings, suggesting that in these two countries, the briefings had a more integrative effect, despite the sense of disruption and uncertainty they provoked. Even though participants were occasionally critical of some of the measures introduced, or resented the constant presence of government officials in the media, they turned to official briefings, the government website, or to the public service broadcaster, to find up-to-date, trustworthy information about the pandemic, or keep up to date on the latest measures they needed to follow.

Suppressing Disagreement: Privatization and De-politicization of Public Debate
Another clear trend highlighted by our data was a significant increase in discussion of news among friends, colleagues, and family members, both on-and off-line. Yet, a closer analysis of the nature of these discussions revealed that they often skirted around controversial and divisive aspects of the pandemic when discussing news publicly, or limited expressions of criticism to private communication among likeminded family members, friends, or colleagues. To put it differently, the significant increase in audience engagement with news went hand in hand with depoliticization and privatization. This was true even in Hungary and Serbia where participants were deeply divided over government responses to the pandemic, but largely avoided making their criticisms public, except when communicating with likeminded friends, family, or colleagues.
As our analysis showed, the increase in discussion was in part driven by lockdown and quarantine restrictions and their impact on daily routines. People who lived with their family members were spending more time with them due to lockdown. As one of our Polish participants explained: "I was stuck at home and there was more time to discuss things" female,38,urban). Many participants also reported more frequent phone calls and video chats with extended family and friends, with one of our Czech participants describing "online debates", accompanied by drinking, during which "we debated everything" male,39,urban). Although political aspects of the pandemic did come up in these discussions, they were often secondary to other issues-such as checking on friends and family and their wellbeing or updating less well-informed relatives of new measures and rules. Some participants also reported purposefully avoiding the discussion of COVID-19 because they found the constant focus on the pandemic stressful and tiring. One of our Polish participants explained this in a manner that echoed the sentiments of many: "I tried avoiding talking about the news so as not to go crazy because literally all of the news was on coronavirus," male,34,urban).
When political issues were broached more directly in the presence of family or friends with differing opinions, this was done in a manner that avoided open expressions of disagreement. For instance, as one of our Hungarian participants explained, his discussions friends and family did focus on different opinions, but this was driven primarily by his interest in other peoples' opinions, and in how they might differ from his own, without emphasizing disagreements: One the other hand, participants who did report discussing controversial topics or voicing critical opinions usually did so with friends and family who shared their views. As one of our Serbian participants wrote: "Of course, this situation led me to spend more time discussing these issues with members of my household, who more or less share my opinions" female,28,rural,media diary 29 March).
Similar tendencies were noticeable in people's use of social media. While many participants reported using social media considerably more often, this was limited primarily to scrolling through the newsfeed, without commenting or sharing. The avoidance of active engagement on social media was evident already in the first wave of our interviews in February, yet this tendency became even more pronounced during the pandemic, with several participants reporting that their active engagement decreased even further. Not all participants articulated why they preferred to use social media passively; the answers of those who did, however, suggest that the principal reason was conflict avoidance. The following exchange with a Czech interviewee provides a case in point: … I joined a group on Facebook, COVID-19. I wouldn't have done it if it weren't for this situation.
Interviewer: Did you comment on it?
No, I do not want to get involved in discussions because I tried it once, a long time ago, and it was not pleasant. They called me names of various animals and so on. So, I only read the discussions from time to time, just for fun. female,44,rural) One of our Hungarian participants also mentioned avoiding sharing on social media for fear of government repression, following reports of arrests of people accused of spreading "fake news" about the pandemic. As she explained later in the interview: I heard on the radio … um … that they arrested and took action against some people because they spread fake news on Facebook, and this happened a few days ago. Well … often, um … I am anxious about … I don't know … they will knock on my door at dawn … the police. (Hun-05, female, 64, rural).
This was an isolated case, and the participant herself noted that the self-censorship did not last, but it can nonetheless be seen as indicative of a more general climate of fear generated by declining standard of democracy and media freedom in Hungary.
While avoiding active engagement through public posting and commenting, several of our participants were rather active in closed groups on Messenger, WhatsApp or Viber. For instance, one of our Polish participants created a closed WhatsApp group called "Family" with her children and their partners, and her sister, and used that to exchange news and political Internet memes (Pol-04, female, 62, rural). Other noticeable trends among some of our participants included joining groups dedicated to topics unrelated to the pandemic, such as arts and crafts, and sharing comments and images intended to counterbalance the negativity of pandemic news with positive messages or humor. A good example of the latter is found in an interview with a Czech participant, who shared a light-hearted message on Facebook friends, which "improved the mood of a few people who were locked in their homes." (Cze-09, male, 54, urban).
