Navigating the ‘infodemic’: how people in six countries access and rate news and information about coronavirus

Infodemic

A man reads a newspaper during the coronavirus outbreak in Madrid. REUTERS/Susana Vera

Professor Rasmus Kleis Nielsen
Dr Richard Fletcher
Nic Newman
Dr. J. Scott Brennen
Dr Philip N. Howard

Executive summary | 1. Introduction | 2. Methodology | 3. Sources of information | 4. The role of platforms| 5. Trust in different sources and platforms. | 6. Concerns over misinformation | 7. Helping people understand the coronavirus crisis | 8. What do people know about coronavirus and how do they respond? | 9. Conclusion. | Footnotes | References | About the Authors | Acknowledgements

Download a PDF version here / Lee en español aquí 

Executive summary

In this report, we use survey data collected in late March and early April 2020 to document and understand how people in six countries (Argentina, Germany, South Korea, Spain, the UK, and the US) accessed news and information about COVID-19 in the early stages of the global pandemic, how they rate the trustworthiness of the different sources and platforms they rely on, how much misinformation they say they encounter, and their knowledge of and responses to the coronavirus crisis.

We show that:

  • News use is up across all six countries, and most people in most countries are using either social media, search engines, video sites, and messaging applications (or combinations of these) to get news and information about coronavirus.
  • In all six countries, people with low levels of formal education are much less likely to say that they rely on news organisations for news and information about coronavirus, and more likely to rely on social media and messaging applications. In Argentina, South Korea, Spain, and the US, young people are much more likely to rely on social media, and in Germany, the UK, and the US, to rely on messaging applications groups.
  • In every country covered, very high numbers of people across age groups, levels of education, and political views rate scientists, doctors, and other health experts as trustworthy sources of information about coronavirus. Three-quarters of respondents trust national or international public health organisations, a majority of respondents rate news organisations relatively trustworthy, and in every country apart from Spain and the United States a majority rates their national government trustworthy as well.
  • While levels of trust in scientists and experts are consistently high, and levels of trust in ordinary people are consistently more limited, there are significant political differences in trust in news organisations and in the government, especially in the United States, where people on the left of the political spectrum trust news organisations much more than they trust the government, and people on the right trust the government much more than they trust news organisations. 
  • When asked how trustworthy they find news and information about coronavirus from different platforms, most respondents rate platforms less trustworthy than experts, health authorities, and news organisations. Results vary significantly across different types of platforms – averaged across the six countries, the 'trust gap' between information from news organisations and information from social media is 33 percentage points, between news and video sites 30 percentage points, and between news and messaging applications 35 percentage points. The gap is 14 percentage points on average between news and search engines.
  • Asked how much false or misleading information about coronavirus (COVID-19), if any, people think they have seen from different sources and platforms, four overall findings stand out.
    • First, for every source and every platform in every country covered, it is a minority who say they have come across a lot or a great deal of false or misleading information around coronavirus.    
    • Second, among sources, 'bottom-up' false or misleading misinformation spread by ordinary people whom respondents do not know personally is most widely identified (though in South Korea, Spain, and the US respondents say individual politicians generate large volumes of 'top-down' misinformation). On average about a third say they have seen a lot or a great deal of false or misleading bottom-up misinformation in the last week.
    • Third, among platforms, concern is focused on social media and messaging applications, where on average about a third of respondents say they have seen a lot or a great deal of false or misleading information in the last week.
    • Fourth, while concern about false or misleading information about coronavirus from news organisations and national government is less widespread than concerns over ordinary people, social media, messaging applications, and in some countries individual politicians, a significant majority are still worried – about a quarter on average for both news and government.
  • A majority of respondents in every country say that the news media have helped them understand the crisis and explain what they can do. However, about one in three also say they feel the news media have exaggerated the pandemic.
  • In terms of how well people do when asked a series of factual questions about coronavirus, most people do relatively well, with a clear majority answering more than half the questions correctly (more than three-quarters in every country apart from South Korea (58%) and the US (65%)). Regression analysis shows that using news organisations as a source of information is associated with a statistically significant increase in coronavirus knowledge in every country except Argentina and Spain. No source of, or platform for, information in our dataset is consistently and significantly associated with lower coronavirus knowledge.
  • For most individual questions, the number of correct answers is associated with level of education, as respondents with low levels of formal education give more incorrect responses to most questions, and for some questions, in particular on issues that high-profile politicians and other prominent public figures have opined on, political orientation plays a large role. Almost a quarter of our respondents incorrectly believe coronavirus was made in a laboratory.


1. Introduction

With the arrival and spread of COVID–19, 'we're not just fighting an epidemic; we're fighting an infodemic', WHO Director-General Tedros Adhanom Ghebreyesus said 15 February.1 In this situation, it is critically important that people have access to news and information that they trust and that can help them understand the coronavirus crisis, what they can do to protect themselves, those they care about, and their wider communities, as well as independent reporting on how governments and others are responding to the pandemic. It is clear that both information and various kinds of misinformation will crucially shape how people understand and respond to the public health crisis, and how they evaluate which institutions are helping address it (and which ones not). As researchers have long known, it is perceptions of risk, not actual risk, that determine how people respond to crises (Glik 2007).

