Background and context

The World Health Organization declared Covid a pandemic on March 11, 2020, and the United Kingdom (UK) government imposed a national lockdown on March 23, 2020 [1, 2]. People were asked to stay home, with only essential outings allowed. All restaurants, pubs and bars closed, except for take-away outlets [3, 4]. These measures significantly impacted the food system, resulting in widespread panic-buying and stockpiling [5]. In the week before lockdown, supermarkets saw record sales [3, 6]. Panic-buying, the hoarding of household goods due to scarcity or fear of shortages [7], was widespread during the pandemic, exacerbating food supply chain issues.

Government and retail response to panic-buying

Panic-buying was identified as a risk during no-deal Brexit preparations [8, 9]. During Covid, the government was criticised for relying on supermarkets, poor communication, and inadequate recognition of the food supply chain [10–15]. The UK’s just-in-time supply chain struggled with fluctuating demand [10, 16] and consumers focused on long shelf-life products amid shortages [17–20]. 83% of consumers had trouble accessing these foods, worsened by limited online delivery slots [11, 19, 21]. To address panic-buying, the government relaxed competition laws and supermarket driver limits, while supermarkets rationed products and introduced special hours for key workers [11–13, 22]. The UK parliamentary Environment Food and Rural Affairs Committee deemed the government’s food system resilience plan for Covid inadequate, citing poor communication and lack of public reassurance [10, 12, 13]. The government argued retailers were best suited to handle supply chain communication but faced criticism for shifting responsibility to them [11, 12, 14].

Panic-buying and social inequity

Panic-buying contributed to food system inequity, with higher socio-economic groups able to stockpile more readily [16, 23–25], while those on Universal Credit couldn’t shop for the best prices due to limited availability [17]. The vulnerable, marginalised, and disabled lacked food access [11, 13, 23–26]. Panic-buying also reduced the stock available at food banks [27]—a food security lifeline, which almost doubled in demand during the pandemic [7, 10, 28].

The media’s role in shaping public behaviour

While panic-buying is often understood as a reaction to uncertainty, the role of the media in potentially amplifying these behaviours is important. The media is a primary source of information during crises, informing public perception and influencing behaviour through framing and reporting choices. Evidence suggests that increased exposure to information about Covid directly increased panic-buying [29]. At the beginning of the pandemic, 24% of the UK population accessed Covid news 20+ times a day, 99% at least once a day and 43% used newspapers as sources of information [30]. Ting et al. [31] refer to this information overload as an ‘infodemic’ and argue that it impinges consumers’ ability to establish trustworthy information sources.

Research gaps in panic-buying studies

Panic-buying is episodic, rare, and often specific to certain locations and disaster circumstances, limiting research potential, especially for empirical studies and lived-experience research [32–35]. Although academic literature on panic-buying peaked during Covid, with 85.7% of articles published between 2020 and 2021 [36], it remains under-researched [26, 32, 37]. Specifically, studies on the relationship between panic-buying and media are sparse [26, 37–42], with no media content analyses of Covid panic-buying in the UK press. However, there is a richer body of research on Covid panic-buying and social media [21, 28, 43–52].

The causes of panic-buying

Research suggests four key factors which drive panic-buying: perceived scarcity, anticipated regret, fear and the need for coping mechanisms [32]. A study by Sherman et al. [29] also found that heightened anxiety contributed to increased panic-buying behaviours. Panic-buying can also be seen as a rational response to unusual circumstances where stocking up for longer periods is a means of survival [26, 43, 53–56]. Sherman et al. [29] argue that labelling the behaviour as panic-buying may be a misrepresentation, as it fails to account for both anxiety-driven actions and rational preparedness, both of which can explain consumer stockpiling during a pandemic.

Sensationalism and panic-buying

The literature suggests that the media can play a role in causing panic-buying by using sensationalised language and images of empty shelves, which create scarcity cues [21, 26, 37, 53, 55, 57, 58]. According to Rajkumar [38] poor quality reporting in Covid likely exacerbated panic-buying. One global media content analysis showed 67.3% of the news articles during the pandemic had images of empty shelves and over half reported specific food items in short supply [37]. Ntontis et al. [33] argue that the media’s description of specific foods increased demand of said goods. Media headlines during Covid were often sensationalised [38, 39] and the media may have magnified the potential risks of panic-buying, encouraging people to stockpile.