It may be tempting to interpret such privatization and politicization as evidence of a "spiral of silence" (Noelle-Neumann 1974), rooted in people's fear of social isolation, which leads individuals to silence their opinions if they perceive that these opinions are in the minority. However, out data lends little support for such an interpretation. Most of our participants perceived public opinion to be divided, without a clear sense of which views might be in the majority or minority. The reluctance to voice their own opinions was therefore motivated not by fear of isolation, but primarily by a desire to avoid conflict, combined with a general sense of exhaustion with the pandemic. Rather than being linked to a "spiral of silence", privatization and depoliticization were arguably linked to the tendency towards unity and consensus in the context of crisis, as discussed in the literature review-a trait characteristic of historical pandemics such as the 2009 H1N1 outbreak, as well as other major disruptive events such as wars, terrorist attacks, or natural catastrophes, which fall into a "sphere of consensus" and are seen as being beyond the realm of legitimate political debate and disagreement (Hallin et al. 2020).

Conclusions
In sum, our analysis brought three interrelated empirical insights which help explain and interpret the causes and consequences of quantitative shifts in news consumption registered by existing research. First, we took a closer look at the reasons for quantitative shifts in news consumption, including the "return" of television and the decline in print media consumption. We showed that these changes were driven largely by the specific nature of policy responses to the COVID-19 pandemic-namely, the imposition of nation-wide lockdowns and the introduction of live briefings. These responses encouraged greater reliance on domestic media technologies, especially television, increased the influence of family members on one's own news consumption, and disrupted access to media outside of the home, thereby creating obstacles to print media consumption. Second, we highlighted disruption and uncertainty as key elements of audience experiences during the early period of the pandemic, and suggested that the pandemic acted as a disruptive, unpredictable, and exhausting media event with a high potential for divisiveness, especially in political contexts marked by low levels of media freedom, declining democratic standards, and politicization of the pandemic. Third, we looked closely at the nature of audience engagement with news and showed that it was marked by privatization and depoliticization; it mostly took place in private settings, was limited to likeminded groups, or skirted around controversial and divisive topics.
Throughout our analysis, we also developed a set of arguments about the COVID-19 pandemic as a "media event" (Dayan and Katz 1992). We noted the features that the pandemic shared with the global rituals examined by Dayan and Katz, including its extraordinary, history-defining nature, the centrality of live coverage, its disruptive effect on the routine flow of daily life, and a sense of collective participation. At the same time, we also highlighted several dimensions that made the pandemic stand apart. The COVID-19 crisis evolved in a media environment in which the centrality of broadcasting has given way to a "hybrid media system" (Chadwick 2013) in which television, radio and print interact with digital platforms. In this context, the pandemic operated as a "hybrid media event" (Vaccari, Chadwick, and O'Loughlin 2015;Sumiala et al. 2018), with audiences combining television broadcasts with social media and news portals.
Our findings also showed that the pandemic was, in contrast to classic media events, marked by a much greater sense of disruption and unpredictability, and as such shared the traits of "disaster marathons" (Liebes 1998). That said, pandemic coverage clearly lasted for longer than the disaster marathons accompanying natural catastrophes or terrorist attacks, and was considerably more global in scope, implicating millions around the globe not merely as spectators, but as directly affected potential victims. As such, it also generated a more widespread and long-lasting sense of uncertainty than a natural catastrophe or terrorist attack could.
Finally, the pandemic also lacked a sense of integration and unity associated with classic media events. Although participants across all the four countries largely avoided openly voicing their disagreements in the public domain, and either circumvented controversial issues altogether or discussed them only with like-minded friends and family members, Hungarian and Serbian participants were clearly divided in their attitudes to governments' handling of the pandemic. In this context, the live briefings did not engender a sense of unity and participation, but rather fostered division, with pro-government participants largely satisfied with official responses, and anti-government participants palpably frustrated and angered. These cross-country differences were arguably rooted in systemic differences, including relatively lower levels of media independence and democratic standards, as well as more visible abuses of the crisis for political gain. This result confirms that the COVID-19 pandemic did not have a uniformly divisive, polarizing effect everywhere in the world, and indicates that its impact varied considerably depending on systemic factors as well as specific crisis management policies and political developments.
We should note that our results reflect the situation during the first wave of the pandemic, and that much has changed since then. In particular, Poland and Czech Republic lost some of their comparative advantage over the course of autumn 2020 and Winter 2021. Having emerged from the first wave relatively unscathed in terms of the number of infections and casualties, all four countries, but especially Poland and the Czech Republic, have found themselves at the opposite side of the international charts during the second wave in Autumn 2020 (Sirotnikova et al. 2020). The government confidence index has consequently dropped dramatically, from 70% in March down to 40% in December 2020 in the Czech Republic, with similar gap recorded in Poland as well (European National Panels n.d.). Further research should therefore focus on the comparison between the different waves and stages of the pandemic, including the vaccine rollout, which might see the above outlined patterns of audience engagement and crosscountry differences shifting, along with the changing context and shifts in the governments' management of the pandemic.