The vast volume of news and information around COVID-19 – and the ambiguity, uncertainty, and sometimes low-quality, misleading character, or outright false nature of some of it – is captured by the WHO’s use of the term 'infodemic'.2 But it is important to recognise up front that while the coronavirus can for practical purposes be identified by medical professionals as a single pathogen, diagnosed and tested for, and its spread thus mapped (provided testing and other means are in place), beyond the most demonstrably false, obviously fabricated, and clearly deliberately manipulated disinformation identified by fact-checkers, most of what we learn about the coronavirus is harder to clearly and cleanly separate into information and misinformation, true and false, reliable and unreliable (Brennen et al. 2020). This is in line with previous research demonstrating that most of the public – arguably rightly – see much of the information they rely on, and many of the sources, whether news media, technology companies, or public authorities, in shades of grey, not black and white (Graves and Nielsen 2017; Newman et al. 2018). Often, subjective and intersubjective factors like trust in sources matter as least as much for how the public gets informed about risk as harder to establish factors like reliability and veracity (Glik 2007).

In this report, we use survey data collected in late March and early April 2020 to understand how people in six countries access news and information about COVID-19, how they rate the trustworthiness of the different sources and platforms they rely on, how much misinformation they say they come across from different sources and on different platforms, and what they themselves know about – and do about – the coronavirus crisis. We also break top-line findings down to identify differences by age, education, and political orientation, and use regression analysis to identify correlations between specific forms of news and information use and what people know about the virus, controlling for other factors like age, education, and political orientation.3 Throughout the report, we will contextualise the current findings in light of what we already know about broader differences and similarities in news and media use in these countries through our annual Reuters Institute Digital News Report (Newman et al. 2019). We hope the analysis will be useful to journalists, news media, platform companies, public authorities, and citizens as they think about the flow of news and information around coronavirus.

The report covers Argentina, Germany, South Korea, Spain, the UK, and the US, six countries with a combined population of more than 600 million and representing different media systems and different political systems. The coronavirus pandemic is also at different stages in each country, and the governments of these countries have implemented different countermeasures. As of 31 March, when our survey went into the field, Our World in Data reported that Argentina had seen 0.53 deaths per million, Germany 6.96, South Korea 3.18, Spain 156.99, the UK 20.74, and the US 9.58.4 As of 7 April, when the last survey closed, the death rates had jumped to 1.17 in Argentina, 19.18 in Germany, 3.74 in South Korea, 279.22 in Spain, 79.15 in the UK, and 33.20 in the US. Spain is by far the hardest hit of the countries covered, Argentina the least hard hit, and South Korea has suffered from the epidemic longer – and, like all other countries, faces the possibility of a ‘second wave’ of cases.


2. Methodology

The report is based on a survey commissioned by the Reuters Institute for the Study of Journalism and the Misinformation, Science and Media project run with the Oxford Internet Institute and supported by the Oxford Martin School. Our purpose is to understand how people access and rate news and information about COVID-19 from different sources. Research was conducted by YouGov using an online questionnaire fielded from 31 March to 7 April 2020 across Argentina, Germany, South Korea, Spain, the UK, and the US.

Samples in each country were assembled using nationally representative quotas for age, gender, and region. The data were also weighted to targets based on census/industry accepted data for the same variables.

Table 1. Countries and sample size

We should note that online samples will tend to under-represent the consumption habits of people who are not online (typically older, less affluent, and with limited formal education). According to the ITU, internet penetration in the six countries range from 93% (Argentina) to 96% (United States).

It is also important to note that online surveys rely on recall, which is often imperfect or subject to biases. Furthermore, questions around misinformation provide only information on people’s perception of the problem, not an objective measure of how much false information they have (perhaps unwittingly) engaged with. We have tried to mitigate these risks through careful questionnaire design and testing.


3. Sources of information

In every country surveyed, we have seen a surge in news consumption compared to what we have seen in the past (e.g. Newman et al. 2019). Many news organisations are working hard to serve growing audiences as well as possible during the crisis, complementing traditional reporting often done under extremely challenging conditions due to lockdowns and social distancing measures (as well as financial pressures on the business of news during the crisis) with many impressive examples of editorial innovation including increased use of data visualisation, interactives, and various forms of simulation. Television and online are the most popular way of getting news in all six countries (see Figure 1). The figures for newspapers are lower than normal, as countries have entered lockdown, complicating print distribution and greatly reducing single copy sales. Though news use overall is up, demographic differences are as expected – younger people overwhelmingly identify online media (and often social media) as their main source of news, and older people broadcast. The coronavirus crisis has not changed the structural shift to digital or changed the demographic profile of different sources of news.

Figure 1. Proportion that used each as a way of getting news in the last week

Q4. Which, if any, of the following have you used in the last week as a source of news? Base: Total sample in each country.