Prior to the first UK lockdown, English newspapers reported that people had started ‘hoarding for a coronavirus outbreak’ [59]. Examples of sensationalised language were words such as ‘rampage’ and ‘riot’. Newspapers featured images of empty shelves and reported food shortages. Panic-buying had also been reported by UK media during Brexit [60], the 2021 lorry driver crisis [61], the 2022/2023 Ukraine war [62], and during the 2023 fruit and vegetable shortages [63].

Theoretical frameworks and media influence

This study draws on agenda-setting theory and framing theory to examine how the media influenced public behaviour during the Covid pandemic. Agenda-setting theory explains how news media directs public attention by emphasising particular issues and how they are presented [64, 65]. Framing theory complements this by focusing on the way media outlets influence public perceptions through their choice of language, imagery, and tone when reporting [66]. Both theories are highly relevant during crises when the media play a critical role in providing information and directing public opinion [64, 65, 67]. Additionally, this study draws on the view that photography and imagery used in media reporting can steer consumer behaviour by shaping visual perceptions of scarcity or abundance [68, 69].

The literature indicated that the media during Covid often blamed consumers, industry and government for panic-buying. Individuals were called greedy and selfish, and accused of depriving others of food [37, 39]. In contrast, supermarkets were portrayed as heroes and were the most quoted stakeholders [39].

By framing panic-buying as a crisis through images of empty shelves and sensationalist headlines, the media may have reinforced public fears and driven behaviours that exacerbated the problem. By applying these theories, the study seeks to understand the media’s role in shaping the public’s response and influencing policy debates around food system resilience.

Research aims and contribution

Arafat et al. [37] argue that in using terms like ‘panic-buying’ and images of empty shelves, the media engaged in ‘harmful reporting’ about the availability of goods. Ntontis et al. [33] confirm these findings in their study of UK stockpilers, and Coleman et al. [60] found that media framing during Brexit may have contributed to panic-buying.

The evidence calls for the implementation of media reporting standards, restrictions on media reporting and monitoring mechanisms [37, 38, 70–73]. Dulam et al.’s [73] modelling study found that reducing media coverage of panic-buying by 50% led to a decrease in demand and increased product availability. The literature emphasises that images of empty shelves (often seen as a symbol of panic-buying) and sensationalist language should be minimised to prevent reinforcing panic-buying behaviour [39, 70, 74].

This study aims to analyse the agendas set by print media that informed public opinion, by analysing text and visual data. The media can set the agenda by making one element of the news more prominent than another or by framing news items in a certain way [65]. Visuals, including photos, symbols, and imagery, influence how the public and policymakers interpret the significance of these events [75–77]. The findings suggest that media analysis can be a valuable tool for understanding food system issues by highlighting: the media’s role in assigning blame, revealing which items were perceived as scarce, showing patterns in consumer behaviour and offering real-time commentary on both causes and effects of panic-buying. The study contributes to agenda-setting theory and holds important implications for food policy, particularly in enhancing resilience and communication strategies for future food system disruptions.

The research also addresses the gap identified in the literature: there has been little analysis of how the British press portrayed panic-buying during Covid. A better understanding of media portrayals is essential for improving future crisis response, as highlighted by the Institute for Fiscal Studies [19].

Research questions (RQ)

1. RQ1: What were the trends in UK media reporting on panic-buying from March to July 2020, and what imagery was used?
   This question examines how often panic-buying was reported and explores the visual cues, such as empty-shelf images, that may have reinforced panic-buying behaviours [26, 53, 57].
2. RQ2: What were the dominant themes and narratives, and how was the issue framed?
   This question looks into the key themes, narratives, and sources used in media reports, evaluating whether the coverage was balanced or sensationalised, and how it may have influenced public and policy views [55, 58].
3. RQ3: What insights from media portrayals of panic-buying can inform future policy planning for food system resilience?
   This question seeks to understand how media coverage informs policymakers in planning for future crises, including resilience strategies, communication improvements, and the role of supermarkets in managing food supply shocks. By answering these questions, the study will provide important insights into the communication of food system crises.