While news use is up and news organisations remain among the most important sources of information, in a public health crisis where most people are online and many different organisations, including public authorities, have websites, social media accounts, and other channels available, news media are not the only sources people rely on. For some, experts like Dr Christian Drosten, chief virologist at the Charité university research hospital in Berlin, and Dr Fernando Simón, the director of Spain’s health emergency centre, have become household names, widely featured in the news and on social media.5

Information-seeking habits are not exclusively centred on news organisations. When asked what sources they have used in the last week for news or information about coronavirus specifically, for example, by going directly to websites or apps, via various platforms, or in person, a complex picture emerges. (See Figure 2.)

Figure 2. Proportion that used each for coronavirus news in the last week

Q7. Which, if any, of the following have you used in the last week as a source of news or information about coronavirus (COVID-19)? Base: Total sample in each country.

Across the six countries we cover, about two-thirds say they have relied on news organisations, ranging from a low 47% in Germany to a high 77% in South Korea. It is remarkable that, in an age of abundant and easily accessible news, and where accurate, credible, reliable information arguably is incredibly important, a large part of the public does not see news organisations themselves as a source of news or information about the crisis. (And news use, even in the midst of the global pandemic, is far from exclusively focused on coronavirus.)6

While news organisations are the single most widely identified source of news and information about coronavirus, many say they do not rely on them, and, collectively, other sources, including various government, health authorities, and expert sources, are as widely relied on as news – and a significant minority say they have relied on ordinary people like themselves. Of course, some of the government, health authority, and expert sources people say they have accessed are almost certainly at least in part accessed via news media, though they can also be accessed via official websites, social media channels, and via advertising and public information campaigns. But it seems likely from the data that for many people, when information from government and experts is featured on the news, it is not necessarily seen as coming from news organisations themselves.

In all the six countries covered, people with low levels of formal education are much less likely to say that they rely on news organisations for news and information about coronavirus, and in all but Germany and the UK, young people are much less likely to say they rely on news organisations.

In most countries, people with higher levels of formal education are also much more likely to say they rely on information from health authorities and various experts. Conversely, among people with low education in the US, more say they rely on ordinary people like themselves than rely on news organisations or the national government. 

With public health experts like Dr Trudie Lang, Professor of Global Health Research at the University of Oxford, arguing that for a public health message to be successful it has to reach 80% of the population, it is clear that no single source or platform can deliver anywhere near that reach alone.7


4. The role of platforms.

Many of the sources people rely on are accessed via the products and services offered by various platform companies, increasingly widely used as sources of news and information as well as for many other purposes (Newman et al. 2019). Just as news organisations and governments have invested heavily in helping people understand coronavirus, Facebook has introduced a 'COVID-19 Information Center' on its main social network; Google Search provides an SOS Alert, new knowledge panels, and an information and resources centre for the virus; YouTube features information from health authorities at least in some countries; and Twitter presents Coronavirus Tweets from news media and from authorities. Many of the major platform companies have also offered governments and health authorities advertising credits to spread public health messages.

While only a small minority of our respondents identify social media as their main source of news (about one in seven on average, 16%), it is clear that various platforms play a key role in how people access and find news and information about the coronavirus. (See Table 2.)

Table 2. Proportion that used selected popular search, social media, and video platforms as ways of getting news about coronavirus in the last week

Q8. Which, if any, of the following have you used in the last week as a source of news or information about coronavirus (COVID-19)? Base: Total sample in each country

Whereas news organisations are less widely used by many young people and many with lower levels of education, social media are more widely used in these groups. Search is also widely used, and in most countries, young people rely more on search than the rest of the population. 

The mix of networks in each country is broadly in line with what we see in our 2019 Digital News Report but there are some interesting differences in general and with particular groups (see Figure 3). Younger people rely much more on newer, more visual networks like Instagram and Snapchat for information about COVID-19. Even TikTok, popular with teenagers, has been used to pass on public health messages such as the Vietnamese song and dance about hand washing which went viral.

Use of Instagram has also increased substantially on our 2019 data – and particularly with younger groups. In Argentina almost half of 18–24s (49%) said they used Instagram to access COVID-19 content. Celebrities and influencers play an outsized role on these networks with some sharing music, running exercise classes as well as commenting on the wider health issues (for good and sometimes ill, as shown by Brennen et al. 2020). Following cases of misinformation on the platform, Instagram has reserved its explore tab only for information from credible health organisations.

Figure 3. Proportion of 18-24s that used each for coronavirus news in the last week

Q8. Which, if any, of the following have you used in the last week as a source of news or information about coronavirus (COVID-19)? Base: 18-24s in each country: UK = 202, USA = 153, Germany = 183, Spain = 80, South Korea = 83, Argentina = 151.

Messaging applications focused on one-to-one communication and smaller, private groups (also found on some social media platforms including Facebook) add further complexity to the situation. WhatsApp has introduced a WHO chatbot and limited users’ ability to forward messages, among other steps, but there are still plenty of examples of misinformation circulating on messaging applications both between individual users and in larger groups (Lytvynenko 2020). Because these environments are designed to allow private communication among individuals and those who participate in specific groups, and because these platforms enable a very large volume of communications, some of it demonstrably false and misleading, there has been a lot of focus on their role, which, due to the dearth of data access and in some cases encryption, is hard for researchers to analyse. 

Survey data provide one way of getting at least a basic understanding of the likely role of these more private forms of communication. About four in ten of our respondents say they have discussed coronavirus in at least one kind of Facebook group, and more than half have discussed COVID-19 in WhatsApp groups (in South Korea in various kinds of Kakao groups). (See Figure 4.) 

In line with what we also see for social media use, messaging groups are much more widely used by young people and people with low levels of formal education. The numbers of people who say they have used private social network groups or messaging groups to discuss COVID-19 are much higher than the percentage who in 2019 said they use these platforms for news (Newman et al. 2019). Take the UK as an example – 31% of our respondents say they have discussed coronavirus in one or more Facebook groups in the last week, compared to just 8% who in 2019 said they discussed news or politics in such groups, and 44% say they have discussed coronavirus in one or more WhatsApp groups in the last week, compared to just 2% who in 2019 said they had discussed news or politics.

Figure 4. Proportion that have discussed coronavirus in different types of group

Q9. Facebook and WhatsApp allow you to set up, join and participate in groups, where you can discuss news or related topics with likeminded people. Which, if any, of the following have you used in the last week to discuss coronavirus (COVID-19)? Base: Total sample in each country

It is potentially important to note here that a clear majority of those who say they have discussed coronavirus via groups on social media or messaging apps specifically name groups with family or friends and colleagues, not groups with strangers. It seems to be the case that many groups originally set up for other purposes now also, at least in part, revolve around discussing COVID-19, underlining how coronavirus conversations are spilling into a wide range of different settings both offline and online.


5. Trust in different sources and platforms.

It is also important to know what sources and platforms people trust, as they will be more likely to form their opinions and behaviour on the basis of information from these (Spiegelhalter 2017). (We know from years of research that people often happily use media that they do not particularly trust, e.g. popular newspapers and some social media platforms, if they find them worthwhile for other reasons, see e.g. Newman et al. 2019.)

Looking at how much people trust different sources first, in every country covered a majority rate news organisations as relatively trustworthy, and in every country apart from Spain and the United States, a majority rate their national government as relatively trustworthy as well. (While not directly comparable, it is worth remembering that, in all six countries, it was only a minority who said before the crisis that they trusted most news most of the time, see Newman et al. 2019.) In every country, health authorities and expert sources are very highly trusted by almost everyone, whereas individual politicians and most ordinary people face greater scepticism. (See Figure 5.)

Figure 5. Proportion that trust each for coronavirus news

Q10. How trustworthy would you say news and information about coronavirus (COVID-19) from the following is? Please use the scale below, where 0 is ‘not at all trustworthy’ and 10 is ‘completely trustworthy’. Base: Total sample in each country. Note: Trusts = 6-10; Neither = 5; Does not trust = 0-4.

There are some socio-economic differences here, with respondents with lower levels of formal education in several countries rating news organisations, and some public authorities and experts, less trustworthy, but holding broadly the same view of ordinary people as everybody else. In most countries, people politically opposed to the government in power (e.g. people on the right in Spain facing a left-wing government, people on the left in the UK facing a right-wing government) rate the national government as much less trustworthy than those who are politically sympathetic to the ruling parties. 

In some countries, the differences between people with different political viewpoints are more pronounced. The United States is the most politically polarised country included, Argentina the least polarised, and Germany exhibits a more asymmetric polarisation.

In the highly (and relatively symmetrically) polarised United States, respondents across the political spectrum have broadly the same view on the (high) trustworthiness of scientists, doctors, and other health experts, and the (more limited) trustworthiness of ordinary people they know personally. But perceptions of news organisations and the national government are strongly aligned with political viewpoints. (See Figure 6.) A clear majority on the left find news organisations trustworthy, whereas more respondents on the left find people they know trustworthy than find the (right-wing) national government trustworthy. Conversely, on the right, a clear majority find the (right-wing) national government trustworthy, but more respondents on the right find people they know trustworthy than find news organisations trustworthy. 

In Argentina, these differences are much less pronounced (though people on the right are sceptical of the left-wing government), and in the more asymmetrically polarised Germany, those on the political right have less trust in both the news media and in the national (Conservative-led) government than people on the centre and on the left.

Figure 6. Proportion that trusts each for coronavirus news in the US by political leaning

Q10. How trustworthy would you say news and information about coronavirus (COVID-19) from each of the following is? Please use the scale below, where 0 is ‘not at all trustworthy’ and 10 is ‘completely trustworthy’. Base: Left/centre/right in the US: Left = 293, Centre = 490, Right = 309. Note: Trusts = 6-10.