Sampling

Despite declining newspaper readership, the national press still reaches a substantial audience, with a daily circulation of 10.4 million and 49% of adults claiming to read an online or print newspaper daily [82]. This makes newspapers a credible and relevant source for media content analysis, particularly during a period of public crisis. The lead author selected six titles and their Sunday equivalents from the highest-circulating newspapers: The Daily Mirror, The Sun, The Independent, The Guardian, The Metro and The Daily Mail [83]–two news brands from each genre. These titles were chosen to reflect a broad range of media genres, political stances, and readership demographics. The study included a mix of popular, mid-market, and quality titles, ensuring a comprehensive analysis across the spectrum of British newspapers [84].

Crucially, the political stance of these newspapers is to direct their framing of panic-buying. For example, The Sun and The Daily Mail are traditionally associated with right-leaning perspectives, whereas The Guardian and The Independent are left-leaning. Such political biases can affect how the causes and consequences of panic-buying are portrayed—whether it is framed as a failure of government policy, consumer behaviour, or supply chain issues. By including a range of political stances, the analysis aims to capture how ideological perspectives shape media coverage of the pandemic.

Furthermore, titles were selected to capture a wide range of readership demographics. The Sun and The Daily Mirror primarily reach C2DE audiences, while The Daily Mail appeals to a higher percentage of 65+ ABC1 audiences. The Guardian and The Independent attract a larger proportion of younger readers, particularly 16–24-year-olds [85]. These variations in readership guide how newspapers frame and present issues such as panic-buying, with each outlet tailoring its content to resonate with the unique concerns and interests of its audience.

Most newspapers operate as multi-platform news brands, offering both print and online editions. To enhance the scope of this study, both print and online titles were included in the analysis. This dual approach not only broadened the reach of the study but also ensured that the portrayal of panic-buying was captured across different formats. Given that online readership tends to be younger compared to traditional print [79], analysing both platforms provided a more nuanced view of how the issue might be perceived by different age groups [82].

A time-period of March 1 2020 to July 31, 2020 was chosen to incorporate the pre-lockdown period, when stockpiling began, and the easing of restrictions in August 2020 [2]. The geographical location was restricted to England. A list of articles used in analysis is available in S1 File.

Search strategy

The lead author used a methodical process to reach a final sample of articles and images. This involved searching for articles, removing irrelevant articles, and reading the remaining articles in full and assessing against an inclusion/exclusion criteria [Fig 1]. The lead author used Nexis, a news and business research database to access data and the #atleast3 search limiter for ‘panic-buying’ or ‘stockpiling’ AND food. Headlines were read and irrelevant articles excluded, with remaining articles exported for further analysis. Articles were excluded if they were 1) advertorials, comments, duplicates or letters 2) not related to panic-buying or stockpiling and food or related to non-food goods or Brexit and 3) not in England.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Flowchart of article selection.

https://doi.org/10.1371/journal.pone.0315142.g001

The search strategy resulted in a final sample of n = 209 articles. For the visual content analysis, the lead author required the media articles in situ to view the images and place the data into relevant contexts [86]. However, one significant limitation of Nexis is that it does not provide access to the visual content, such as photographs or images, that accompanies the text in articles. This lack of access posed a challenge to conducting a comprehensive visual analysis. To address this limitation, the lead author manually searched for the corresponding online articles using Google, entering headlines and partial content to locate the original sources. This method allowed for 85% of the articles (n = 179) to be replicated online. Of these, 125 articles (70% of the sample) contained photography relevant to panic-buying. While this approach was effective for a majority of the sample, it’s important to note that the inability to retrieve the full visual content directly through Nexis may have limited the comprehensiveness of the visual data collection.

Development of the coding framework

A coding framework was developed to allow for the systematic recording of data and imagery within the articles (S2 File), closely following the frameworks developed by Buckton et al. [81] and Wells and Caraher [65]. Manifest data in this study included the newspaper title, publication date, stakeholder voices, headline and image category. This allowed for the development of temporal and reporting trends analysis.

Analytical framework overview

To provide clarity and transparency, the research methodology was organised into three key stages, illustrated in Table 1. This tiered approach integrated content and visual analysis, systematically coding manifest and latent data, and linked these findings to the research objectives.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Analytical framework overview.

https://doi.org/10.1371/journal.pone.0315142.t001

The lead author also tracked sentiment and sensationalism using the framework in Table 2. This analysis of ‘loaded words’ [87] provides an understanding of how language and texts influence and sustain social meanings and knowledge structures [88].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Framework for defining sentiment and sensationalism.

https://doi.org/10.1371/journal.pone.0315142.t002

For the visual content analysis, the author reviewed the body of images and coded these into five categories. These categories included, for example: images of empty shelves (which signal scarcity or panic), images of long queues (which suggest public anxiety) or photographs of stockpiled goods (which visually represent the behaviours being reported on). The systematic coding of images allowed the research to explore how visual cues may have been employed to influence public perception. For example, repeated imagery of empty shelves or crowded supermarkets created scarcity cues that may have influenced readers’ decisions to panic-buy. This aligns with the study’s core aim to understand how media visuals contributed to agenda-setting and the public’s perception of food access during the crisis.

For the latent content, the author drew on Entman’s [66] definition of framing, where narratives are constructed to promote the interpretation of an issue in a particular way. The articles were examined in detail to identify frames and agendas. The coding process revealed six frames: prominence of supermarkets; food supply/access; food policy; individual behaviour; socio-economic impacts of panic-buying and panic-buying drivers. Outside of these, certain tensions and contradictions emerged, such as media outlets advising customers to stockpile [59] and subsequently criticising those who did [89], these were also logged onto the coding frame.

The lead author tested the coding framework on 15 articles to validate the coding categories. In the final coding phase, the lead author carefully read each article, systematically recording both manifest and latent data in Excel. For each thematic code, the author noted whether the article explicitly contained content related to that theme. For example, The Sun, 4.03.2020, which specified ‘supermarkets have plans to ration food by working with suppliers to cut back on variety—focusing instead on their supplies of staple products,’ was coded as mentioning food supply, supermarkets and supermarket food policy. Both quantitative and qualitative research methods were employed, with statistical analysis performed manually in Excel. This dual approach—quantitatively measuring reporting trends alongside qualitative exploration of narratives, themes, and frames—provided a comprehensive interpretation of agenda-setting during Covid.

Reporting trends

The database search returned n = 788 articles. Following the inclusion/exclusion criteria the manifest data shows a total of n = 209 relevant articles were printed between March 1 –July 31, 2020. Table 3 shows the breakdown of articles across newspaper titles with their genre and political stance. The popular press comprised over half the articles and coverage was led by left-wing press (46%). The Daily Mirror reported most frequently in this period, with 35% of articles, followed by The Daily Mail (19%) and The Sun (19%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Article breakdown across publications.

https://doi.org/10.1371/journal.pone.0315142.t003

Although the study cannot definitively explain why The Daily Mirror led in coverage, existing literature provides some hypotheses. Tabloids like The Daily Mirror focus on sensational, emotionally charged stories that resonate with their audience [90]. Panic-buying, a significant public concern during the pandemic, aligns with this focus. The Daily Mirror’s readership, primarily C2DE social groups [82, 85], may have been more impacted by shortages, prompting the newspaper to prioritise this issue to reflect their concerns. Additionally, as a left-leaning paper, The Daily Mirror often critiques Conservative policies [91], and panic-buying provided a lens for highlighting governmental shortcomings. These factors may have contributed to its extensive coverage, though these remain informed hypotheses.

96% of the articles were published in March/April; March accounted for 82%. Given the concentration in March/April, Fig 2 shows the reporting peaks for this period only and identifies political, news, or global announcements. Between March 18–21, coverage peaked at 62 articles. This intense focus could reflect intra-media agenda setting, where media outlets not only report on the same issues but influence each other’s focus and framing of topics, leading to a consistent narrative across various sources. These articles focused on supermarket rationing policies and human-interest stories about the elderly or vulnerable. A social media video by nurse Dawn Bilbrough [92] pleading with shoppers to stop panic-buying may have driven coverage, alongside announcements by government figures Sir Patrick Vallance (former Government Chief Scientific Advisor), Matt Hancock (former Health Secretary) and George Eustace (former Secretary of State for Department for Environment, Food and Rural Affairs).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Number of articles in March and April, annotated with potentially triggering events.