Overall, when asked how trustworthy they find news and information about coronavirus from different platforms, most respondents rate platforms less trustworthy than news organisations, health authorities, and various experts. (See Figure 7.) Results vary significantly across different types of platforms. 

In every country, many more people find news and information on coronavirus from news organisations trustworthy than say the same about things they find on social media. Across the six countries, the 'trust gap' between news organisations and social media is, on average, 33 percentage points. The results are broadly similar for video sites and messaging applications. In every country covered, more people find news and information on coronavirus from news organisations trustworthy than say the same about content from video sites. On average, the trust gap is 30 percentage points. Similarly, for news and information accessed via messaging applications, the  trust gap is on average 35 percentage points. 

For search engines, the results are comparable, but the differences much less pronounced. In Argentina, there is no significant difference between search and news organisations, and averaged across all six countries, the trust gap is 14 percentage points. This is in line with previous research suggesting that while in many countries fewer people say they trust the news they find via search than say they trust the news media in general, trust in news in search is significantly higher than trust in news found via social media (Newman et al. 2019).

As Figure 7 shows, these patterns vary somewhat from country to country.

Figure 7. Proportion that trust each as a way of getting news about coronavirus

Q11. How trustworthy would you say news and information about coronavirus (COVID-19) on each of the following is? Please use the scale below, where 0 is 'not at all trustworthy' and 10 is 'completely trustworthy'. Base: Total sample in each country. Note: Trusts = 6-10; Neither = 5; Does not trust = 0-4.


6. Concerns over misinformation.

As Sylvie Briand, director of Infectious Hazards Management at WHO's Health Emergencies Programme and architect of WHO's strategy to counter the infodemic risk, has told The Lancet, 'We know that every outbreak will be accompanied by a kind of tsunami of information, but also within this information you always have misinformation, rumours, etc.'8

Our results demonstrate that much of the public is keenly aware of this. In addition to countless examples of various kinds of misinformation circulating online and via various platforms, incidents unfortunately also include some news organisations publishing misleading information about the virus, and many examples of high-level politicians, celebrities, and other prominent public figures pushing false or misleading information via press conferences, official channels, and social media (Brennen et al. 2020).

Given these problems and previous research showing that many people think of what they see as poor journalism and hyperpartisan political propaganda as parts of wider misinformation problems (Graves and Nielsen 2017), and that people are often very worried about the authenticity and veracity of much of the information they come across online (Newman et al. 2019), we have sought to document how much misinformation people say they come across from different sources and across different platforms. 

It is important to reiterate here that our survey data only capture people’s perceptions of these problems, and do not provide hard evidence of the actual veracity or reliability of the information people come across. Nonetheless, as with many parts of the misinformation problems plaguing digital media and our societies more broadly, perception is part of reality. If people see something as misinformation, even if it is true, they react to it differently. If they see something as reliable, even if it isn’t, they may well act on it.

Asked how much false or misleading information about coronavirus (COVID-19), if any, people think they have seen from different sources and platforms, four overall findings stand out.

  1. For every source and every platform in every country covered, it is a minority who say they have come across a lot or a great deal of false or misleading information. 
  2. Among sources, bottom-up misinformation from ordinary people respondents do not know personally is most widely identified as a source of false or misleading information (on average by about a third). (In South Korea, Spain, and the US respondents say individual politicians generate large volumes of top-down misinformation.)
  3. Among platforms, concern is focused on social media and messaging applications, where on average about a third of respondents say they have seen a lot or a great deal of false or misleading information in the last week. 
  4. While concern about false or misleading information about coronavirus from news organisations and national government is less widespread than concerns over ordinary people, social media, messaging applications, and in some countries individual politicians, a significant majority are still worried – about a quarter on average for both news and government.

Figure 8. Proportion that say they have seen ‘a lot’ or ‘a great deal’ of false or misleading information about coronavirus from each in the last week

Q12. How much false or misleading information about coronavirus (COVID-19), if any, do you think you have seen on each of the following in the last week? Base: Total sample in each country.

Looking in more details at sources of information (see Figure 8), very significant minorities of respondents, especially in Argentina, Spain, and the US, are concerned about what they see as false or misleading information from news organisations; in Spain and the US, many are also concerned about misinformation from the national government (in Spain those on the right are particularly concerned about the left-wing government, in the US, those on the left particularly concerned about the right-wing government), and in several countries, many are concerned over misinformation from individual politicians. 

In the United States, as many respondents say they have seen false or misleading information from news organisations, the national government, and/or individual politicians as say the same about ordinary people they do not know (or about social media and messaging applications). The political patterns here are similar to those found for trust – people on the left worry about what they see as misinformation from individual politicians and the national government, people on the right about what they see as misinformation from news organisations. Argentina is again less polarised (though some on the right do worry about what they see as false or misleading information from the left-wing national government). Germany, again, exhibits more asymmetric polarisation, with few on the left and in the centre concerned, but more than a third of those on the right concerned about what they see as false or misleading information from news organisations and from the national government led by the Conservative Angela Merkel.