https://doi.org/10.1371/journal.pone.0315142.g002

Imagery trends

There were certain images which became symbols of the pandemic, such as empty shelves and queues. Analysis in Fig 3 shows empty shelves were in 64% of the sample and queues in 40%. Online shopping (19%) varied from images of unavailable delivery slots on supermarket websites to images of social media users complaining about a lack of slots. Newspapers often used the public’s social media posts of empty shelves to illustrate specific food shortages. 91% of empty shelf imagery was shown by the popular and mid-market genres–there was a notable gap in empty shelf imagery in the quality press. Wright argues that visual imagery conveys information and can increase media impact [93]. The frequent use of empty shelf imagery is an example of how visuals can reinforce the public’s sense of urgency, potentially driving consumer behaviour. This supports the agenda-setting theory by showing how media can direct attention to specific issues.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Percentage of articles in sample which showed image type.

https://doi.org/10.1371/journal.pone.0315142.g003

Sentiment analysis and sensationalism

Sentiment and sensationalism were coded according to Table 4. Sentiment analysis showed most articles were negative (68%). Negative articles were driven by four main areas: panic-buyers described as self-serving; the chaotic situation; impacts on the vulnerable and inadequate government response. Positive stories (2%) recounted supermarket efforts to help the elderly or consumer confessions about not panic-buying. 20% combined both negative and positive slants and 9% were neutral.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Examples of sensationalised language.

https://doi.org/10.1371/journal.pone.0315142.t004

Sensationalised language may imply impending threat, dramatise situations or vilify people. Table 4 gives an indication of the sensationalist language used in headlines. This provides insights into how language is used to produce meaning [88]. There was a high degree of sensationalism across the articles [89%], evenly spread across all genres. Quality press, such as The Independent, had sensationalism scores similar to popular titles, such as The Daily Mirror (84% and 88%, respectively). This could be indicative of ‘tabloidization’, where mid-market and quality newspapers compromise on quality output to increase readership [94]. The high level of sensationalism found in this study aligns with previous research on how emotionally charged reporting may have affected public panic and stockpiling behaviours, further contributing to the media’s agenda-setting role during crises [21, 26, 37–39, 53, 57, 58, 60].

Stakeholder voices

The author analysed stakeholder voices across industry (supermarkets and retail organisations), political (government representatives or organisations), experts (food systems and logistics specialists), and consumers. The most frequently cited source was industry, appearing in 62% of the articles. Supermarkets aimed to reassure the public about the absence of food shortages and discouraged stockpiling.

Supermarkets often dominated the headlines, oscillating between arbiters of shopping rules (’VERY LITTLE HELPS STORES LIMIT STOCKS AMID BUG FEARS–Tesco rations pasta & veg to curb panic buys’—The Sun, 9.3.2020), and heroes who kept the nation fed (‘UK supermarkets draw up plan to ’feed the nation’ as coronavirus spreads’ -The Guardian, 2.03.2020). A joint letter from supermarkets via the British Retail Consortium [5] urging shoppers to be considerate also dominated headlines, which may suggest that retailers, rather than the government, were leading the response to panic-buying.

Political voices appeared in 27% of articles, with government sources stating that panic-buying was unnecessary and supply chains were resilient. Consumer voices featured in over a third of articles, with detailed accounts of panic-buying behaviours by individuals (‘filled his boot with 15 kg of penne, 48 bags of crisps, 16 tins of beans, and litres of Dettol,’–The Daily Mail, 2.03.2020) and distinctions between panic-buying and cautious preparation (’It’s not panic-buying, I think people are just being cautious and buying stuff for storage,’–The Guardian, 2.03.2020). Evidence suggests that social media was a significant platform for sharing panic-buying behaviours [45], and this was reflected in this study, with many consumer voices presented through social media extracts and reprints.

Experts appeared in 30% of articles, often quoting Bruno Monteyne, a City of London analyst and former Tesco supply chain director. His quotes were used to create sensational headlines, such as ’Coronavirus: Supermarkets preparing for ’food riots’ as panic-buying Brits strip shelves’ (The Daily Mirror, 6.03.2020).

The prominence of supermarkets as the primary source of reassurance suggests that media narratives framed them as central to resolving the crisis, highlighting the media’s role in shaping public understanding of key institutions during the pandemic.

Thematic analysis

This section describes the main themes and sub-themes in newspaper coverage, analysed through the lens of agenda-setting theory [75], which posits that media story selection shapes the public agenda [95]. It examines how media frames the news agenda through specific narratives [65, 77]. It also offers analysis of the visual imagery accompanying reporting.

Overall findings

Six main themes were identified in the newspaper coverage. Supermarkets were mentioned in 88% of articles, followed by food supply/access (79%), food policy (74%), individual behaviour (72%), socio-economic impacts of panic-buying (70%), and panic-buying drivers (41%). Table 5 outlines these themes and sub-themes, with example headlines illustrating their presentation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Panic-buying main themes and sub-themes.

https://doi.org/10.1371/journal.pone.0315142.t005

Sub-themes

Table 6 shows which sub-themes were mentioned most. Supermarkets, for example, were mentioned in 88% of articles and 67% of articles mentioned supermarkets’ food policy.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Percentage of articles that include the most mentioned sub-theme within each main theme.

https://doi.org/10.1371/journal.pone.0315142.t006

Fig 4 shows the breakdown of all sub-theme mentions.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Percentage of articles which mention sub-theme.

https://doi.org/10.1371/journal.pone.0315142.g004

Visual imagery

64% of the sample included images of empty shelves, queues, and unavailable online delivery slots, creating a frightening perception of food shortages and increased demand. Emotive photos, particularly of the elderly in front of empty shelves, frequently appeared in newspapers and became prominent in the news agenda. For example:

’ (Elderly man staring at empty Sainsbury’s shelves shows panic-buying impact)’

(The Daily Mirror, 18.03.2020).

Additionally, social media posts detailing specific locations with empty shelves and showing images of food in people’s boots or trolleys vividly illustrated the panic-buying phenomenon. The visual imagery acts as a concrete manifestation of the reported panic-buying. These may have reinforced public perceptions of scarcity, illustrating how visual framing by the media may have intensified consumer anxiety and driven behaviour, in line with agenda-setting theory.

Tensions and contradictions

There were several tensions and contradictions within 60% of articles, which sat outside the themes but provided important insights. These were i) emotional pleas to cease panic-buying and encouragement of altruism (46%), ii) mentions of ‘no shortage of food’ and adequate supply (27%), iii) advising/discussing what consumers should stockpile (7%), and iv) mention of contradictory advice by the government (7%).

The plea for consumers to stop panic-buying (46%) came from multiple sources: retailers, government, newspapers and consumers. This tension firmly portrayed the panic-buying crisis as a fault of individuals and not a breakdown of the food system. 27% of the articles mentioned that there was no shortage of food or an adequate supply. These phrases are often attributed to retail or government sources. Images of empty shelves ran alongside articles which said ‘adequate supply’, in apparent contradiction.

This confusion may have been compounded by contradictory government advice (7% of articles). For example, newspapers reported on Public Health England’s advice to stock up on two weeks’ food supply in case of isolation, which sent ‘mixed messages’ (The Daily Mail, 2.03.2020). Furthermore, the Prime Minister advised people to visit the supermarket in person to make slots available for the vulnerable–which supermarkets disagreed with.

(…) The Prime Minister urged everyone to use food delivery services wherever possible, but the reality is that current demand vastly exceeds supply.

(The Independent, 17.03.2020)

The final contradiction relates to coverage in the first two weeks of March. There was an explicit paradox where newspapers either asked if people were stockpiling or advised them what to stockpile (in 7% of articles). The Daily Mail printed a ‘What should you stockpile in a pandemic’ advice guide, which listed cereals, grains, beans, tinned food and soft drinks (The Daily Mail, 13.03.2020). Similarly, The Sun printed a food stockpile checklist of ‘things you might want to include in your coronavirus stockpile’. This included items such as pasta (‘high in carbs and stores well’), rice (‘can be used in loads of different meals’) and baking goods (‘to make bread’), (The Sun, 10.03.2020). In subsequent weeks, the newspapers would then vilify those who stockpiled, despite having advised them to do so. The products which the newspapers advised people to buy are also those which were in high demand during the pandemic.