Turning to platforms for information second (see Figure 9), very significant minorities of respondents, especially in the US, are concerned about what they see as false or misleading information from across social media, video sharing sites, and messaging applications (and to a lesser extent search engines). Again, concern is particularly focused on social media and messaging applications, two types of platform that as many say they see misinformation on as say they see misinformation from ordinary people they don’t know (and from news organisations and the national government in the US).

The picture for video sites and search is more mixed. Across the six countries, concerns over false and misleading information from news organisations and video sites are about equal, with differences from country to country and limited gaps in most cases. In four of six countries, more people say they have come across false or misleading information from news organisations than say the same about search engines, and in two countries, more are concerned about misinformation from the national government than found via search engines.

Figure 9. Proportion that say they have seen ‘a lot’ or ‘a great deal’ of false or misleading information about coronavirus on each in the last week

Q13. How much false or misleading information about coronavirus (COVID-19), if any, do you think you have seen on each of the following in the last week? Base: Total sample in each country.


7. Helping people understand the coronavirus crisis.

Beyond what sources and platforms people use, which ones they trust, and which ones they feel spread misinformation, it is important to understand which sources people feel help them understand and navigate the crisis in a reasonable way. Here, we can use our survey data to compare what people say about how well they feel the news media in their country help them, and how well they feel their government helps them.

Figure 10. Attitudes towards media coverage of the coronavirus pandemic (average across countries)

Q14. To what extent do you agree or disagree with the following statements about the coronavirus (COVID-19)? Base: Total sample.

Looking at the news media first, a majority of respondents in every country say that the news media have helped them understand the crisis and explain what they can do in response. However, about one in three also say they feel the news media have exaggerated the pandemic. Figure 10 provides the average across six countries, Figure 11 data for each.

Figure 11. Attitudes towards media coverage of the coronavirus pandemic

Q14. To what extent do you agree or disagree with the following statements about the coronavirus (COVID-19)? Base: Total sample in each country.

Comparing these responses to how people rate their national government is interesting. Generally, more people say that news media have helped them understand the crisis and explain how they can react to it than say the government has done the same. The UK is the only exception. However, in almost every country, significantly more people say that the news media have exaggerated the crisis than say the same about the national government. Figure 12 provides the average across six countries, Figure 13 data for each.

Figure 12. Attitudes towards government communications around the coronavirus pandemic (average across countries)

Q15. To what extent do you agree or disagree with the following statements about the coronavirus (COVID-19)? Base: Total sample.

Figure 13. Attitudes towards government communications around the coronavirus pandemic

Q15. To what extent do you agree or disagree with the following statements about the coronavirus (COVID-19)? Base: Total sample in each country.

There are some partisan differences in how well people say the news media and the government has helped them understand the crisis, especially in the United States and in Spain, but, far more consistently (the UK is the only exception), young people and people with low levels of formal education are much less likely to say the news media or the national government has helped them understand the pandemic.


8. What do people know about coronavirus and how do they respond?

In the end, perhaps the most important thing is how knowledgeable people are about coronavirus, and how they respond to the pandemic. As long as there is no vaccine, COVID-19 can only be contained, and the curve flattened, if people understand the situation and react accordingly. Clearly, news and information are not the only factors that affect this, as people will come to the situation from different socio-economic positions, and other factors, including the relative precarity of their private and professional situation, the views of their family and friends, and the communities they live in, will shape both knowledge and behaviour.9 But in a public health crisis, access to accurate, relevant, reliable, and trusted information is literally a question of life and death.

Looking first at what people know, we have asked people a series of five factual questions about the coronavirus, four of them drawn from the WHO’s official coronavirus 'myth buster' site, and one the false claim that COVID-19 was made in a laboratory, a myth promoted by some high-level politicians, partisan pundits, and other prominent figures.10 For each statement, we ask respondents whether it is true, false, with an option to answer “don’t know”. How many correct answers people give to these five factual questions provides an indication of how much they know about the coronavirus.

The first thing to note is that most people do relatively well, with a clear majority answering three or more questions correctly (about three-quarters or more in every country apart from South Korea (58%) and the US (65%)). (See Table 3.) For most individual questions, the number of correct answers is strongly associated with level of education, as respondents with low levels of formal education give fewer correct responses. 

The main outlier here is the false claim that COVID-19 was made in a laboratory – here there are very significant differences by political viewpoint, especially in the United States, where some prominent right-wing voices have promoted this false notion. In four of six countries, the majority of respondents do not give the correct answer when asked about this. Almost a quarter of our respondents incorrectly say coronavirus was made in a laboratory, rising to about a third among those on the political right. 

In a few other instances, many give incorrect answers to some of our knowledge questions  – in Argentina, Germany, Spain, and South Korea, for example, a large minority wrongly believe that coronavirus cannot be transmitted in areas with hot weather.

Table 3. Answers to questions about coronavirus.