The media’s contradictory messaging—encouraging stockpiling while vilifying panic-buyers—highlights how inconsistent framing can create confusion and social tension, further emphasising the media’s power in shaping public understanding.

Reporting trends

82% of articles analysed were printed in March 2020, peaking between March 18–21. This coincided with a surge in consumer spending, with supermarkets experiencing their highest sales in the week before the first lockdown on March 23 [3]. Coverage spikes were linked to government announcements, a British Retail Consortium’s open letter, and human-interest stories like nurse Dawn Bilbrough’s plea against panic-buying. These peaks highlight how quickly news can steer media and public agendas, which is valuable information for policymakers. Uysal’s study [43], also showed a correlation between media exposure and panic-buying, suggesting that understanding the types of content driving coverage can aid government communication planning. This research reinforces the need for policymakers to engage proactively with media outlets to promote preparedness rather than panic. Such efforts have the potential to foster food system resilience in times of crisis.

Sensationalism and visual imagery in media coverage

While sensationalism is common in crisis reporting, the level observed in this study is particularly significant. 89% of the analysed articles used sensationalised language with terms like ’waves,’ ’panic,’ ’empty shelves,’ and ’riots.’ This aligns with Rajkumar’s [38] and Phillips et al.’s findings [39], which noted high levels of sensationalism in media. Sensationalism, coupled with visual cues like empty shelves (64% of articles) and queues (40%), and may have heightened public anxiety and stimulated panic-buying behaviours [36, 57, 60].

Arafat et al.’s study found that 67.3% of articles included empty shelves and Rajkumar found that 65% used imagery encouraging panic-buying [37, 38]. Notably, images of empty shelves often accompanied articles stating there was no food shortage, sending mixed messages. Consumers shared their panic-buying experiences in supermarkets, often with social media excerpts showing empty shelves, reinforcing perceptions of food shortages and encouraging panic-buying. Social media and print media portrayed food scarcity, with social media playing a key role in spreading panic-buying information [21, 45]. Combining sensational text and imagery may have compounded public uncertainty, further emphasising the need for responsible media reporting. Policymakers could consider introducing media guidelines to limit sensationalism and encourage accurate portrayals of food availability [60], which would help mitigate unnecessary public anxiety and panic-buying [36, 60].

Supermarkets’ dominance in coverage

Supermarkets appeared in 62% of articles, a finding consistent with Phillips et al. [39]. This finding supports Dixon and Banwell’s theory of ’supermarketisation’, where supermarkets hold significant power in the food system [96]. Supermarkets were framed mostly positively in 88% of articles, portrayed as ’feeding the nation’, while negativity was often directed at individuals instead. Supermarket policies—such as rationing, streamlining product ranges, and distinct operating hours—were highlighted more than government policies (67% vs. 20%), with the government criticised for leaving the panic-buying response to retailers [11, 14].

Given the supermarkets’ dominant role in media coverage and public perceptions during the pandemic, policymakers could strengthen collaboration frameworks with major retailers. This would enable consistent communication strategies to manage public expectations and avoid consumer panic in future crises.

Themes and framing

Coverage predominantly blamed individuals for panic-buying, portraying them as immoral and supermarkets as heroes. Phillips et al. [39] noted that communities often blame specific groups while portraying others as victims or champions. This aligns with other studies on panic-buying media content [37, 39, 60]. In 46% of articles, the government, retailers, and consumers urged the public to stop panic-buying, blaming individuals for its impact on vulnerable populations. Common descriptors for stockpilers included ‘selfish,’ ‘waves,’ ‘hoards,’ and ‘locusts.’ Arafat and Kar support naming and shaming in their proposed media guidelines for panic-buying [97]. However, the authors argue that focusing on individuals diverts attention from the systemic nature of panic-buying, which requires active government involvement. Policymakers could use these insights to develop future public messaging, emphasising the need for systemic responses rather than individual blame.