Q17. The following statements have been said about coronavirus (COVID-19). To the best of your knowledge, do you think each one of them is true or false? Base: Total sample in each country.

In order to try and disentangle the influence of different factors on how informed people are about the disease, we can use linear regression analysis to see the effect of different sources of information on coronavirus knowledge (measured as number of correct answers to the five questions) while controlling for socio-demographic variables like age, gender, education, and political leaning. This is cross-sectional analysis, so we cannot reveal causation, but it does allow us to estimate the association between coronavirus knowledge and sources of information like news organisations, government, health organisations, ordinary people, and different platforms. A summary of the analysis is shown in Table 4.

Table 4. Summary of OLS regression models where the dependent variable is coronavirus knowledge.

Note: Columns showing unstandardised beta coefficients followed by standard error. Search engines = Google, Bing, Naver and Daum; Social media = Facebook, Twitter, and Instagram; Video sites = YouTube; Messaging apps = WhatsApp, Facebook Messenger, Snapchat and Kakao Talk.  * p < .05, ** p < .01, *** p < .001.

The results in Table 4 show that in every country except Argentina and Spain, using news organisations as a source of information about coronavirus is associated with a statistically significant increase in coronavirus knowledge. In the US, where the effect is strongest, relying on news organisations for news and information about coronavirus is associated with around half a point on the 0-5 knowledge scale (B = .47, p < .001). Although the effects are not always statistically significant, relying on information from scientists, doctors and health experts, and health organisations is typically associated with having a higher level of coronavirus knowledge. The same is true in just two countries, Germany and the UK, for relying on information from the national government. 

No source of information is consistently and significantly associated with lower coronavirus knowledge, but in some countries, there is a negative effect from relying on ordinary people, video sites, and messaging apps. Where associations exist for search, they suggest a positive effect.

In addition to the associations between knowledge and different sources of information, it is also important to consider social and political factors. In every country apart from Spain, higher levels of education are associated with statistically significant slightly higher coronavirus knowledge. In four of six countries, those who identify as on the political right score significantly lower on coronavirus knowledge, but, far more consistently and dramatically, the sizable minority who, when asked to identify themselves politically on a left-right scale answer 'don’t know' (between 15% and 20% in most countries) score significantly lower. This could point to the effect of low political interest and/or alienation from established political parties. As is clear from Table 4, we find no consistent pattern around age or gender when controlling for other factors.

Like knowledge, behaviour is of course influenced by many other factors than news and information, including again private circumstances, socio-economic position, and the like, and survey responses are not always particularly reliable guides to what people actually do. That said, people’s responses to a set of coronavirus-relevant questions about what they have done yesterday suggests that most people are following social distancing recommendations and many are sheltering in place. (See Table 5.)

Table 5. Proportion that did each activity yesterday.

Q16. Which of the following, if any, did you do yesterday? Base: Total sample in each country.

Country-to-country variations here are probably in part to do with different government recommendations and policies in addition to any effect of news and information. But as said, other factors surely matter too – in the UK, for example, respondents with low formal education are much more likely to say they have gone in to work in the last week, a reminder that not everyone is in a position where they are able to work from home.


9. Conclusion.

We have found that people rely on many different sources of news and information about coronavirus, and access them via many different platforms. News use is up, news organisations remain central, and news media are trusted by a majority in all six countries. Most of our respondents also rely on various platforms, but regard the content they access via social media, video sites, and messaging applications (and to a lesser extent search engines) as much less trustworthy than information from news organisations. 

The majority say they feel news media have helped them understand the pandemic, and in fact our regression analysis suggests that they have, as relying on news organisations is associated with significantly higher levels of knowledge about coronavirus in most countries. Concerns that those who rely on other sources of information may be misinformed and more likely to hold false beliefs about the pandemic and the coronavirus are not borne out by our data and analysis in any consistent way. Social media use, for example, is not associated with lower levels of knowledge in any of the six countries, and where statistically significant associations exist with search, they are positive.

We have also found that people express very high levels of trust in scientists, doctors, and other experts, and often high levels of trust in health authorities and in global health organisations like the WHO – sources of information that many platform companies are currently promoting. Uncontroversial expertise is clearly having a comeback after years in which some politicians and parts of the public allegedly had had 'enough of experts'. 

How long this comeback will last if high-level politicians start publicly attacking and questioning experts, health authorities, and international organisations is an open question - our data collection was over before President Donald Trump stepped up his criticism on the WHO on 7 April, but even before this, just 51% of Americans on the political right rated it trustworthy, compared to 68% in the US public at large. Sustained attacks from political leaders on specific sources will almost certainly reduce trust in these, at least among those sympathetic to the politicians in question. As professional journalists have long known from their own experience, uncontroversial expertise may be highly and widely trusted, but more controversial expertise is not necessarily highly, let alone widely, trusted.

In addition to the sources and platforms of information that people rely on, we find that more basic social and political factors also influence understanding of the pandemic – in terms of social factors, those with lower levels of education know less about coronavirus than those with higher levels of education, in terms of political factors, those who are uninterested in politics, or alienated from established parties, know less about coronavirus than others. In some countries, there are also very significant partisan differences, as people with different political persuasions see the situation very differently, particularly in the United States. It is clear that high-level politicians bear a special responsibility for ensuring that they communicate clearly, accurately, and honestly about the pandemic. If they don’t, the consequences can be severe, as people will misunderstand the situation and put themselves, their loved ones, and their communities at greater risk.

A less widely recognised but equally concerning phenomenon is that young people and people with limited formal education in almost every country rely less on news organisations for news and information about coronavirus, trust both news media and the government less, are less likely to say that either the media or the government have helped them understand the pandemic, and that especially those with low education often know less about coronavirus than the rest of the population. Large minorities in every country do not engage with news (and do not trust it), and do not engage with government advice (and do not trust it), and, in turn, often know less about the crisis. This has to be addressed if important information is to reach  – and be taken seriously by – everyone.

So clearly, there is still much work to do – for news organisations, governments, platform companies, and others – to ensure that everyone knows that they need to know about coronavirus and on that basis can act to protect themselves, those they care about, and their wider communities. 

This report documents that news organisations play a crucial role in this. They are important in terms of providing people with news and information about coronavirus, information most people  trust (in many countries more than they trust the government), and information that we find helps people be more informed. But their importance goes well beyond this. Independent news organisations help people understand the crisis, even as they, at their best, also help hold national governments, health authorities, and other powerful actors to account for how they respond to the crisis, both in terms of the efficiency of their response and in terms of how transparent and truthful they are in their public communication around the crisis. This is crucial, as social distancing and similar public health measures to remain effective in the long-run needs to be credible as well as intelligible to the public at large.

We don’t have a vaccine for the virus yet, but at least we now know that access to accurate, relevant, reliable, trusted information from independent news media and other sources can help fight the “infodemic” and thus help people help themselves and their societies.


Footnotes.

1 https://www.who.int/dg/speeches/detail/munich-security-conference

2 The term 'infodemic' is not an established concept in social science research. A Google Scholar search suggests almost no use of it in academic research prior to 2020. The portmanteau was offered in 2003 in a newspaper comment by David J. Rothkopf to describe 'information epidemics', where 'a few facts, mixed with fear, speculation and rumor, [were] amplified and relayed swiftly worldwide by modern information technologies'. Note that, in Rothkopf’s original version, it suggested overreaction where infodemics supposedly 'have affected national and international economies, politics and even security in ways that are utterly disproportionate with the root realities' (Rothkopf 2003).

3As our data are cross-sectional, these correlations are not necessarily evidence of causality.

4 https://ourworldindata.org/grapher/total-covid-deaths-per-million?tab=chart&year=70&time=43..78&country=ARG+DEU+KOR+ESP+GBR+USA

5 https://www.nytimes.com/2020/04/05/world/europe/scientists-coronavirus-heroes.html

6In early April, Chartbeat estimated that coronavirus articles accounted for 29% of time spent reading news across their client network https://blog.chartbeat.com/2020/04/08/coronavirus-data-news-traffic-trends/

7 https://reutersinstitute.politics.ox.ac.uk/risj-review/how-journalists-can-help-stop-spread-coronavirus-outbreak/

8 https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)30461-X/fulltext/

9 E.g. the Pew Research Center in the United States has documented large differences in how people perceive and respond to the coronavirus not only on the basis of the media they use (Jurkowitz and Mitchell 2020), but also their partisan sympathies, religious beliefs, etc. (Parker et al. 2020).

10 https://www.who.int/emergencies/diseases/novel-coronavirus-2019/advice-for-public/myth-busters

11 https://www.nytimes.com/2020/04/07/us/politics/coronavirus-trump-who.html

References

Acknowledgements

We are very grateful to the Reuters Institute team and the team at YouGov –everyone involved in this report has worked hard early mornings, late evenings, over weekends and holidays, to make sure we could collect, analyse, and present the data as quickly as possible.

About the Authors

Rasmus Kleis Nielsen is the Director of the Reuters Institute for the Study of Journalism and Professor of Political Communication at the University of Oxford.

Richard Fletcher is a Senior Research Fellow at the Reuters Institute for the Study of Journalism and leads the Institute’s research team.

Nic Newman is a Senior Research Associate at the Reuters Institute for the Study of Journalism, where he is lead author of the annual Digital News Report and the annual Journalism Trends and Predictions report.

J. Scott Brennen is a Research Fellow at the Reuters Institute for the Study of Journalism and the Oxford Internet Institute at the University of Oxford.

Philip N. Howard is the Director of the Oxford Internet Institute and a Professor of Sociology, Information and International Affairs at the University of Oxford.


Report published by the Reuters Institute for the Study of Journalism as part of the Oxford Martin Programme on Misinformation, Science and Media, a three-year research collaboration between the Reuters Institute, the Oxford Internet Institute, and the Oxford Martin School.